diff --git "a/log/log-train-2023-03-31-18-51-54-3" "b/log/log-train-2023-03-31-18-51-54-3" new file mode 100644--- /dev/null +++ "b/log/log-train-2023-03-31-18-51-54-3" @@ -0,0 +1,25303 @@ +2023-03-31 18:51:54,787 INFO [train.py:975] (3/4) Training started +2023-03-31 18:51:54,787 INFO [train.py:985] (3/4) Device: cuda:3 +2023-03-31 18:51:54,825 INFO [train.py:994] (3/4) {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'warm_step': 2000, 'env_info': {'k2-version': '1.23.3', 'k2-build-type': 'Debug', 'k2-with-cuda': True, 'k2-git-sha1': '1c9950559223ec24d187f56bc424c3b43904bed3', 'k2-git-date': 'Thu Jan 26 22:00:26 2023', 'lhotse-version': '1.13.0.dev+git.ca98c73.dirty', 'torch-version': '2.0.0+cu117', 'torch-cuda-available': True, 'torch-cuda-version': '11.7', 'python-version': '3.8', 'icefall-git-branch': 'surt', 'icefall-git-sha1': '51e6a8a-dirty', 'icefall-git-date': 'Fri Mar 17 11:23:13 2023', 'icefall-path': '/exp/draj/mini_scale_2022/icefall', 'k2-path': '/exp/draj/mini_scale_2022/k2/k2/python/k2/__init__.py', 'lhotse-path': '/exp/draj/mini_scale_2022/lhotse/lhotse/__init__.py', 'hostname': 'r2n03', 'IP address': '10.1.2.3'}, 'world_size': 4, 'master_port': 54321, 'tensorboard': True, 'num_epochs': 30, 'start_epoch': 1, 'start_batch': 0, 'exp_dir': PosixPath('pruned_transducer_stateless7_streaming/exp/v2'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'base_lr': 0.05, 'lr_batches': 5000, 'lr_epochs': 3.5, 'context_size': 2, 'prune_range': 5, 'lm_scale': 0.25, 'am_scale': 0.0, 'simple_loss_scale': 0.5, 'seed': 42, 'print_diagnostics': False, 'inf_check': False, 'save_every_n': 2000, 'keep_last_k': 10, 'average_period': 200, 'use_fp16': True, 'num_encoder_layers': '2,2,2,2,2', 'feedforward_dims': '768,768,768,768,768', 'nhead': '8,8,8,8,8', 'encoder_dims': '256,256,256,256,256', 'attention_dims': '192,192,192,192,192', 'encoder_unmasked_dims': '192,192,192,192,192', 'zipformer_downsampling_factors': '1,2,4,8,2', 'cnn_module_kernels': '31,31,31,31,31', 'decoder_dim': 512, 'joiner_dim': 512, 'short_chunk_size': 50, 'num_left_chunks': 4, 'decode_chunk_len': 32, 'full_libri': True, 'manifest_dir': PosixPath('data/manifests'), 'max_duration': 800, 'bucketing_sampler': True, 'num_buckets': 30, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'drop_last': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'input_strategy': 'PrecomputedFeatures', 'blank_id': 0, 'vocab_size': 500} +2023-03-31 18:51:54,826 INFO [train.py:996] (3/4) About to create model +2023-03-31 18:51:55,665 INFO [zipformer.py:405] (3/4) At encoder stack 4, which has downsampling_factor=2, we will combine the outputs of layers 1 and 3, with downsampling_factors=2 and 8. +2023-03-31 18:51:55,676 INFO [train.py:1000] (3/4) Number of model parameters: 20697573 +2023-03-31 18:52:03,011 INFO [train.py:1019] (3/4) Using DDP +2023-03-31 18:52:03,648 INFO [asr_datamodule.py:429] (3/4) About to get the shuffled train-clean-100, train-clean-360 and train-other-500 cuts, combined with their reverberated versions +2023-03-31 18:52:03,689 INFO [asr_datamodule.py:224] (3/4) Enable MUSAN +2023-03-31 18:52:03,689 INFO [asr_datamodule.py:225] (3/4) About to get Musan cuts +2023-03-31 18:52:05,941 INFO [asr_datamodule.py:249] (3/4) Enable SpecAugment +2023-03-31 18:52:05,941 INFO [asr_datamodule.py:250] (3/4) Time warp factor: 80 +2023-03-31 18:52:05,941 INFO [asr_datamodule.py:260] (3/4) Num frame mask: 10 +2023-03-31 18:52:05,941 INFO [asr_datamodule.py:273] (3/4) About to create train dataset +2023-03-31 18:52:05,941 INFO [asr_datamodule.py:300] (3/4) Using DynamicBucketingSampler. +2023-03-31 18:52:08,275 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-03-31 18:52:08,739 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-03-31 18:52:09,020 INFO [asr_datamodule.py:315] (3/4) About to create train dataloader +2023-03-31 18:52:09,021 INFO [asr_datamodule.py:440] (3/4) About to get dev-clean cuts +2023-03-31 18:52:09,022 INFO [asr_datamodule.py:447] (3/4) About to get dev-other cuts +2023-03-31 18:52:09,023 INFO [asr_datamodule.py:346] (3/4) About to create dev dataset +2023-03-31 18:52:09,471 INFO [asr_datamodule.py:363] (3/4) About to create dev dataloader +2023-03-31 18:52:23,899 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-03-31 18:52:24,322 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-03-31 18:52:36,261 INFO [train.py:903] (3/4) Epoch 1, batch 0, loss[loss=7.077, simple_loss=6.406, pruned_loss=6.699, over 19303.00 frames. ], tot_loss[loss=7.077, simple_loss=6.406, pruned_loss=6.699, over 19303.00 frames. ], batch size: 44, lr: 2.50e-02, grad_scale: 2.0 +2023-03-31 18:52:36,261 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 18:52:49,135 INFO [train.py:937] (3/4) Epoch 1, validation: loss=6.888, simple_loss=6.229, pruned_loss=6.575, over 944034.00 frames. +2023-03-31 18:52:49,136 INFO [train.py:938] (3/4) Maximum memory allocated so far is 11586MB +2023-03-31 18:53:02,997 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-03-31 18:53:58,980 INFO [train.py:903] (3/4) Epoch 1, batch 50, loss[loss=1.355, simple_loss=1.199, pruned_loss=1.39, over 19320.00 frames. ], tot_loss[loss=2.158, simple_loss=1.949, pruned_loss=1.999, over 874003.52 frames. ], batch size: 70, lr: 2.75e-02, grad_scale: 0.125 +2023-03-31 18:54:00,856 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 18:54:25,976 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1992, 4.1942, 4.1928, 4.1986, 4.1974, 4.1933, 4.1975, 4.1975], + device='cuda:3'), covar=tensor([0.0069, 0.0043, 0.0079, 0.0073, 0.0193, 0.0106, 0.0119, 0.0103], + device='cuda:3'), in_proj_covar=tensor([0.0013, 0.0013, 0.0013, 0.0014, 0.0014, 0.0013, 0.0013, 0.0013], + device='cuda:3'), out_proj_covar=tensor([8.6617e-06, 8.9851e-06, 8.8598e-06, 8.8768e-06, 8.9125e-06, 8.9549e-06, + 8.8366e-06, 8.7795e-06], device='cuda:3') +2023-03-31 18:54:36,611 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 18:54:41,879 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-03-31 18:54:47,823 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=4.44 vs. limit=2.0 +2023-03-31 18:54:59,744 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=42.44 vs. limit=5.0 +2023-03-31 18:55:06,318 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=17.55 vs. limit=2.0 +2023-03-31 18:55:11,352 INFO [train.py:903] (3/4) Epoch 1, batch 100, loss[loss=1.141, simple_loss=0.9946, pruned_loss=1.181, over 13389.00 frames. ], tot_loss[loss=1.632, simple_loss=1.453, pruned_loss=1.614, over 1525776.39 frames. ], batch size: 136, lr: 3.00e-02, grad_scale: 0.25 +2023-03-31 18:55:11,617 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 18:55:17,835 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 9.479e+01 1.678e+02 3.237e+02 1.260e+03 8.630e+04, threshold=6.475e+02, percent-clipped=0.0 +2023-03-31 18:55:25,893 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-03-31 18:55:43,476 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=6.18 vs. limit=2.0 +2023-03-31 18:56:20,076 INFO [train.py:903] (3/4) Epoch 1, batch 150, loss[loss=1.034, simple_loss=0.882, pruned_loss=1.101, over 19442.00 frames. ], tot_loss[loss=1.391, simple_loss=1.223, pruned_loss=1.423, over 2036100.23 frames. ], batch size: 70, lr: 3.25e-02, grad_scale: 0.25 +2023-03-31 18:57:32,422 INFO [train.py:903] (3/4) Epoch 1, batch 200, loss[loss=1.032, simple_loss=0.8756, pruned_loss=1.053, over 19610.00 frames. ], tot_loss[loss=1.255, simple_loss=1.093, pruned_loss=1.291, over 2435973.89 frames. ], batch size: 57, lr: 3.50e-02, grad_scale: 0.5 +2023-03-31 18:57:33,129 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-03-31 18:57:39,439 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 7.067e+01 1.186e+02 1.653e+02 2.090e+02 5.158e+02, threshold=3.307e+02, percent-clipped=0.0 +2023-03-31 18:57:48,568 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-31 18:58:43,211 INFO [train.py:903] (3/4) Epoch 1, batch 250, loss[loss=1.004, simple_loss=0.8457, pruned_loss=0.9909, over 19586.00 frames. ], tot_loss[loss=1.167, simple_loss=1.009, pruned_loss=1.192, over 2749483.49 frames. ], batch size: 61, lr: 3.75e-02, grad_scale: 0.5 +2023-03-31 18:59:51,889 INFO [train.py:903] (3/4) Epoch 1, batch 300, loss[loss=0.9671, simple_loss=0.8077, pruned_loss=0.9349, over 19679.00 frames. ], tot_loss[loss=1.102, simple_loss=0.946, pruned_loss=1.114, over 2992759.98 frames. ], batch size: 58, lr: 4.00e-02, grad_scale: 1.0 +2023-03-31 18:59:56,691 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 8.717e+01 1.166e+02 1.521e+02 1.991e+02 3.277e+02, threshold=3.043e+02, percent-clipped=0.0 +2023-03-31 18:59:58,411 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=306.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:00:09,467 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=314.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:00:49,610 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.11 vs. limit=2.0 +2023-03-31 19:00:58,531 INFO [train.py:903] (3/4) Epoch 1, batch 350, loss[loss=0.9795, simple_loss=0.8117, pruned_loss=0.9264, over 19346.00 frames. ], tot_loss[loss=1.056, simple_loss=0.8993, pruned_loss=1.054, over 3193706.32 frames. ], batch size: 66, lr: 4.25e-02, grad_scale: 1.0 +2023-03-31 19:01:05,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-03-31 19:01:47,854 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4088, 4.0422, 4.2233, 4.2358, 4.4600, 4.4360, 4.4329, 4.4024], + device='cuda:3'), covar=tensor([0.0124, 0.0192, 0.0141, 0.0187, 0.0123, 0.0095, 0.0107, 0.0117], + device='cuda:3'), in_proj_covar=tensor([0.0013, 0.0014, 0.0013, 0.0014, 0.0014, 0.0013, 0.0014, 0.0013], + device='cuda:3'), out_proj_covar=tensor([8.6472e-06, 9.1172e-06, 8.9155e-06, 8.9413e-06, 8.7632e-06, 8.8641e-06, + 8.7775e-06, 8.7957e-06], device='cuda:3') +2023-03-31 19:02:08,273 INFO [train.py:903] (3/4) Epoch 1, batch 400, loss[loss=0.9656, simple_loss=0.7948, pruned_loss=0.8932, over 19545.00 frames. ], tot_loss[loss=1.026, simple_loss=0.8681, pruned_loss=1.008, over 3321103.40 frames. ], batch size: 54, lr: 4.50e-02, grad_scale: 2.0 +2023-03-31 19:02:13,391 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 9.539e+01 1.278e+02 1.546e+02 1.978e+02 5.474e+02, threshold=3.091e+02, percent-clipped=7.0 +2023-03-31 19:02:13,799 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=405.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:02:33,793 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=421.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:03:05,099 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=445.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:03:12,885 INFO [train.py:903] (3/4) Epoch 1, batch 450, loss[loss=0.9599, simple_loss=0.7894, pruned_loss=0.8566, over 19608.00 frames. ], tot_loss[loss=0.9995, simple_loss=0.8405, pruned_loss=0.9638, over 3444940.34 frames. ], batch size: 61, lr: 4.75e-02, grad_scale: 2.0 +2023-03-31 19:03:41,214 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=4.25 vs. limit=2.0 +2023-03-31 19:03:51,035 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-03-31 19:03:51,734 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-03-31 19:04:19,396 INFO [train.py:903] (3/4) Epoch 1, batch 500, loss[loss=0.8779, simple_loss=0.7256, pruned_loss=0.7473, over 19395.00 frames. ], tot_loss[loss=0.9832, simple_loss=0.8237, pruned_loss=0.9259, over 3522392.64 frames. ], batch size: 48, lr: 4.99e-02, grad_scale: 2.0 +2023-03-31 19:04:25,207 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 9.539e+01 1.386e+02 1.860e+02 2.529e+02 4.736e+02, threshold=3.719e+02, percent-clipped=12.0 +2023-03-31 19:05:27,838 INFO [train.py:903] (3/4) Epoch 1, batch 550, loss[loss=0.8769, simple_loss=0.7334, pruned_loss=0.7027, over 19672.00 frames. ], tot_loss[loss=0.9673, simple_loss=0.8088, pruned_loss=0.8861, over 3587717.64 frames. ], batch size: 53, lr: 4.98e-02, grad_scale: 2.0 +2023-03-31 19:05:41,522 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=560.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:06:04,508 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=580.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:06:12,408 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=586.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:06:32,557 INFO [train.py:903] (3/4) Epoch 1, batch 600, loss[loss=0.8956, simple_loss=0.7518, pruned_loss=0.6932, over 19720.00 frames. ], tot_loss[loss=0.9435, simple_loss=0.7891, pruned_loss=0.8385, over 3644417.99 frames. ], batch size: 63, lr: 4.98e-02, grad_scale: 2.0 +2023-03-31 19:06:36,898 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.910e+02 4.086e+02 6.136e+02 1.097e+03, threshold=8.173e+02, percent-clipped=60.0 +2023-03-31 19:06:41,681 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=608.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:06:43,962 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=2.18 vs. limit=2.0 +2023-03-31 19:06:47,614 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=3.69 vs. limit=2.0 +2023-03-31 19:07:11,460 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-03-31 19:07:38,089 INFO [train.py:903] (3/4) Epoch 1, batch 650, loss[loss=0.7179, simple_loss=0.6089, pruned_loss=0.5294, over 19352.00 frames. ], tot_loss[loss=0.9193, simple_loss=0.7701, pruned_loss=0.7918, over 3680038.14 frames. ], batch size: 47, lr: 4.98e-02, grad_scale: 2.0 +2023-03-31 19:07:47,307 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=658.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:08:13,494 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=677.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:08:41,843 INFO [train.py:903] (3/4) Epoch 1, batch 700, loss[loss=0.7126, simple_loss=0.6074, pruned_loss=0.5094, over 15511.00 frames. ], tot_loss[loss=0.8901, simple_loss=0.7477, pruned_loss=0.7424, over 3718904.32 frames. ], batch size: 34, lr: 4.98e-02, grad_scale: 2.0 +2023-03-31 19:08:43,434 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=702.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:08:46,601 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.000e+02 5.173e+02 6.580e+02 8.914e+02 3.039e+03, threshold=1.316e+03, percent-clipped=29.0 +2023-03-31 19:09:43,565 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=749.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:09:45,325 INFO [train.py:903] (3/4) Epoch 1, batch 750, loss[loss=0.8078, simple_loss=0.6888, pruned_loss=0.5669, over 19651.00 frames. ], tot_loss[loss=0.8617, simple_loss=0.7266, pruned_loss=0.6956, over 3747947.31 frames. ], batch size: 58, lr: 4.97e-02, grad_scale: 2.0 +2023-03-31 19:10:14,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=773.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:10:48,968 INFO [train.py:903] (3/4) Epoch 1, batch 800, loss[loss=0.7224, simple_loss=0.6229, pruned_loss=0.487, over 19846.00 frames. ], tot_loss[loss=0.8357, simple_loss=0.7072, pruned_loss=0.6544, over 3765221.58 frames. ], batch size: 52, lr: 4.97e-02, grad_scale: 4.0 +2023-03-31 19:10:53,085 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.110e+02 5.827e+02 7.991e+02 1.030e+03 2.888e+03, threshold=1.598e+03, percent-clipped=14.0 +2023-03-31 19:11:01,619 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-03-31 19:11:08,448 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=816.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:11:23,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9111, 0.8740, 1.0417, 0.9486, 0.8724, 0.8623, 0.5636, 0.7401], + device='cuda:3'), covar=tensor([0.5532, 0.6191, 0.5872, 0.8007, 0.7043, 0.7320, 0.9413, 0.6373], + device='cuda:3'), in_proj_covar=tensor([0.0056, 0.0059, 0.0061, 0.0067, 0.0064, 0.0063, 0.0071, 0.0057], + device='cuda:3'), out_proj_covar=tensor([4.1324e-05, 3.7400e-05, 3.8572e-05, 4.7434e-05, 4.3410e-05, 3.9747e-05, + 5.0992e-05, 3.7452e-05], device='cuda:3') +2023-03-31 19:11:40,315 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=841.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:11:52,170 INFO [train.py:903] (3/4) Epoch 1, batch 850, loss[loss=0.7706, simple_loss=0.6636, pruned_loss=0.5139, over 19735.00 frames. ], tot_loss[loss=0.8073, simple_loss=0.6859, pruned_loss=0.6137, over 3780006.56 frames. ], batch size: 63, lr: 4.96e-02, grad_scale: 4.0 +2023-03-31 19:12:10,443 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=864.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:12:12,289 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=865.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:12:15,720 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=5.18 vs. limit=5.0 +2023-03-31 19:12:43,340 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-03-31 19:12:51,333 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2825, 1.5032, 2.5942, 2.0387, 2.1583, 2.5438, 2.5565, 2.5166], + device='cuda:3'), covar=tensor([0.4292, 0.7630, 0.3867, 0.4796, 0.3235, 0.2659, 0.3149, 0.3112], + device='cuda:3'), in_proj_covar=tensor([0.0061, 0.0065, 0.0058, 0.0064, 0.0061, 0.0053, 0.0062, 0.0054], + device='cuda:3'), out_proj_covar=tensor([3.9114e-05, 4.7181e-05, 3.5075e-05, 4.5042e-05, 3.7678e-05, 3.0497e-05, + 3.8307e-05, 3.4056e-05], device='cuda:3') +2023-03-31 19:12:54,612 INFO [train.py:903] (3/4) Epoch 1, batch 900, loss[loss=0.623, simple_loss=0.5409, pruned_loss=0.4035, over 19099.00 frames. ], tot_loss[loss=0.7853, simple_loss=0.6698, pruned_loss=0.5801, over 3803596.91 frames. ], batch size: 42, lr: 4.96e-02, grad_scale: 4.0 +2023-03-31 19:12:59,594 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.676e+02 6.072e+02 7.456e+02 9.579e+02 1.181e+04, threshold=1.491e+03, percent-clipped=3.0 +2023-03-31 19:13:21,569 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=924.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:13:30,349 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=930.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:13:53,705 INFO [train.py:903] (3/4) Epoch 1, batch 950, loss[loss=0.6101, simple_loss=0.5342, pruned_loss=0.3842, over 19763.00 frames. ], tot_loss[loss=0.7656, simple_loss=0.6553, pruned_loss=0.5508, over 3805964.47 frames. ], batch size: 46, lr: 4.96e-02, grad_scale: 4.0 +2023-03-31 19:13:53,739 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-03-31 19:13:56,064 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=952.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:14:51,779 INFO [train.py:903] (3/4) Epoch 1, batch 1000, loss[loss=0.6795, simple_loss=0.5952, pruned_loss=0.4237, over 19597.00 frames. ], tot_loss[loss=0.7491, simple_loss=0.6435, pruned_loss=0.5257, over 3813667.56 frames. ], batch size: 52, lr: 4.95e-02, grad_scale: 4.0 +2023-03-31 19:14:56,987 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.794e+02 5.980e+02 7.509e+02 1.052e+03 2.029e+03, threshold=1.502e+03, percent-clipped=4.0 +2023-03-31 19:15:25,818 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1029.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:15:38,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1039.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:15:41,715 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-03-31 19:15:45,014 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1045.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:15:52,657 INFO [train.py:903] (3/4) Epoch 1, batch 1050, loss[loss=0.6339, simple_loss=0.5601, pruned_loss=0.3856, over 19730.00 frames. ], tot_loss[loss=0.7304, simple_loss=0.6305, pruned_loss=0.4999, over 3818770.20 frames. ], batch size: 51, lr: 4.95e-02, grad_scale: 4.0 +2023-03-31 19:15:56,697 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1054.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:16:12,569 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1067.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:16:20,721 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-03-31 19:16:53,091 INFO [train.py:903] (3/4) Epoch 1, batch 1100, loss[loss=0.7131, simple_loss=0.6114, pruned_loss=0.4534, over 17940.00 frames. ], tot_loss[loss=0.7157, simple_loss=0.6198, pruned_loss=0.4793, over 3830342.97 frames. ], batch size: 83, lr: 4.94e-02, grad_scale: 4.0 +2023-03-31 19:16:57,396 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.036e+02 7.117e+02 8.563e+02 1.068e+03 2.368e+03, threshold=1.713e+03, percent-clipped=4.0 +2023-03-31 19:17:14,922 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1120.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:17:45,068 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1145.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:17:51,773 INFO [train.py:903] (3/4) Epoch 1, batch 1150, loss[loss=0.658, simple_loss=0.5725, pruned_loss=0.4048, over 19480.00 frames. ], tot_loss[loss=0.7004, simple_loss=0.6089, pruned_loss=0.4596, over 3826876.39 frames. ], batch size: 49, lr: 4.94e-02, grad_scale: 4.0 +2023-03-31 19:18:13,023 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1171.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:18:47,606 INFO [train.py:903] (3/4) Epoch 1, batch 1200, loss[loss=0.6695, simple_loss=0.5957, pruned_loss=0.3945, over 19714.00 frames. ], tot_loss[loss=0.6864, simple_loss=0.5983, pruned_loss=0.4427, over 3827428.76 frames. ], batch size: 63, lr: 4.93e-02, grad_scale: 8.0 +2023-03-31 19:18:52,216 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.229e+02 7.433e+02 9.314e+02 1.239e+03 3.000e+03, threshold=1.863e+03, percent-clipped=16.0 +2023-03-31 19:18:56,109 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=1209.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:19:16,395 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-03-31 19:19:42,385 INFO [train.py:903] (3/4) Epoch 1, batch 1250, loss[loss=0.5681, simple_loss=0.505, pruned_loss=0.3334, over 19315.00 frames. ], tot_loss[loss=0.6725, simple_loss=0.5889, pruned_loss=0.4256, over 3830534.52 frames. ], batch size: 44, lr: 4.92e-02, grad_scale: 8.0 +2023-03-31 19:20:32,325 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1295.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:20:39,090 INFO [train.py:903] (3/4) Epoch 1, batch 1300, loss[loss=0.7669, simple_loss=0.6349, pruned_loss=0.4938, over 19688.00 frames. ], tot_loss[loss=0.6656, simple_loss=0.584, pruned_loss=0.4151, over 3814461.74 frames. ], batch size: 60, lr: 4.92e-02, grad_scale: 8.0 +2023-03-31 19:20:39,475 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1301.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:20:43,725 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.522e+02 7.699e+02 1.048e+03 1.379e+03 4.741e+03, threshold=2.097e+03, percent-clipped=13.0 +2023-03-31 19:21:00,035 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1320.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:21:03,692 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1323.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:21:04,879 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1324.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:21:06,836 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1326.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:21:33,403 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1348.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:21:35,927 INFO [train.py:903] (3/4) Epoch 1, batch 1350, loss[loss=0.5934, simple_loss=0.528, pruned_loss=0.3443, over 19673.00 frames. ], tot_loss[loss=0.6528, simple_loss=0.5751, pruned_loss=0.4009, over 3805375.39 frames. ], batch size: 53, lr: 4.91e-02, grad_scale: 8.0 +2023-03-31 19:22:31,180 INFO [train.py:903] (3/4) Epoch 1, batch 1400, loss[loss=0.5378, simple_loss=0.5005, pruned_loss=0.291, over 19767.00 frames. ], tot_loss[loss=0.6393, simple_loss=0.5662, pruned_loss=0.3864, over 3815595.34 frames. ], batch size: 54, lr: 4.91e-02, grad_scale: 8.0 +2023-03-31 19:22:35,196 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.777e+02 7.620e+02 9.515e+02 1.230e+03 4.278e+03, threshold=1.903e+03, percent-clipped=3.0 +2023-03-31 19:23:24,308 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-03-31 19:23:25,237 INFO [train.py:903] (3/4) Epoch 1, batch 1450, loss[loss=0.5174, simple_loss=0.481, pruned_loss=0.28, over 19593.00 frames. ], tot_loss[loss=0.6269, simple_loss=0.5571, pruned_loss=0.3741, over 3805106.09 frames. ], batch size: 52, lr: 4.90e-02, grad_scale: 8.0 +2023-03-31 19:23:26,469 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1452.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:24:18,027 INFO [train.py:903] (3/4) Epoch 1, batch 1500, loss[loss=0.548, simple_loss=0.5053, pruned_loss=0.2996, over 19604.00 frames. ], tot_loss[loss=0.6201, simple_loss=0.5528, pruned_loss=0.3657, over 3804530.53 frames. ], batch size: 50, lr: 4.89e-02, grad_scale: 8.0 +2023-03-31 19:24:23,061 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.214e+02 9.104e+02 1.060e+03 1.370e+03 5.981e+03, threshold=2.119e+03, percent-clipped=12.0 +2023-03-31 19:24:35,378 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=1515.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:24:52,460 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0535, 1.5494, 1.9822, 1.8383, 1.1566, 1.7706, 0.7405, 1.2794], + device='cuda:3'), covar=tensor([0.1695, 0.2352, 0.2132, 0.3198, 0.3904, 0.2827, 0.7146, 0.3230], + device='cuda:3'), in_proj_covar=tensor([0.0087, 0.0082, 0.0094, 0.0102, 0.0104, 0.0093, 0.0138, 0.0103], + device='cuda:3'), out_proj_covar=tensor([5.9787e-05, 4.9858e-05, 6.0910e-05, 7.3433e-05, 7.5147e-05, 6.2697e-05, + 1.0091e-04, 7.3868e-05], device='cuda:3') +2023-03-31 19:25:14,429 INFO [train.py:903] (3/4) Epoch 1, batch 1550, loss[loss=0.6387, simple_loss=0.5659, pruned_loss=0.3663, over 19535.00 frames. ], tot_loss[loss=0.6103, simple_loss=0.5468, pruned_loss=0.3552, over 3817228.39 frames. ], batch size: 54, lr: 4.89e-02, grad_scale: 8.0 +2023-03-31 19:25:24,673 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-31 19:25:43,604 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1580.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:25:47,367 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1584.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:26:05,886 INFO [train.py:903] (3/4) Epoch 1, batch 1600, loss[loss=0.5786, simple_loss=0.5347, pruned_loss=0.3144, over 19679.00 frames. ], tot_loss[loss=0.603, simple_loss=0.5419, pruned_loss=0.3475, over 3814099.19 frames. ], batch size: 59, lr: 4.88e-02, grad_scale: 8.0 +2023-03-31 19:26:10,808 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.811e+02 9.198e+02 1.152e+03 1.497e+03 2.578e+03, threshold=2.303e+03, percent-clipped=3.0 +2023-03-31 19:26:11,232 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1605.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:26:25,452 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-03-31 19:26:35,468 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1629.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:26:36,305 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1630.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:26:57,907 INFO [train.py:903] (3/4) Epoch 1, batch 1650, loss[loss=0.6326, simple_loss=0.5755, pruned_loss=0.3496, over 19671.00 frames. ], tot_loss[loss=0.5958, simple_loss=0.5378, pruned_loss=0.3397, over 3816120.20 frames. ], batch size: 60, lr: 4.87e-02, grad_scale: 8.0 +2023-03-31 19:27:52,335 INFO [train.py:903] (3/4) Epoch 1, batch 1700, loss[loss=0.5354, simple_loss=0.494, pruned_loss=0.2905, over 19522.00 frames. ], tot_loss[loss=0.5899, simple_loss=0.5349, pruned_loss=0.333, over 3827695.72 frames. ], batch size: 54, lr: 4.86e-02, grad_scale: 8.0 +2023-03-31 19:27:56,182 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.719e+02 9.402e+02 1.223e+03 1.535e+03 2.582e+03, threshold=2.447e+03, percent-clipped=3.0 +2023-03-31 19:28:26,633 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-03-31 19:28:33,164 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8223, 1.5478, 1.6529, 3.0672, 3.9571, 2.2233, 3.0576, 3.4278], + device='cuda:3'), covar=tensor([0.0472, 0.4525, 0.6759, 0.2284, 0.0764, 0.6087, 0.1176, 0.1415], + device='cuda:3'), in_proj_covar=tensor([0.0091, 0.0143, 0.0183, 0.0129, 0.0129, 0.0220, 0.0125, 0.0115], + device='cuda:3'), out_proj_covar=tensor([5.2270e-05, 9.9517e-05, 1.2725e-04, 9.1926e-05, 7.6963e-05, 1.4293e-04, + 8.1368e-05, 7.6814e-05], device='cuda:3') +2023-03-31 19:28:46,832 INFO [train.py:903] (3/4) Epoch 1, batch 1750, loss[loss=0.4959, simple_loss=0.4671, pruned_loss=0.2628, over 19726.00 frames. ], tot_loss[loss=0.5817, simple_loss=0.5301, pruned_loss=0.3252, over 3827032.20 frames. ], batch size: 46, lr: 4.86e-02, grad_scale: 8.0 +2023-03-31 19:29:38,259 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=1796.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:29:43,455 INFO [train.py:903] (3/4) Epoch 1, batch 1800, loss[loss=0.5446, simple_loss=0.5126, pruned_loss=0.2887, over 19699.00 frames. ], tot_loss[loss=0.5757, simple_loss=0.5257, pruned_loss=0.3199, over 3822081.22 frames. ], batch size: 59, lr: 4.85e-02, grad_scale: 8.0 +2023-03-31 19:29:47,626 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.927e+02 9.266e+02 1.209e+03 1.539e+03 2.564e+03, threshold=2.418e+03, percent-clipped=2.0 +2023-03-31 19:30:36,181 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-03-31 19:30:40,566 INFO [train.py:903] (3/4) Epoch 1, batch 1850, loss[loss=0.5217, simple_loss=0.4942, pruned_loss=0.2746, over 19521.00 frames. ], tot_loss[loss=0.5701, simple_loss=0.5229, pruned_loss=0.3143, over 3832692.97 frames. ], batch size: 54, lr: 4.84e-02, grad_scale: 8.0 +2023-03-31 19:30:45,775 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1856.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:31:08,136 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1875.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:31:12,894 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-03-31 19:31:20,292 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1886.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:31:22,097 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1888.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:31:36,176 INFO [train.py:903] (3/4) Epoch 1, batch 1900, loss[loss=0.4852, simple_loss=0.4853, pruned_loss=0.2413, over 19688.00 frames. ], tot_loss[loss=0.5637, simple_loss=0.519, pruned_loss=0.3087, over 3810251.14 frames. ], batch size: 60, lr: 4.83e-02, grad_scale: 8.0 +2023-03-31 19:31:40,292 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.070e+02 9.078e+02 1.104e+03 1.499e+03 2.754e+03, threshold=2.207e+03, percent-clipped=2.0 +2023-03-31 19:31:47,483 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1911.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:31:47,504 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1911.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:31:52,293 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-03-31 19:31:57,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-03-31 19:32:06,359 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=1928.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:32:19,677 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-03-31 19:32:31,502 INFO [train.py:903] (3/4) Epoch 1, batch 1950, loss[loss=0.5276, simple_loss=0.5077, pruned_loss=0.2735, over 19779.00 frames. ], tot_loss[loss=0.5583, simple_loss=0.5154, pruned_loss=0.3042, over 3801877.59 frames. ], batch size: 56, lr: 4.83e-02, grad_scale: 8.0 +2023-03-31 19:32:57,190 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=1973.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:33:29,169 INFO [train.py:903] (3/4) Epoch 1, batch 2000, loss[loss=0.5919, simple_loss=0.5391, pruned_loss=0.3223, over 17434.00 frames. ], tot_loss[loss=0.5554, simple_loss=0.5146, pruned_loss=0.3009, over 3795301.12 frames. ], batch size: 101, lr: 4.82e-02, grad_scale: 8.0 +2023-03-31 19:33:33,541 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.917e+02 1.007e+03 1.260e+03 1.703e+03 3.255e+03, threshold=2.521e+03, percent-clipped=11.0 +2023-03-31 19:34:04,902 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1658, 1.5290, 1.9460, 1.3531, 1.9862, 2.8498, 2.4229, 1.8742], + device='cuda:3'), covar=tensor([0.2902, 0.1803, 0.2877, 0.3281, 0.2211, 0.0529, 0.1367, 0.2707], + device='cuda:3'), in_proj_covar=tensor([0.0094, 0.0074, 0.0095, 0.0106, 0.0106, 0.0053, 0.0081, 0.0108], + device='cuda:3'), out_proj_covar=tensor([6.2455e-05, 4.8035e-05, 6.3418e-05, 7.3631e-05, 7.2993e-05, 3.0281e-05, + 5.5912e-05, 7.1019e-05], device='cuda:3') +2023-03-31 19:34:18,983 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2043.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:34:25,201 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-03-31 19:34:27,597 INFO [train.py:903] (3/4) Epoch 1, batch 2050, loss[loss=0.5009, simple_loss=0.4922, pruned_loss=0.2548, over 19671.00 frames. ], tot_loss[loss=0.5471, simple_loss=0.5104, pruned_loss=0.294, over 3793023.90 frames. ], batch size: 58, lr: 4.81e-02, grad_scale: 16.0 +2023-03-31 19:34:43,698 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-03-31 19:34:43,733 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-03-31 19:34:54,258 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8024, 2.0519, 2.7261, 1.6033, 2.2401, 3.9762, 3.5296, 3.7114], + device='cuda:3'), covar=tensor([0.2452, 0.2058, 0.1234, 0.2859, 0.1250, 0.0138, 0.0309, 0.0323], + device='cuda:3'), in_proj_covar=tensor([0.0127, 0.0109, 0.0095, 0.0128, 0.0094, 0.0061, 0.0077, 0.0066], + device='cuda:3'), out_proj_covar=tensor([8.1477e-05, 7.0637e-05, 5.8991e-05, 8.1920e-05, 5.9912e-05, 3.1654e-05, + 4.1392e-05, 3.4848e-05], device='cuda:3') +2023-03-31 19:35:06,985 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-03-31 19:35:12,786 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2088.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:35:27,385 INFO [train.py:903] (3/4) Epoch 1, batch 2100, loss[loss=0.5812, simple_loss=0.5395, pruned_loss=0.3114, over 19514.00 frames. ], tot_loss[loss=0.5382, simple_loss=0.5056, pruned_loss=0.2871, over 3798742.34 frames. ], batch size: 56, lr: 4.80e-02, grad_scale: 16.0 +2023-03-31 19:35:31,661 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.951e+02 9.211e+02 1.091e+03 1.524e+03 2.851e+03, threshold=2.182e+03, percent-clipped=6.0 +2023-03-31 19:35:49,309 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8165, 2.1435, 2.8044, 1.6266, 2.1407, 4.0233, 4.0525, 3.6483], + device='cuda:3'), covar=tensor([0.2652, 0.2220, 0.1281, 0.2987, 0.1498, 0.0242, 0.0245, 0.0490], + device='cuda:3'), in_proj_covar=tensor([0.0137, 0.0118, 0.0102, 0.0137, 0.0103, 0.0065, 0.0081, 0.0070], + device='cuda:3'), out_proj_covar=tensor([8.8146e-05, 7.6440e-05, 6.3669e-05, 8.8146e-05, 6.5860e-05, 3.3896e-05, + 4.3853e-05, 3.6932e-05], device='cuda:3') +2023-03-31 19:35:56,700 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-03-31 19:36:17,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-03-31 19:36:24,969 INFO [train.py:903] (3/4) Epoch 1, batch 2150, loss[loss=0.4576, simple_loss=0.4535, pruned_loss=0.2309, over 19620.00 frames. ], tot_loss[loss=0.5298, simple_loss=0.5012, pruned_loss=0.2805, over 3803119.06 frames. ], batch size: 50, lr: 4.79e-02, grad_scale: 16.0 +2023-03-31 19:36:45,574 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2167.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:37:13,945 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2192.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:37:19,351 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6498, 3.1804, 2.0092, 2.7386, 1.4826, 3.4738, 2.9719, 3.1045], + device='cuda:3'), covar=tensor([0.0695, 0.1382, 0.2956, 0.0873, 0.3014, 0.0551, 0.0895, 0.0756], + device='cuda:3'), in_proj_covar=tensor([0.0150, 0.0191, 0.0190, 0.0131, 0.0206, 0.0116, 0.0130, 0.0118], + device='cuda:3'), out_proj_covar=tensor([1.2349e-04, 1.4822e-04, 1.2768e-04, 9.5103e-05, 1.4342e-04, 8.3670e-05, + 9.3502e-05, 8.8158e-05], device='cuda:3') +2023-03-31 19:37:24,837 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=2200.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:37:25,819 INFO [train.py:903] (3/4) Epoch 1, batch 2200, loss[loss=0.4543, simple_loss=0.4589, pruned_loss=0.2248, over 19671.00 frames. ], tot_loss[loss=0.5223, simple_loss=0.4969, pruned_loss=0.2749, over 3818536.49 frames. ], batch size: 53, lr: 4.78e-02, grad_scale: 16.0 +2023-03-31 19:37:31,694 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.347e+02 9.332e+02 1.145e+03 1.435e+03 3.303e+03, threshold=2.290e+03, percent-clipped=7.0 +2023-03-31 19:37:49,524 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=2219.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:38:03,609 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=2232.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:38:05,862 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0741, 3.4956, 1.9345, 3.2460, 1.4866, 3.8295, 3.4263, 3.4624], + device='cuda:3'), covar=tensor([0.0479, 0.0978, 0.2569, 0.0571, 0.2548, 0.0400, 0.0657, 0.0736], + device='cuda:3'), in_proj_covar=tensor([0.0153, 0.0188, 0.0193, 0.0132, 0.0205, 0.0118, 0.0129, 0.0118], + device='cuda:3'), out_proj_covar=tensor([1.2445e-04, 1.4557e-04, 1.2955e-04, 9.6870e-05, 1.4336e-04, 8.6310e-05, + 9.2721e-05, 8.8126e-05], device='cuda:3') +2023-03-31 19:38:27,463 INFO [train.py:903] (3/4) Epoch 1, batch 2250, loss[loss=0.4659, simple_loss=0.4789, pruned_loss=0.2265, over 19699.00 frames. ], tot_loss[loss=0.518, simple_loss=0.4945, pruned_loss=0.2715, over 3821303.84 frames. ], batch size: 59, lr: 4.77e-02, grad_scale: 16.0 +2023-03-31 19:39:24,144 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2299.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:39:25,913 INFO [train.py:903] (3/4) Epoch 1, batch 2300, loss[loss=0.5031, simple_loss=0.495, pruned_loss=0.2556, over 19301.00 frames. ], tot_loss[loss=0.5142, simple_loss=0.4923, pruned_loss=0.2687, over 3805063.08 frames. ], batch size: 66, lr: 4.77e-02, grad_scale: 8.0 +2023-03-31 19:39:31,322 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.883e+02 9.458e+02 1.205e+03 1.557e+03 3.326e+03, threshold=2.410e+03, percent-clipped=10.0 +2023-03-31 19:39:39,152 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-03-31 19:39:41,702 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2315.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:39:54,010 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2324.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:40:04,986 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2334.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:40:16,272 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-03-31 19:40:17,288 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2344.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:40:20,648 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2347.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:40:24,837 INFO [train.py:903] (3/4) Epoch 1, batch 2350, loss[loss=0.4928, simple_loss=0.4908, pruned_loss=0.2474, over 19735.00 frames. ], tot_loss[loss=0.5107, simple_loss=0.4902, pruned_loss=0.266, over 3809235.26 frames. ], batch size: 63, lr: 4.76e-02, grad_scale: 8.0 +2023-03-31 19:40:33,273 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2358.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:40:48,340 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2369.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:41:07,209 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-03-31 19:41:23,353 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-03-31 19:41:26,487 INFO [train.py:903] (3/4) Epoch 1, batch 2400, loss[loss=0.4212, simple_loss=0.4217, pruned_loss=0.2104, over 19731.00 frames. ], tot_loss[loss=0.5035, simple_loss=0.486, pruned_loss=0.2609, over 3820760.20 frames. ], batch size: 45, lr: 4.75e-02, grad_scale: 8.0 +2023-03-31 19:41:33,163 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.460e+02 9.458e+02 1.226e+03 1.613e+03 2.603e+03, threshold=2.451e+03, percent-clipped=4.0 +2023-03-31 19:42:26,145 INFO [train.py:903] (3/4) Epoch 1, batch 2450, loss[loss=0.4952, simple_loss=0.4916, pruned_loss=0.2494, over 19513.00 frames. ], tot_loss[loss=0.4993, simple_loss=0.4838, pruned_loss=0.2577, over 3818890.36 frames. ], batch size: 56, lr: 4.74e-02, grad_scale: 8.0 +2023-03-31 19:43:24,714 INFO [train.py:903] (3/4) Epoch 1, batch 2500, loss[loss=0.4905, simple_loss=0.4878, pruned_loss=0.2466, over 18859.00 frames. ], tot_loss[loss=0.4937, simple_loss=0.4805, pruned_loss=0.2537, over 3822051.45 frames. ], batch size: 74, lr: 4.73e-02, grad_scale: 8.0 +2023-03-31 19:43:30,998 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.049e+02 1.082e+03 1.390e+03 1.742e+03 4.873e+03, threshold=2.779e+03, percent-clipped=5.0 +2023-03-31 19:43:37,153 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-31 19:44:22,072 INFO [train.py:903] (3/4) Epoch 1, batch 2550, loss[loss=0.436, simple_loss=0.4401, pruned_loss=0.2159, over 19864.00 frames. ], tot_loss[loss=0.4931, simple_loss=0.48, pruned_loss=0.2532, over 3810120.13 frames. ], batch size: 52, lr: 4.72e-02, grad_scale: 8.0 +2023-03-31 19:44:47,236 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2571.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:44:51,869 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.62 vs. limit=5.0 +2023-03-31 19:45:09,096 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2590.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:45:14,145 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-03-31 19:45:15,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2596.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:45:21,672 INFO [train.py:903] (3/4) Epoch 1, batch 2600, loss[loss=0.466, simple_loss=0.4737, pruned_loss=0.2291, over 19739.00 frames. ], tot_loss[loss=0.4901, simple_loss=0.4786, pruned_loss=0.2509, over 3800060.88 frames. ], batch size: 63, lr: 4.71e-02, grad_scale: 8.0 +2023-03-31 19:45:25,701 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2603.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:45:28,247 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.353e+02 9.154e+02 1.259e+03 1.710e+03 2.682e+03, threshold=2.519e+03, percent-clipped=0.0 +2023-03-31 19:45:39,484 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2615.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:45:44,378 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.40 vs. limit=5.0 +2023-03-31 19:45:46,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2522, 1.0848, 1.2938, 1.4340, 1.9845, 0.9427, 2.0066, 2.1586], + device='cuda:3'), covar=tensor([0.0419, 0.2541, 0.3026, 0.2109, 0.0642, 0.3217, 0.0947, 0.0692], + device='cuda:3'), in_proj_covar=tensor([0.0115, 0.0198, 0.0216, 0.0205, 0.0148, 0.0288, 0.0183, 0.0155], + device='cuda:3'), out_proj_covar=tensor([8.3711e-05, 1.4422e-04, 1.5926e-04, 1.6019e-04, 1.0830e-04, 1.9321e-04, + 1.4648e-04, 1.1630e-04], device='cuda:3') +2023-03-31 19:45:55,231 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2628.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:46:00,887 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.62 vs. limit=5.0 +2023-03-31 19:46:22,935 INFO [train.py:903] (3/4) Epoch 1, batch 2650, loss[loss=0.5801, simple_loss=0.5287, pruned_loss=0.3157, over 13589.00 frames. ], tot_loss[loss=0.4846, simple_loss=0.4756, pruned_loss=0.2469, over 3808439.78 frames. ], batch size: 136, lr: 4.70e-02, grad_scale: 8.0 +2023-03-31 19:46:28,025 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-31 19:46:39,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-03-31 19:47:23,183 INFO [train.py:903] (3/4) Epoch 1, batch 2700, loss[loss=0.4772, simple_loss=0.4804, pruned_loss=0.237, over 18710.00 frames. ], tot_loss[loss=0.4811, simple_loss=0.4736, pruned_loss=0.2443, over 3812619.74 frames. ], batch size: 74, lr: 4.69e-02, grad_scale: 8.0 +2023-03-31 19:47:24,589 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=2702.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:47:25,667 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2703.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:47:29,727 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.007e+02 8.490e+02 1.133e+03 1.436e+03 3.154e+03, threshold=2.267e+03, percent-clipped=3.0 +2023-03-31 19:48:24,813 INFO [train.py:903] (3/4) Epoch 1, batch 2750, loss[loss=0.5299, simple_loss=0.5019, pruned_loss=0.279, over 13256.00 frames. ], tot_loss[loss=0.4769, simple_loss=0.471, pruned_loss=0.2414, over 3813471.07 frames. ], batch size: 136, lr: 4.68e-02, grad_scale: 8.0 +2023-03-31 19:49:25,709 INFO [train.py:903] (3/4) Epoch 1, batch 2800, loss[loss=0.4942, simple_loss=0.499, pruned_loss=0.2447, over 19613.00 frames. ], tot_loss[loss=0.473, simple_loss=0.4683, pruned_loss=0.2389, over 3806723.15 frames. ], batch size: 57, lr: 4.67e-02, grad_scale: 8.0 +2023-03-31 19:49:31,047 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.211e+02 1.002e+03 1.265e+03 1.511e+03 4.462e+03, threshold=2.529e+03, percent-clipped=7.0 +2023-03-31 19:49:45,708 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2817.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:50:15,447 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2842.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:50:26,202 INFO [train.py:903] (3/4) Epoch 1, batch 2850, loss[loss=0.4004, simple_loss=0.4269, pruned_loss=0.1869, over 19664.00 frames. ], tot_loss[loss=0.4711, simple_loss=0.4672, pruned_loss=0.2375, over 3810657.18 frames. ], batch size: 55, lr: 4.66e-02, grad_scale: 8.0 +2023-03-31 19:51:16,440 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.61 vs. limit=5.0 +2023-03-31 19:51:22,285 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-03-31 19:51:25,262 INFO [train.py:903] (3/4) Epoch 1, batch 2900, loss[loss=0.4738, simple_loss=0.445, pruned_loss=0.2512, over 19770.00 frames. ], tot_loss[loss=0.4688, simple_loss=0.4658, pruned_loss=0.2359, over 3799133.40 frames. ], batch size: 47, lr: 4.65e-02, grad_scale: 8.0 +2023-03-31 19:51:30,476 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.237e+02 1.045e+03 1.349e+03 1.754e+03 3.463e+03, threshold=2.699e+03, percent-clipped=4.0 +2023-03-31 19:52:25,021 INFO [train.py:903] (3/4) Epoch 1, batch 2950, loss[loss=0.5112, simple_loss=0.4994, pruned_loss=0.2614, over 18207.00 frames. ], tot_loss[loss=0.4671, simple_loss=0.465, pruned_loss=0.2346, over 3797566.85 frames. ], batch size: 83, lr: 4.64e-02, grad_scale: 8.0 +2023-03-31 19:52:30,839 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2955.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:53:02,076 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0237, 1.6132, 2.4952, 1.5123, 2.4328, 4.5148, 3.3487, 2.5994], + device='cuda:3'), covar=tensor([0.2997, 0.1720, 0.2164, 0.2782, 0.1448, 0.0171, 0.1126, 0.1610], + device='cuda:3'), in_proj_covar=tensor([0.0160, 0.0126, 0.0146, 0.0173, 0.0159, 0.0078, 0.0143, 0.0158], + device='cuda:3'), out_proj_covar=tensor([1.0920e-04, 8.5020e-05, 1.0301e-04, 1.2028e-04, 1.0772e-04, 5.1987e-05, + 9.6555e-05, 1.0380e-04], device='cuda:3') +2023-03-31 19:53:26,136 INFO [train.py:903] (3/4) Epoch 1, batch 3000, loss[loss=0.4323, simple_loss=0.443, pruned_loss=0.2108, over 19713.00 frames. ], tot_loss[loss=0.4628, simple_loss=0.4625, pruned_loss=0.2316, over 3801473.12 frames. ], batch size: 51, lr: 4.63e-02, grad_scale: 8.0 +2023-03-31 19:53:26,137 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 19:53:38,705 INFO [train.py:937] (3/4) Epoch 1, validation: loss=0.3995, simple_loss=0.4801, pruned_loss=0.1594, over 944034.00 frames. +2023-03-31 19:53:38,705 INFO [train.py:938] (3/4) Maximum memory allocated so far is 15470MB +2023-03-31 19:53:43,180 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-03-31 19:53:45,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.188e+02 9.060e+02 1.151e+03 1.550e+03 2.691e+03, threshold=2.303e+03, percent-clipped=0.0 +2023-03-31 19:53:56,937 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8757, 2.0647, 2.0163, 3.0921, 4.3899, 1.3692, 2.4858, 4.1978], + device='cuda:3'), covar=tensor([0.0254, 0.2358, 0.2864, 0.1656, 0.0291, 0.3593, 0.1022, 0.0323], + device='cuda:3'), in_proj_covar=tensor([0.0138, 0.0218, 0.0229, 0.0231, 0.0164, 0.0309, 0.0202, 0.0169], + device='cuda:3'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 19:54:23,283 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3037.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:54:35,539 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3047.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:54:40,057 INFO [train.py:903] (3/4) Epoch 1, batch 3050, loss[loss=0.4171, simple_loss=0.4193, pruned_loss=0.2075, over 19728.00 frames. ], tot_loss[loss=0.459, simple_loss=0.4603, pruned_loss=0.2288, over 3814212.25 frames. ], batch size: 46, lr: 4.62e-02, grad_scale: 8.0 +2023-03-31 19:54:46,090 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-03-31 19:55:07,353 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3073.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 19:55:36,294 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3098.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 19:55:39,210 INFO [train.py:903] (3/4) Epoch 1, batch 3100, loss[loss=0.497, simple_loss=0.4809, pruned_loss=0.2565, over 19658.00 frames. ], tot_loss[loss=0.4613, simple_loss=0.4615, pruned_loss=0.2305, over 3820186.25 frames. ], batch size: 53, lr: 4.61e-02, grad_scale: 8.0 +2023-03-31 19:55:45,836 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.197e+02 1.021e+03 1.362e+03 1.815e+03 5.785e+03, threshold=2.723e+03, percent-clipped=14.0 +2023-03-31 19:56:41,857 INFO [train.py:903] (3/4) Epoch 1, batch 3150, loss[loss=0.4683, simple_loss=0.4724, pruned_loss=0.2321, over 18723.00 frames. ], tot_loss[loss=0.4583, simple_loss=0.4596, pruned_loss=0.2285, over 3816716.02 frames. ], batch size: 74, lr: 4.60e-02, grad_scale: 8.0 +2023-03-31 19:56:54,540 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3162.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:57:03,670 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4357, 0.9079, 1.0135, 1.1918, 1.3666, 1.5957, 1.3953, 1.4780], + device='cuda:3'), covar=tensor([0.1086, 0.2613, 0.2398, 0.1675, 0.2570, 0.1006, 0.1502, 0.1212], + device='cuda:3'), in_proj_covar=tensor([0.0119, 0.0160, 0.0174, 0.0145, 0.0200, 0.0133, 0.0143, 0.0133], + device='cuda:3'), out_proj_covar=tensor([8.8973e-05, 1.1752e-04, 1.2172e-04, 1.0654e-04, 1.4492e-04, 9.5291e-05, + 1.0151e-04, 9.6116e-05], device='cuda:3') +2023-03-31 19:57:08,973 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-03-31 19:57:24,065 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3186.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:57:31,522 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7914, 1.2726, 1.2453, 1.6260, 1.8342, 1.6934, 1.8321, 1.5687], + device='cuda:3'), covar=tensor([0.1494, 0.2530, 0.2651, 0.2335, 0.2963, 0.2584, 0.3346, 0.1732], + device='cuda:3'), in_proj_covar=tensor([0.0200, 0.0254, 0.0252, 0.0269, 0.0353, 0.0249, 0.0323, 0.0219], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 19:57:40,816 INFO [train.py:903] (3/4) Epoch 1, batch 3200, loss[loss=0.4037, simple_loss=0.4277, pruned_loss=0.1898, over 19765.00 frames. ], tot_loss[loss=0.4564, simple_loss=0.4588, pruned_loss=0.227, over 3807910.90 frames. ], batch size: 56, lr: 4.59e-02, grad_scale: 8.0 +2023-03-31 19:57:46,449 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.226e+02 9.158e+02 1.127e+03 1.418e+03 2.574e+03, threshold=2.253e+03, percent-clipped=0.0 +2023-03-31 19:58:32,927 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.73 vs. limit=2.0 +2023-03-31 19:58:41,766 INFO [train.py:903] (3/4) Epoch 1, batch 3250, loss[loss=0.5085, simple_loss=0.5016, pruned_loss=0.2577, over 19704.00 frames. ], tot_loss[loss=0.4554, simple_loss=0.4584, pruned_loss=0.2263, over 3798699.47 frames. ], batch size: 63, lr: 4.58e-02, grad_scale: 8.0 +2023-03-31 19:59:40,390 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3299.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:59:42,437 INFO [train.py:903] (3/4) Epoch 1, batch 3300, loss[loss=0.3753, simple_loss=0.4036, pruned_loss=0.1735, over 19499.00 frames. ], tot_loss[loss=0.451, simple_loss=0.4555, pruned_loss=0.2233, over 3799373.19 frames. ], batch size: 49, lr: 4.57e-02, grad_scale: 8.0 +2023-03-31 19:59:43,737 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3301.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 19:59:48,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.005e+02 9.991e+02 1.183e+03 1.562e+03 4.237e+03, threshold=2.366e+03, percent-clipped=7.0 +2023-03-31 19:59:48,809 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-03-31 19:59:49,125 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3306.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:00:29,865 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-31 20:00:43,748 INFO [train.py:903] (3/4) Epoch 1, batch 3350, loss[loss=0.4988, simple_loss=0.4872, pruned_loss=0.2552, over 19353.00 frames. ], tot_loss[loss=0.4509, simple_loss=0.4561, pruned_loss=0.2228, over 3801688.69 frames. ], batch size: 66, lr: 4.56e-02, grad_scale: 8.0 +2023-03-31 20:01:16,587 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.3750, 4.8479, 3.1252, 4.5286, 1.6523, 5.3666, 4.7548, 5.2739], + device='cuda:3'), covar=tensor([0.0451, 0.0950, 0.2123, 0.0486, 0.3187, 0.0470, 0.0529, 0.0441], + device='cuda:3'), in_proj_covar=tensor([0.0197, 0.0220, 0.0242, 0.0182, 0.0255, 0.0169, 0.0146, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:3') +2023-03-31 20:01:22,261 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3381.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:01:46,120 INFO [train.py:903] (3/4) Epoch 1, batch 3400, loss[loss=0.466, simple_loss=0.449, pruned_loss=0.2415, over 19581.00 frames. ], tot_loss[loss=0.4488, simple_loss=0.4545, pruned_loss=0.2216, over 3795050.85 frames. ], batch size: 52, lr: 4.55e-02, grad_scale: 8.0 +2023-03-31 20:01:52,574 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-31 20:01:52,898 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.525e+02 9.967e+02 1.253e+03 1.611e+03 4.007e+03, threshold=2.507e+03, percent-clipped=3.0 +2023-03-31 20:02:02,173 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3414.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:02:07,709 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3418.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:02:38,951 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3443.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:02:48,483 INFO [train.py:903] (3/4) Epoch 1, batch 3450, loss[loss=0.404, simple_loss=0.4256, pruned_loss=0.1912, over 19728.00 frames. ], tot_loss[loss=0.4445, simple_loss=0.4511, pruned_loss=0.219, over 3809038.87 frames. ], batch size: 47, lr: 4.54e-02, grad_scale: 8.0 +2023-03-31 20:02:50,770 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-03-31 20:03:43,984 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3496.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:03:50,182 INFO [train.py:903] (3/4) Epoch 1, batch 3500, loss[loss=0.4473, simple_loss=0.4699, pruned_loss=0.2124, over 19677.00 frames. ], tot_loss[loss=0.4433, simple_loss=0.4503, pruned_loss=0.2181, over 3810675.17 frames. ], batch size: 59, lr: 4.53e-02, grad_scale: 8.0 +2023-03-31 20:03:56,683 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.496e+02 9.703e+02 1.213e+03 1.703e+03 9.610e+03, threshold=2.427e+03, percent-clipped=9.0 +2023-03-31 20:04:16,296 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9988, 2.1675, 2.4133, 1.9159, 2.0936, 1.3075, 1.4820, 2.0989], + device='cuda:3'), covar=tensor([0.1272, 0.0556, 0.0425, 0.1002, 0.0930, 0.1181, 0.1606, 0.0845], + device='cuda:3'), in_proj_covar=tensor([0.0226, 0.0141, 0.0143, 0.0177, 0.0136, 0.0190, 0.0211, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:04:24,132 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2369, 2.9464, 3.0982, 2.5469, 2.5409, 0.9832, 1.0707, 2.0154], + device='cuda:3'), covar=tensor([0.2053, 0.0742, 0.0480, 0.1163, 0.1497, 0.2000, 0.3079, 0.1634], + device='cuda:3'), in_proj_covar=tensor([0.0225, 0.0140, 0.0143, 0.0177, 0.0135, 0.0191, 0.0212, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:04:52,066 INFO [train.py:903] (3/4) Epoch 1, batch 3550, loss[loss=0.4017, simple_loss=0.4089, pruned_loss=0.1972, over 19753.00 frames. ], tot_loss[loss=0.4446, simple_loss=0.4515, pruned_loss=0.2189, over 3805643.68 frames. ], batch size: 45, lr: 4.51e-02, grad_scale: 8.0 +2023-03-31 20:05:00,231 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3557.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:05:07,871 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3564.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:05:31,015 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3582.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:05:33,604 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.93 vs. limit=5.0 +2023-03-31 20:05:47,590 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-31 20:05:53,943 INFO [train.py:903] (3/4) Epoch 1, batch 3600, loss[loss=0.4518, simple_loss=0.4622, pruned_loss=0.2207, over 19531.00 frames. ], tot_loss[loss=0.4595, simple_loss=0.4606, pruned_loss=0.2292, over 3807128.78 frames. ], batch size: 54, lr: 4.50e-02, grad_scale: 8.0 +2023-03-31 20:06:00,965 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.467e+02 9.459e+02 1.417e+03 1.964e+03 2.103e+04, threshold=2.834e+03, percent-clipped=17.0 +2023-03-31 20:06:55,033 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3650.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:06:55,794 INFO [train.py:903] (3/4) Epoch 1, batch 3650, loss[loss=0.5684, simple_loss=0.5259, pruned_loss=0.3055, over 19344.00 frames. ], tot_loss[loss=0.4575, simple_loss=0.46, pruned_loss=0.2275, over 3817666.48 frames. ], batch size: 66, lr: 4.49e-02, grad_scale: 8.0 +2023-03-31 20:07:18,765 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3670.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:07:49,511 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3694.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:07:50,860 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3695.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:07:56,857 INFO [train.py:903] (3/4) Epoch 1, batch 3700, loss[loss=0.4817, simple_loss=0.4819, pruned_loss=0.2408, over 19654.00 frames. ], tot_loss[loss=0.4602, simple_loss=0.4618, pruned_loss=0.2293, over 3832395.91 frames. ], batch size: 58, lr: 4.48e-02, grad_scale: 8.0 +2023-03-31 20:08:01,726 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1521, 3.7532, 4.6521, 4.3360, 2.4872, 4.1924, 3.9122, 4.1897], + device='cuda:3'), covar=tensor([0.0219, 0.0442, 0.0224, 0.0167, 0.2195, 0.0176, 0.0285, 0.0433], + device='cuda:3'), in_proj_covar=tensor([0.0136, 0.0179, 0.0217, 0.0148, 0.0312, 0.0124, 0.0170, 0.0199], + device='cuda:3'), out_proj_covar=tensor([9.5508e-05, 1.2382e-04, 1.4064e-04, 9.0595e-05, 1.7691e-04, 8.3968e-05, + 1.1319e-04, 1.2267e-04], device='cuda:3') +2023-03-31 20:08:05,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.255e+02 1.022e+03 1.666e+03 2.666e+03 1.441e+04, threshold=3.331e+03, percent-clipped=22.0 +2023-03-31 20:09:01,160 INFO [train.py:903] (3/4) Epoch 1, batch 3750, loss[loss=0.4791, simple_loss=0.4914, pruned_loss=0.2334, over 19400.00 frames. ], tot_loss[loss=0.4572, simple_loss=0.4604, pruned_loss=0.227, over 3822081.50 frames. ], batch size: 70, lr: 4.47e-02, grad_scale: 8.0 +2023-03-31 20:09:02,765 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3752.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:09:03,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2927, 1.4333, 1.5108, 1.2295, 1.3350, 0.7307, 0.5908, 1.6228], + device='cuda:3'), covar=tensor([0.1512, 0.0794, 0.0963, 0.1345, 0.1169, 0.1971, 0.2440, 0.1233], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0153, 0.0154, 0.0192, 0.0142, 0.0212, 0.0229, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0001, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:09:19,438 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3765.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:09:34,307 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3777.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:10:05,359 INFO [train.py:903] (3/4) Epoch 1, batch 3800, loss[loss=0.4516, simple_loss=0.452, pruned_loss=0.2256, over 19838.00 frames. ], tot_loss[loss=0.4528, simple_loss=0.4577, pruned_loss=0.2239, over 3823407.22 frames. ], batch size: 52, lr: 4.46e-02, grad_scale: 8.0 +2023-03-31 20:10:12,576 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.844e+02 1.035e+03 1.394e+03 1.973e+03 4.112e+03, threshold=2.788e+03, percent-clipped=1.0 +2023-03-31 20:10:13,349 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-31 20:10:41,878 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-03-31 20:11:08,752 INFO [train.py:903] (3/4) Epoch 1, batch 3850, loss[loss=0.5392, simple_loss=0.519, pruned_loss=0.2797, over 13564.00 frames. ], tot_loss[loss=0.4503, simple_loss=0.4562, pruned_loss=0.2222, over 3819650.53 frames. ], batch size: 136, lr: 4.45e-02, grad_scale: 8.0 +2023-03-31 20:12:13,118 INFO [train.py:903] (3/4) Epoch 1, batch 3900, loss[loss=0.4532, simple_loss=0.46, pruned_loss=0.2232, over 18794.00 frames. ], tot_loss[loss=0.4463, simple_loss=0.4536, pruned_loss=0.2195, over 3831982.65 frames. ], batch size: 74, lr: 4.44e-02, grad_scale: 8.0 +2023-03-31 20:12:17,965 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9181, 1.3678, 1.6597, 1.1008, 2.5556, 3.1863, 3.0343, 3.0571], + device='cuda:3'), covar=tensor([0.1710, 0.2551, 0.1965, 0.2992, 0.0665, 0.0196, 0.0251, 0.0317], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0258, 0.0260, 0.0287, 0.0197, 0.0130, 0.0151, 0.0128], + device='cuda:3'), out_proj_covar=tensor([2.0124e-04, 1.8870e-04, 1.9291e-04, 2.1106e-04, 1.6942e-04, 9.0144e-05, + 1.1416e-04, 9.9066e-05], device='cuda:3') +2023-03-31 20:12:22,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.917e+02 1.152e+03 1.441e+03 1.935e+03 3.736e+03, threshold=2.883e+03, percent-clipped=2.0 +2023-03-31 20:12:22,948 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-31 20:12:23,361 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=3908.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:13:08,746 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3944.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:13:11,155 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6985, 2.8633, 3.0551, 2.8540, 0.9436, 2.7169, 2.4524, 2.7103], + device='cuda:3'), covar=tensor([0.0358, 0.0484, 0.0519, 0.0386, 0.3057, 0.0294, 0.0558, 0.0891], + device='cuda:3'), in_proj_covar=tensor([0.0138, 0.0175, 0.0215, 0.0152, 0.0313, 0.0116, 0.0171, 0.0208], + device='cuda:3'), out_proj_covar=tensor([9.5347e-05, 1.2038e-04, 1.4004e-04, 9.3064e-05, 1.7663e-04, 8.0132e-05, + 1.1329e-04, 1.2646e-04], device='cuda:3') +2023-03-31 20:13:18,226 INFO [train.py:903] (3/4) Epoch 1, batch 3950, loss[loss=0.4506, simple_loss=0.478, pruned_loss=0.2116, over 18848.00 frames. ], tot_loss[loss=0.4501, simple_loss=0.4554, pruned_loss=0.2224, over 3818411.27 frames. ], batch size: 74, lr: 4.43e-02, grad_scale: 8.0 +2023-03-31 20:13:24,005 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-03-31 20:14:04,514 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-31 20:14:24,018 INFO [train.py:903] (3/4) Epoch 1, batch 4000, loss[loss=0.3673, simple_loss=0.3899, pruned_loss=0.1724, over 15078.00 frames. ], tot_loss[loss=0.4441, simple_loss=0.4514, pruned_loss=0.2184, over 3806991.61 frames. ], batch size: 33, lr: 4.42e-02, grad_scale: 8.0 +2023-03-31 20:14:26,763 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4328, 1.1325, 1.2442, 1.4639, 2.1764, 1.1293, 2.0243, 2.1709], + device='cuda:3'), covar=tensor([0.0431, 0.2297, 0.2640, 0.1670, 0.0474, 0.2312, 0.0781, 0.0561], + device='cuda:3'), in_proj_covar=tensor([0.0164, 0.0253, 0.0250, 0.0255, 0.0184, 0.0317, 0.0219, 0.0189], + device='cuda:3'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:14:30,932 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.942e+02 1.053e+03 1.358e+03 1.948e+03 9.883e+03, threshold=2.717e+03, percent-clipped=12.0 +2023-03-31 20:14:49,347 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4021.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:14:52,460 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4023.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:15:11,867 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=4038.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:15:12,800 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-03-31 20:15:22,664 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4046.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:15:28,264 INFO [train.py:903] (3/4) Epoch 1, batch 4050, loss[loss=0.4455, simple_loss=0.4575, pruned_loss=0.2167, over 19738.00 frames. ], tot_loss[loss=0.4434, simple_loss=0.451, pruned_loss=0.2179, over 3805387.31 frames. ], batch size: 63, lr: 4.41e-02, grad_scale: 8.0 +2023-03-31 20:16:32,907 INFO [train.py:903] (3/4) Epoch 1, batch 4100, loss[loss=0.3507, simple_loss=0.3786, pruned_loss=0.1614, over 19754.00 frames. ], tot_loss[loss=0.439, simple_loss=0.4477, pruned_loss=0.2151, over 3802247.97 frames. ], batch size: 46, lr: 4.40e-02, grad_scale: 8.0 +2023-03-31 20:16:41,801 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.649e+02 1.198e+03 1.458e+03 1.833e+03 3.490e+03, threshold=2.915e+03, percent-clipped=3.0 +2023-03-31 20:17:08,516 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-03-31 20:17:38,859 INFO [train.py:903] (3/4) Epoch 1, batch 4150, loss[loss=0.5073, simple_loss=0.4876, pruned_loss=0.2635, over 13562.00 frames. ], tot_loss[loss=0.4355, simple_loss=0.4456, pruned_loss=0.2127, over 3806231.33 frames. ], batch size: 136, lr: 4.39e-02, grad_scale: 8.0 +2023-03-31 20:17:41,668 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4153.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:17:49,797 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4159.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:18:44,275 INFO [train.py:903] (3/4) Epoch 1, batch 4200, loss[loss=0.4557, simple_loss=0.4723, pruned_loss=0.2196, over 19743.00 frames. ], tot_loss[loss=0.4336, simple_loss=0.4447, pruned_loss=0.2112, over 3815992.70 frames. ], batch size: 63, lr: 4.38e-02, grad_scale: 8.0 +2023-03-31 20:18:46,651 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-03-31 20:18:51,465 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.163e+02 8.919e+02 1.098e+03 1.489e+03 3.268e+03, threshold=2.196e+03, percent-clipped=3.0 +2023-03-31 20:19:01,556 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-03-31 20:19:47,736 INFO [train.py:903] (3/4) Epoch 1, batch 4250, loss[loss=0.5242, simple_loss=0.5089, pruned_loss=0.2698, over 19732.00 frames. ], tot_loss[loss=0.4358, simple_loss=0.4462, pruned_loss=0.2127, over 3816797.43 frames. ], batch size: 63, lr: 4.36e-02, grad_scale: 8.0 +2023-03-31 20:20:02,065 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-03-31 20:20:15,223 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-03-31 20:20:25,113 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4279.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:20:36,686 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=4288.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:20:52,970 INFO [train.py:903] (3/4) Epoch 1, batch 4300, loss[loss=0.3724, simple_loss=0.4014, pruned_loss=0.1717, over 19586.00 frames. ], tot_loss[loss=0.4341, simple_loss=0.4449, pruned_loss=0.2116, over 3826429.28 frames. ], batch size: 52, lr: 4.35e-02, grad_scale: 8.0 +2023-03-31 20:20:57,843 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4304.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:21:01,264 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4306.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:21:02,055 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.021e+02 1.171e+03 1.478e+03 2.100e+03 3.660e+03, threshold=2.957e+03, percent-clipped=20.0 +2023-03-31 20:21:44,198 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7471, 3.4161, 2.2393, 3.1828, 1.4185, 3.3734, 3.0578, 3.0665], + device='cuda:3'), covar=tensor([0.0638, 0.1200, 0.2062, 0.0613, 0.3129, 0.0794, 0.0572, 0.0847], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0231, 0.0256, 0.0205, 0.0273, 0.0207, 0.0155, 0.0171], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:3') +2023-03-31 20:21:46,494 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-03-31 20:21:59,415 INFO [train.py:903] (3/4) Epoch 1, batch 4350, loss[loss=0.4379, simple_loss=0.4468, pruned_loss=0.2145, over 19682.00 frames. ], tot_loss[loss=0.4316, simple_loss=0.4436, pruned_loss=0.2097, over 3834136.49 frames. ], batch size: 58, lr: 4.34e-02, grad_scale: 8.0 +2023-03-31 20:22:23,983 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.59 vs. limit=5.0 +2023-03-31 20:23:03,133 INFO [train.py:903] (3/4) Epoch 1, batch 4400, loss[loss=0.4151, simple_loss=0.4287, pruned_loss=0.2007, over 19621.00 frames. ], tot_loss[loss=0.4275, simple_loss=0.441, pruned_loss=0.2071, over 3846217.39 frames. ], batch size: 50, lr: 4.33e-02, grad_scale: 8.0 +2023-03-31 20:23:05,873 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4403.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:23:11,485 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.454e+02 9.391e+02 1.114e+03 1.514e+03 3.216e+03, threshold=2.228e+03, percent-clipped=1.0 +2023-03-31 20:23:14,566 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4409.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:23:29,276 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-03-31 20:23:39,404 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-03-31 20:23:46,934 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4434.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:23:50,503 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8610, 1.2293, 1.5997, 1.1610, 2.5087, 3.3376, 2.9669, 3.1636], + device='cuda:3'), covar=tensor([0.1767, 0.3135, 0.2442, 0.2797, 0.0704, 0.0160, 0.0258, 0.0207], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0268, 0.0278, 0.0297, 0.0203, 0.0126, 0.0153, 0.0117], + device='cuda:3'), out_proj_covar=tensor([2.1958e-04, 2.0524e-04, 2.1240e-04, 2.2676e-04, 1.8258e-04, 9.1234e-05, + 1.2155e-04, 9.7951e-05], device='cuda:3') +2023-03-31 20:24:04,974 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3956, 2.3672, 1.6425, 2.7726, 1.9345, 2.4326, 1.9434, 1.8206], + device='cuda:3'), covar=tensor([0.1161, 0.0850, 0.1044, 0.0923, 0.1477, 0.0778, 0.2042, 0.1161], + device='cuda:3'), in_proj_covar=tensor([0.0112, 0.0097, 0.0127, 0.0141, 0.0147, 0.0085, 0.0146, 0.0121], + device='cuda:3'), out_proj_covar=tensor([7.9044e-05, 6.7667e-05, 8.3553e-05, 9.2726e-05, 9.3568e-05, 5.2100e-05, + 1.0632e-04, 8.2738e-05], device='cuda:3') +2023-03-31 20:24:06,852 INFO [train.py:903] (3/4) Epoch 1, batch 4450, loss[loss=0.377, simple_loss=0.3959, pruned_loss=0.1791, over 19466.00 frames. ], tot_loss[loss=0.427, simple_loss=0.4404, pruned_loss=0.2068, over 3830011.39 frames. ], batch size: 49, lr: 4.32e-02, grad_scale: 8.0 +2023-03-31 20:25:00,175 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-31 20:25:09,796 INFO [train.py:903] (3/4) Epoch 1, batch 4500, loss[loss=0.4654, simple_loss=0.4729, pruned_loss=0.229, over 17358.00 frames. ], tot_loss[loss=0.4256, simple_loss=0.4395, pruned_loss=0.2058, over 3825692.41 frames. ], batch size: 101, lr: 4.31e-02, grad_scale: 8.0 +2023-03-31 20:25:12,511 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=4503.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:25:18,130 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.500e+02 1.049e+03 1.357e+03 1.620e+03 3.962e+03, threshold=2.713e+03, percent-clipped=8.0 +2023-03-31 20:25:31,639 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9759, 1.1511, 2.1793, 1.3018, 2.9090, 3.4130, 3.4486, 2.3913], + device='cuda:3'), covar=tensor([0.1840, 0.1843, 0.1516, 0.1772, 0.1055, 0.0534, 0.1029, 0.1451], + device='cuda:3'), in_proj_covar=tensor([0.0228, 0.0220, 0.0215, 0.0239, 0.0239, 0.0179, 0.0248, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:26:14,040 INFO [train.py:903] (3/4) Epoch 1, batch 4550, loss[loss=0.3664, simple_loss=0.3946, pruned_loss=0.1691, over 19752.00 frames. ], tot_loss[loss=0.4221, simple_loss=0.4373, pruned_loss=0.2034, over 3814276.75 frames. ], batch size: 46, lr: 4.30e-02, grad_scale: 8.0 +2023-03-31 20:26:24,212 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-03-31 20:26:47,301 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-03-31 20:27:16,034 INFO [train.py:903] (3/4) Epoch 1, batch 4600, loss[loss=0.4272, simple_loss=0.4442, pruned_loss=0.2051, over 18236.00 frames. ], tot_loss[loss=0.4218, simple_loss=0.4369, pruned_loss=0.2034, over 3821479.90 frames. ], batch size: 84, lr: 4.29e-02, grad_scale: 4.0 +2023-03-31 20:27:24,059 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.210e+02 9.691e+02 1.279e+03 1.723e+03 8.130e+03, threshold=2.557e+03, percent-clipped=7.0 +2023-03-31 20:27:36,635 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4618.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:28:17,136 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=4650.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:28:18,168 INFO [train.py:903] (3/4) Epoch 1, batch 4650, loss[loss=0.5297, simple_loss=0.5122, pruned_loss=0.2736, over 13040.00 frames. ], tot_loss[loss=0.4211, simple_loss=0.4366, pruned_loss=0.2028, over 3821080.93 frames. ], batch size: 136, lr: 4.28e-02, grad_scale: 4.0 +2023-03-31 20:28:27,866 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4659.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:28:34,636 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-03-31 20:28:44,787 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-03-31 20:28:58,636 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4684.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:29:18,957 INFO [train.py:903] (3/4) Epoch 1, batch 4700, loss[loss=0.4136, simple_loss=0.4476, pruned_loss=0.1898, over 19666.00 frames. ], tot_loss[loss=0.4194, simple_loss=0.4357, pruned_loss=0.2015, over 3825140.98 frames. ], batch size: 55, lr: 4.27e-02, grad_scale: 4.0 +2023-03-31 20:29:21,846 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9736, 1.2687, 1.1690, 1.7903, 1.5671, 2.0137, 1.8596, 1.9420], + device='cuda:3'), covar=tensor([0.0824, 0.1789, 0.1953, 0.1563, 0.2178, 0.1147, 0.1887, 0.0933], + device='cuda:3'), in_proj_covar=tensor([0.0235, 0.0302, 0.0301, 0.0328, 0.0395, 0.0280, 0.0355, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 20:29:28,007 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.618e+02 9.658e+02 1.202e+03 1.526e+03 2.859e+03, threshold=2.405e+03, percent-clipped=1.0 +2023-03-31 20:29:39,216 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-03-31 20:30:21,804 INFO [train.py:903] (3/4) Epoch 1, batch 4750, loss[loss=0.5142, simple_loss=0.494, pruned_loss=0.2672, over 13861.00 frames. ], tot_loss[loss=0.4171, simple_loss=0.4342, pruned_loss=0.2, over 3835963.16 frames. ], batch size: 135, lr: 4.26e-02, grad_scale: 4.0 +2023-03-31 20:30:39,345 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4765.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:31:18,393 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6875, 3.7415, 4.2463, 3.9432, 1.3755, 3.4972, 3.3933, 3.6619], + device='cuda:3'), covar=tensor([0.0273, 0.0358, 0.0284, 0.0211, 0.2799, 0.0257, 0.0336, 0.0691], + device='cuda:3'), in_proj_covar=tensor([0.0166, 0.0211, 0.0252, 0.0181, 0.0344, 0.0131, 0.0186, 0.0261], + device='cuda:3'), out_proj_covar=tensor([1.1039e-04, 1.4284e-04, 1.6781e-04, 1.1135e-04, 1.9166e-04, 8.4401e-05, + 1.2011e-04, 1.5564e-04], device='cuda:3') +2023-03-31 20:31:23,850 INFO [train.py:903] (3/4) Epoch 1, batch 4800, loss[loss=0.4281, simple_loss=0.445, pruned_loss=0.2056, over 19530.00 frames. ], tot_loss[loss=0.419, simple_loss=0.4355, pruned_loss=0.2012, over 3819027.62 frames. ], batch size: 54, lr: 4.25e-02, grad_scale: 8.0 +2023-03-31 20:31:32,988 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 7.504e+02 1.038e+03 1.224e+03 1.522e+03 3.175e+03, threshold=2.447e+03, percent-clipped=5.0 +2023-03-31 20:32:25,501 INFO [train.py:903] (3/4) Epoch 1, batch 4850, loss[loss=0.4042, simple_loss=0.4363, pruned_loss=0.186, over 19554.00 frames. ], tot_loss[loss=0.4199, simple_loss=0.4364, pruned_loss=0.2017, over 3817384.57 frames. ], batch size: 54, lr: 4.24e-02, grad_scale: 8.0 +2023-03-31 20:32:46,764 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-03-31 20:32:55,280 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4874.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:33:06,423 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-03-31 20:33:12,764 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-03-31 20:33:12,785 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-03-31 20:33:23,803 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-03-31 20:33:25,335 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4899.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:33:27,238 INFO [train.py:903] (3/4) Epoch 1, batch 4900, loss[loss=0.4417, simple_loss=0.4547, pruned_loss=0.2143, over 18819.00 frames. ], tot_loss[loss=0.4165, simple_loss=0.4341, pruned_loss=0.1994, over 3819940.15 frames. ], batch size: 74, lr: 4.23e-02, grad_scale: 8.0 +2023-03-31 20:33:37,034 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.319e+02 9.845e+02 1.167e+03 1.485e+03 2.856e+03, threshold=2.333e+03, percent-clipped=2.0 +2023-03-31 20:33:44,810 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-03-31 20:34:29,502 INFO [train.py:903] (3/4) Epoch 1, batch 4950, loss[loss=0.4733, simple_loss=0.4596, pruned_loss=0.2435, over 19028.00 frames. ], tot_loss[loss=0.4152, simple_loss=0.4335, pruned_loss=0.1985, over 3817645.17 frames. ], batch size: 42, lr: 4.21e-02, grad_scale: 8.0 +2023-03-31 20:34:38,656 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3520, 1.3868, 1.1565, 1.3785, 1.2458, 1.3829, 1.2002, 1.4433], + device='cuda:3'), covar=tensor([0.0791, 0.1361, 0.1399, 0.0965, 0.1608, 0.0728, 0.1367, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0205, 0.0307, 0.0262, 0.0224, 0.0293, 0.0219, 0.0247, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:34:40,548 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-03-31 20:35:05,873 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-03-31 20:35:31,813 INFO [train.py:903] (3/4) Epoch 1, batch 5000, loss[loss=0.3701, simple_loss=0.4075, pruned_loss=0.1663, over 19662.00 frames. ], tot_loss[loss=0.4154, simple_loss=0.4338, pruned_loss=0.1985, over 3819532.07 frames. ], batch size: 55, lr: 4.20e-02, grad_scale: 8.0 +2023-03-31 20:35:35,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-03-31 20:35:40,124 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.692e+02 8.720e+02 1.063e+03 1.451e+03 3.452e+03, threshold=2.125e+03, percent-clipped=4.0 +2023-03-31 20:35:46,914 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-03-31 20:35:56,112 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=5021.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:36:27,134 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=5046.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:36:33,266 INFO [train.py:903] (3/4) Epoch 1, batch 5050, loss[loss=0.4106, simple_loss=0.4416, pruned_loss=0.1898, over 19076.00 frames. ], tot_loss[loss=0.415, simple_loss=0.4337, pruned_loss=0.1982, over 3809980.30 frames. ], batch size: 69, lr: 4.19e-02, grad_scale: 8.0 +2023-03-31 20:37:02,501 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-03-31 20:37:34,520 INFO [train.py:903] (3/4) Epoch 1, batch 5100, loss[loss=0.4575, simple_loss=0.4594, pruned_loss=0.2278, over 18241.00 frames. ], tot_loss[loss=0.4145, simple_loss=0.4333, pruned_loss=0.1979, over 3818416.74 frames. ], batch size: 83, lr: 4.18e-02, grad_scale: 8.0 +2023-03-31 20:37:39,778 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-03-31 20:37:43,121 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.551e+02 1.065e+03 1.254e+03 1.490e+03 3.647e+03, threshold=2.509e+03, percent-clipped=6.0 +2023-03-31 20:37:43,168 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-03-31 20:37:47,677 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-03-31 20:38:34,795 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1210, 1.2025, 0.9722, 1.0665, 0.9196, 1.0692, 0.9494, 1.1616], + device='cuda:3'), covar=tensor([0.1363, 0.1731, 0.2029, 0.1428, 0.2065, 0.1125, 0.1662, 0.1138], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0313, 0.0262, 0.0226, 0.0291, 0.0214, 0.0245, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:38:36,972 INFO [train.py:903] (3/4) Epoch 1, batch 5150, loss[loss=0.4098, simple_loss=0.4415, pruned_loss=0.1891, over 19567.00 frames. ], tot_loss[loss=0.4145, simple_loss=0.433, pruned_loss=0.198, over 3812774.18 frames. ], batch size: 61, lr: 4.17e-02, grad_scale: 8.0 +2023-03-31 20:38:45,548 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5527, 4.1655, 2.1219, 3.6772, 1.1275, 4.0342, 3.5751, 3.7442], + device='cuda:3'), covar=tensor([0.0605, 0.1080, 0.2365, 0.0662, 0.3634, 0.0641, 0.0657, 0.0622], + device='cuda:3'), in_proj_covar=tensor([0.0238, 0.0238, 0.0269, 0.0224, 0.0289, 0.0226, 0.0167, 0.0178], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:3') +2023-03-31 20:38:46,435 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-03-31 20:39:20,328 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-03-31 20:39:37,969 INFO [train.py:903] (3/4) Epoch 1, batch 5200, loss[loss=0.4184, simple_loss=0.4451, pruned_loss=0.1958, over 19517.00 frames. ], tot_loss[loss=0.416, simple_loss=0.4336, pruned_loss=0.1992, over 3814801.00 frames. ], batch size: 56, lr: 4.16e-02, grad_scale: 8.0 +2023-03-31 20:39:45,881 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.640e+02 1.028e+03 1.252e+03 1.630e+03 4.880e+03, threshold=2.504e+03, percent-clipped=1.0 +2023-03-31 20:39:50,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-03-31 20:40:08,859 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0872, 1.3703, 2.0750, 1.5844, 2.9304, 4.6446, 4.3042, 5.0426], + device='cuda:3'), covar=tensor([0.1839, 0.3023, 0.2361, 0.2570, 0.0618, 0.0124, 0.0145, 0.0089], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0280, 0.0299, 0.0307, 0.0203, 0.0123, 0.0173, 0.0127], + device='cuda:3'), out_proj_covar=tensor([2.4248e-04, 2.2567e-04, 2.3750e-04, 2.4721e-04, 1.8835e-04, 9.8263e-05, + 1.4135e-04, 1.1334e-04], device='cuda:3') +2023-03-31 20:40:32,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-03-31 20:40:39,287 INFO [train.py:903] (3/4) Epoch 1, batch 5250, loss[loss=0.3402, simple_loss=0.3721, pruned_loss=0.1542, over 19473.00 frames. ], tot_loss[loss=0.4112, simple_loss=0.4309, pruned_loss=0.1958, over 3822795.69 frames. ], batch size: 49, lr: 4.15e-02, grad_scale: 8.0 +2023-03-31 20:40:42,832 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0143, 4.4536, 5.7239, 5.4948, 1.4754, 5.1421, 4.7287, 4.9699], + device='cuda:3'), covar=tensor([0.0172, 0.0354, 0.0273, 0.0151, 0.3180, 0.0133, 0.0262, 0.0650], + device='cuda:3'), in_proj_covar=tensor([0.0172, 0.0217, 0.0267, 0.0190, 0.0358, 0.0134, 0.0196, 0.0277], + device='cuda:3'), out_proj_covar=tensor([1.1452e-04, 1.4342e-04, 1.7712e-04, 1.1366e-04, 1.9788e-04, 8.7248e-05, + 1.2384e-04, 1.6197e-04], device='cuda:3') +2023-03-31 20:41:17,744 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7302, 1.2350, 1.1464, 1.6020, 1.4303, 1.7274, 1.8689, 1.6517], + device='cuda:3'), covar=tensor([0.0985, 0.1627, 0.1789, 0.1549, 0.2145, 0.1324, 0.1880, 0.1082], + device='cuda:3'), in_proj_covar=tensor([0.0239, 0.0296, 0.0297, 0.0322, 0.0387, 0.0272, 0.0358, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 20:41:39,595 INFO [train.py:903] (3/4) Epoch 1, batch 5300, loss[loss=0.3871, simple_loss=0.4185, pruned_loss=0.1778, over 19784.00 frames. ], tot_loss[loss=0.4115, simple_loss=0.4308, pruned_loss=0.196, over 3824272.50 frames. ], batch size: 56, lr: 4.14e-02, grad_scale: 8.0 +2023-03-31 20:41:48,686 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.962e+02 9.394e+02 1.191e+03 1.647e+03 4.206e+03, threshold=2.383e+03, percent-clipped=5.0 +2023-03-31 20:41:53,356 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-03-31 20:42:41,818 INFO [train.py:903] (3/4) Epoch 1, batch 5350, loss[loss=0.3897, simple_loss=0.4241, pruned_loss=0.1777, over 19693.00 frames. ], tot_loss[loss=0.4088, simple_loss=0.4288, pruned_loss=0.1944, over 3834310.18 frames. ], batch size: 63, lr: 4.13e-02, grad_scale: 8.0 +2023-03-31 20:43:13,710 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-03-31 20:43:43,593 INFO [train.py:903] (3/4) Epoch 1, batch 5400, loss[loss=0.3595, simple_loss=0.4037, pruned_loss=0.1576, over 19533.00 frames. ], tot_loss[loss=0.4051, simple_loss=0.4264, pruned_loss=0.1919, over 3841058.89 frames. ], batch size: 56, lr: 4.12e-02, grad_scale: 8.0 +2023-03-31 20:43:51,076 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.073e+02 9.364e+02 1.084e+03 1.611e+03 4.795e+03, threshold=2.168e+03, percent-clipped=7.0 +2023-03-31 20:44:28,519 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-31 20:44:44,653 INFO [train.py:903] (3/4) Epoch 1, batch 5450, loss[loss=0.4572, simple_loss=0.4651, pruned_loss=0.2247, over 19681.00 frames. ], tot_loss[loss=0.4057, simple_loss=0.4263, pruned_loss=0.1926, over 3838458.15 frames. ], batch size: 59, lr: 4.11e-02, grad_scale: 8.0 +2023-03-31 20:44:46,096 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1585, 2.8531, 1.7206, 2.6774, 0.9357, 2.8982, 2.5974, 2.7001], + device='cuda:3'), covar=tensor([0.0824, 0.1206, 0.2286, 0.0798, 0.3366, 0.0968, 0.0662, 0.0797], + device='cuda:3'), in_proj_covar=tensor([0.0236, 0.0240, 0.0279, 0.0225, 0.0292, 0.0231, 0.0170, 0.0180], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:3') +2023-03-31 20:45:46,525 INFO [train.py:903] (3/4) Epoch 1, batch 5500, loss[loss=0.524, simple_loss=0.5066, pruned_loss=0.2707, over 13203.00 frames. ], tot_loss[loss=0.407, simple_loss=0.4276, pruned_loss=0.1932, over 3835304.93 frames. ], batch size: 135, lr: 4.10e-02, grad_scale: 8.0 +2023-03-31 20:45:54,009 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.842e+02 9.504e+02 1.107e+03 1.412e+03 4.004e+03, threshold=2.214e+03, percent-clipped=7.0 +2023-03-31 20:46:08,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-03-31 20:46:30,122 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3518, 1.5315, 1.0127, 1.6081, 1.5813, 1.7257, 1.5350, 1.8238], + device='cuda:3'), covar=tensor([0.1081, 0.1911, 0.1850, 0.1152, 0.1738, 0.0735, 0.1209, 0.0748], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0318, 0.0263, 0.0227, 0.0292, 0.0223, 0.0243, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:46:46,622 INFO [train.py:903] (3/4) Epoch 1, batch 5550, loss[loss=0.4217, simple_loss=0.4463, pruned_loss=0.1986, over 19350.00 frames. ], tot_loss[loss=0.4055, simple_loss=0.4264, pruned_loss=0.1923, over 3839187.64 frames. ], batch size: 66, lr: 4.09e-02, grad_scale: 8.0 +2023-03-31 20:46:53,475 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-03-31 20:47:42,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-03-31 20:47:47,486 INFO [train.py:903] (3/4) Epoch 1, batch 5600, loss[loss=0.4692, simple_loss=0.466, pruned_loss=0.2362, over 17782.00 frames. ], tot_loss[loss=0.4027, simple_loss=0.4245, pruned_loss=0.1905, over 3852812.35 frames. ], batch size: 101, lr: 4.08e-02, grad_scale: 8.0 +2023-03-31 20:47:56,556 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.862e+02 1.009e+03 1.185e+03 1.400e+03 2.216e+03, threshold=2.370e+03, percent-clipped=2.0 +2023-03-31 20:48:48,828 INFO [train.py:903] (3/4) Epoch 1, batch 5650, loss[loss=0.3272, simple_loss=0.3568, pruned_loss=0.1488, over 19313.00 frames. ], tot_loss[loss=0.4045, simple_loss=0.4253, pruned_loss=0.1918, over 3848895.43 frames. ], batch size: 44, lr: 4.07e-02, grad_scale: 8.0 +2023-03-31 20:49:09,595 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1628, 3.9034, 2.7234, 3.3927, 1.8217, 3.5240, 3.3365, 3.4802], + device='cuda:3'), covar=tensor([0.0577, 0.0877, 0.1849, 0.0745, 0.2935, 0.0931, 0.0653, 0.0695], + device='cuda:3'), in_proj_covar=tensor([0.0237, 0.0231, 0.0277, 0.0224, 0.0291, 0.0225, 0.0170, 0.0183], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:3') +2023-03-31 20:49:09,666 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=5668.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 20:49:32,790 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-03-31 20:49:49,711 INFO [train.py:903] (3/4) Epoch 1, batch 5700, loss[loss=0.4137, simple_loss=0.4363, pruned_loss=0.1955, over 19649.00 frames. ], tot_loss[loss=0.4069, simple_loss=0.4264, pruned_loss=0.1937, over 3842715.70 frames. ], batch size: 60, lr: 4.06e-02, grad_scale: 8.0 +2023-03-31 20:49:57,493 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.193e+02 1.084e+03 1.385e+03 1.754e+03 4.325e+03, threshold=2.770e+03, percent-clipped=14.0 +2023-03-31 20:50:51,575 INFO [train.py:903] (3/4) Epoch 1, batch 5750, loss[loss=0.3466, simple_loss=0.3805, pruned_loss=0.1563, over 19755.00 frames. ], tot_loss[loss=0.4043, simple_loss=0.4249, pruned_loss=0.1918, over 3848497.14 frames. ], batch size: 46, lr: 4.05e-02, grad_scale: 8.0 +2023-03-31 20:50:51,594 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-03-31 20:50:59,658 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-03-31 20:51:05,170 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-03-31 20:51:18,557 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=5773.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:51:52,607 INFO [train.py:903] (3/4) Epoch 1, batch 5800, loss[loss=0.4147, simple_loss=0.445, pruned_loss=0.1922, over 19615.00 frames. ], tot_loss[loss=0.4046, simple_loss=0.4255, pruned_loss=0.1918, over 3838484.89 frames. ], batch size: 57, lr: 4.04e-02, grad_scale: 8.0 +2023-03-31 20:52:02,181 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.310e+02 8.969e+02 1.169e+03 1.352e+03 2.735e+03, threshold=2.337e+03, percent-clipped=0.0 +2023-03-31 20:52:02,647 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1417, 1.2154, 1.8097, 1.3638, 2.6279, 2.5828, 2.6389, 1.8300], + device='cuda:3'), covar=tensor([0.1599, 0.1732, 0.1233, 0.1616, 0.0707, 0.0633, 0.0864, 0.1326], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0269, 0.0254, 0.0284, 0.0296, 0.0231, 0.0317, 0.0290], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:52:25,497 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0051, 2.0684, 2.0558, 2.6907, 4.6283, 1.0363, 2.2130, 4.0742], + device='cuda:3'), covar=tensor([0.0222, 0.2330, 0.2628, 0.1549, 0.0287, 0.2753, 0.1287, 0.0366], + device='cuda:3'), in_proj_covar=tensor([0.0188, 0.0284, 0.0271, 0.0263, 0.0214, 0.0319, 0.0236, 0.0219], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 20:52:53,460 INFO [train.py:903] (3/4) Epoch 1, batch 5850, loss[loss=0.3526, simple_loss=0.3773, pruned_loss=0.164, over 19758.00 frames. ], tot_loss[loss=0.4059, simple_loss=0.4263, pruned_loss=0.1928, over 3830433.81 frames. ], batch size: 45, lr: 4.03e-02, grad_scale: 8.0 +2023-03-31 20:53:23,813 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=5876.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:53:55,088 INFO [train.py:903] (3/4) Epoch 1, batch 5900, loss[loss=0.4274, simple_loss=0.4528, pruned_loss=0.201, over 19679.00 frames. ], tot_loss[loss=0.4023, simple_loss=0.4236, pruned_loss=0.1905, over 3823595.74 frames. ], batch size: 58, lr: 4.02e-02, grad_scale: 8.0 +2023-03-31 20:53:58,588 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-03-31 20:54:03,122 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.633e+02 8.668e+02 1.127e+03 1.397e+03 3.736e+03, threshold=2.255e+03, percent-clipped=4.0 +2023-03-31 20:54:21,183 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-03-31 20:54:55,351 INFO [train.py:903] (3/4) Epoch 1, batch 5950, loss[loss=0.3684, simple_loss=0.3975, pruned_loss=0.1697, over 19497.00 frames. ], tot_loss[loss=0.3997, simple_loss=0.4217, pruned_loss=0.1889, over 3826874.76 frames. ], batch size: 49, lr: 4.01e-02, grad_scale: 8.0 +2023-03-31 20:55:02,437 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9278, 1.8206, 1.7824, 2.5209, 1.9895, 2.5499, 1.7596, 1.4414], + device='cuda:3'), covar=tensor([0.1060, 0.0847, 0.0593, 0.0639, 0.1058, 0.0429, 0.1595, 0.1217], + device='cuda:3'), in_proj_covar=tensor([0.0183, 0.0161, 0.0182, 0.0230, 0.0227, 0.0128, 0.0250, 0.0194], + device='cuda:3'), out_proj_covar=tensor([1.3641e-04, 1.1881e-04, 1.2153e-04, 1.5438e-04, 1.4831e-04, 8.7549e-05, + 1.7839e-04, 1.3483e-04], device='cuda:3') +2023-03-31 20:55:11,945 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-31 20:55:57,492 INFO [train.py:903] (3/4) Epoch 1, batch 6000, loss[loss=0.4021, simple_loss=0.4281, pruned_loss=0.1881, over 17173.00 frames. ], tot_loss[loss=0.4013, simple_loss=0.4226, pruned_loss=0.19, over 3814819.82 frames. ], batch size: 101, lr: 4.00e-02, grad_scale: 8.0 +2023-03-31 20:55:57,492 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 20:56:10,585 INFO [train.py:937] (3/4) Epoch 1, validation: loss=0.2784, simple_loss=0.3626, pruned_loss=0.09714, over 944034.00 frames. +2023-03-31 20:56:10,586 INFO [train.py:938] (3/4) Maximum memory allocated so far is 17850MB +2023-03-31 20:56:19,577 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.849e+02 9.012e+02 1.240e+03 1.620e+03 2.952e+03, threshold=2.480e+03, percent-clipped=5.0 +2023-03-31 20:56:19,903 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6008.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:56:24,078 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6012.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:57:10,738 INFO [train.py:903] (3/4) Epoch 1, batch 6050, loss[loss=0.511, simple_loss=0.4962, pruned_loss=0.2629, over 18733.00 frames. ], tot_loss[loss=0.4043, simple_loss=0.4246, pruned_loss=0.192, over 3824837.91 frames. ], batch size: 74, lr: 3.99e-02, grad_scale: 8.0 +2023-03-31 20:57:24,367 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6061.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:57:40,479 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-03-31 20:58:12,899 INFO [train.py:903] (3/4) Epoch 1, batch 6100, loss[loss=0.36, simple_loss=0.3876, pruned_loss=0.1662, over 19787.00 frames. ], tot_loss[loss=0.4013, simple_loss=0.4226, pruned_loss=0.19, over 3824367.90 frames. ], batch size: 48, lr: 3.98e-02, grad_scale: 8.0 +2023-03-31 20:58:20,963 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.479e+02 9.509e+02 1.169e+03 1.489e+03 2.977e+03, threshold=2.338e+03, percent-clipped=4.0 +2023-03-31 20:58:32,477 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6117.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:58:44,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6127.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 20:59:13,361 INFO [train.py:903] (3/4) Epoch 1, batch 6150, loss[loss=0.4424, simple_loss=0.4481, pruned_loss=0.2184, over 13825.00 frames. ], tot_loss[loss=0.4003, simple_loss=0.422, pruned_loss=0.1893, over 3811244.75 frames. ], batch size: 137, lr: 3.97e-02, grad_scale: 8.0 +2023-03-31 20:59:16,562 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6153.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 20:59:42,678 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-03-31 20:59:58,501 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6188.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:00:13,783 INFO [train.py:903] (3/4) Epoch 1, batch 6200, loss[loss=0.3569, simple_loss=0.404, pruned_loss=0.155, over 19652.00 frames. ], tot_loss[loss=0.4003, simple_loss=0.4218, pruned_loss=0.1894, over 3816237.82 frames. ], batch size: 55, lr: 3.96e-02, grad_scale: 8.0 +2023-03-31 21:00:22,728 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.993e+02 9.585e+02 1.181e+03 1.511e+03 2.920e+03, threshold=2.362e+03, percent-clipped=2.0 +2023-03-31 21:00:38,384 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6220.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:00:39,683 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6221.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:00:49,059 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6229.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:00:53,421 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6232.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:01:16,025 INFO [train.py:903] (3/4) Epoch 1, batch 6250, loss[loss=0.3691, simple_loss=0.4115, pruned_loss=0.1633, over 19658.00 frames. ], tot_loss[loss=0.3973, simple_loss=0.4201, pruned_loss=0.1872, over 3814354.95 frames. ], batch size: 55, lr: 3.95e-02, grad_scale: 8.0 +2023-03-31 21:01:19,936 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6651, 1.3844, 1.2105, 1.5786, 1.3748, 1.5030, 1.3620, 1.6914], + device='cuda:3'), covar=tensor([0.0927, 0.1920, 0.1555, 0.1211, 0.1641, 0.0963, 0.1505, 0.0830], + device='cuda:3'), in_proj_covar=tensor([0.0223, 0.0342, 0.0272, 0.0236, 0.0296, 0.0229, 0.0266, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:01:46,723 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-03-31 21:01:56,504 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6284.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:02:18,542 INFO [train.py:903] (3/4) Epoch 1, batch 6300, loss[loss=0.3345, simple_loss=0.3696, pruned_loss=0.1497, over 19400.00 frames. ], tot_loss[loss=0.3918, simple_loss=0.416, pruned_loss=0.1837, over 3821989.57 frames. ], batch size: 47, lr: 3.94e-02, grad_scale: 8.0 +2023-03-31 21:02:20,118 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8237, 1.1747, 1.1680, 1.8890, 1.7918, 1.8340, 1.9206, 1.8075], + device='cuda:3'), covar=tensor([0.1073, 0.2218, 0.1992, 0.1572, 0.2035, 0.1379, 0.1911, 0.1054], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0304, 0.0301, 0.0325, 0.0392, 0.0277, 0.0344, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 21:02:26,555 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.984e+02 8.812e+02 1.125e+03 1.363e+03 2.149e+03, threshold=2.249e+03, percent-clipped=0.0 +2023-03-31 21:02:58,304 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6335.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:03:00,493 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2929, 1.3589, 1.5386, 2.0131, 2.8829, 1.2359, 1.8315, 2.7713], + device='cuda:3'), covar=tensor([0.0308, 0.2731, 0.2736, 0.1729, 0.0418, 0.2502, 0.1167, 0.0483], + device='cuda:3'), in_proj_covar=tensor([0.0190, 0.0285, 0.0272, 0.0267, 0.0220, 0.0318, 0.0241, 0.0227], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:03:08,921 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3234, 1.3630, 1.0877, 1.4305, 1.2828, 1.4318, 1.3214, 1.4722], + device='cuda:3'), covar=tensor([0.1038, 0.1779, 0.1434, 0.1086, 0.1735, 0.0719, 0.1417, 0.0808], + device='cuda:3'), in_proj_covar=tensor([0.0233, 0.0350, 0.0267, 0.0237, 0.0308, 0.0234, 0.0269, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:03:17,632 INFO [train.py:903] (3/4) Epoch 1, batch 6350, loss[loss=0.3828, simple_loss=0.4246, pruned_loss=0.1704, over 19673.00 frames. ], tot_loss[loss=0.3941, simple_loss=0.4174, pruned_loss=0.1854, over 3825088.47 frames. ], batch size: 60, lr: 3.93e-02, grad_scale: 8.0 +2023-03-31 21:03:18,874 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6352.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:03:58,112 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6383.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:04:00,298 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0524, 1.1788, 0.8796, 1.1149, 0.9065, 1.1327, 0.8987, 1.0655], + device='cuda:3'), covar=tensor([0.1247, 0.1555, 0.1849, 0.1172, 0.1661, 0.0815, 0.1662, 0.0939], + device='cuda:3'), in_proj_covar=tensor([0.0235, 0.0351, 0.0270, 0.0239, 0.0307, 0.0236, 0.0273, 0.0212], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:04:07,474 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-03-31 21:04:08,228 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0020, 0.9767, 1.6674, 1.1625, 2.1930, 2.2063, 2.4273, 1.3984], + device='cuda:3'), covar=tensor([0.1524, 0.1923, 0.1134, 0.1545, 0.0713, 0.0756, 0.0762, 0.1359], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0299, 0.0282, 0.0301, 0.0318, 0.0262, 0.0353, 0.0314], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:04:18,895 INFO [train.py:903] (3/4) Epoch 1, batch 6400, loss[loss=0.4386, simple_loss=0.455, pruned_loss=0.2111, over 19272.00 frames. ], tot_loss[loss=0.3944, simple_loss=0.4186, pruned_loss=0.1851, over 3818808.15 frames. ], batch size: 66, lr: 3.92e-02, grad_scale: 8.0 +2023-03-31 21:04:24,491 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6405.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:04:27,870 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.880e+02 9.359e+02 1.206e+03 1.547e+03 5.333e+03, threshold=2.412e+03, percent-clipped=7.0 +2023-03-31 21:04:28,298 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6408.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 21:05:19,064 INFO [train.py:903] (3/4) Epoch 1, batch 6450, loss[loss=0.4477, simple_loss=0.4663, pruned_loss=0.2145, over 19757.00 frames. ], tot_loss[loss=0.3925, simple_loss=0.4173, pruned_loss=0.1838, over 3819062.82 frames. ], batch size: 63, lr: 3.91e-02, grad_scale: 8.0 +2023-03-31 21:05:23,287 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-03-31 21:05:39,909 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6467.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:05:55,637 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6480.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:01,942 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6485.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:05,061 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-03-31 21:06:05,473 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6488.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:15,204 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6497.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:21,207 INFO [train.py:903] (3/4) Epoch 1, batch 6500, loss[loss=0.4721, simple_loss=0.4585, pruned_loss=0.2429, over 19866.00 frames. ], tot_loss[loss=0.392, simple_loss=0.4169, pruned_loss=0.1835, over 3820790.86 frames. ], batch size: 52, lr: 3.90e-02, grad_scale: 8.0 +2023-03-31 21:06:25,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-03-31 21:06:28,780 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.341e+02 9.704e+02 1.201e+03 1.443e+03 2.205e+03, threshold=2.402e+03, percent-clipped=0.0 +2023-03-31 21:06:35,156 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6513.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:43,913 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6520.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:06:58,411 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6532.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:07:21,802 INFO [train.py:903] (3/4) Epoch 1, batch 6550, loss[loss=0.4512, simple_loss=0.4533, pruned_loss=0.2245, over 13473.00 frames. ], tot_loss[loss=0.3946, simple_loss=0.4188, pruned_loss=0.1852, over 3813824.29 frames. ], batch size: 135, lr: 3.89e-02, grad_scale: 8.0 +2023-03-31 21:07:39,353 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6565.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:07:48,423 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6573.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:08:10,379 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6591.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:08:22,326 INFO [train.py:903] (3/4) Epoch 1, batch 6600, loss[loss=0.4163, simple_loss=0.4509, pruned_loss=0.1909, over 19474.00 frames. ], tot_loss[loss=0.3946, simple_loss=0.4195, pruned_loss=0.1849, over 3822363.20 frames. ], batch size: 64, lr: 3.89e-02, grad_scale: 16.0 +2023-03-31 21:08:31,045 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.089e+02 8.736e+02 1.082e+03 1.231e+03 3.386e+03, threshold=2.164e+03, percent-clipped=2.0 +2023-03-31 21:08:36,148 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6612.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:08:39,570 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5294, 1.5454, 1.5536, 2.1666, 3.1914, 1.0517, 1.9741, 3.2127], + device='cuda:3'), covar=tensor([0.0314, 0.2714, 0.2943, 0.1474, 0.0437, 0.2710, 0.1217, 0.0389], + device='cuda:3'), in_proj_covar=tensor([0.0191, 0.0288, 0.0277, 0.0270, 0.0222, 0.0317, 0.0243, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:08:40,744 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6616.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:08:56,079 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6628.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:09:19,949 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6647.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:09:20,015 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6647.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:09:24,251 INFO [train.py:903] (3/4) Epoch 1, batch 6650, loss[loss=0.3407, simple_loss=0.3706, pruned_loss=0.1554, over 19753.00 frames. ], tot_loss[loss=0.3909, simple_loss=0.417, pruned_loss=0.1824, over 3825994.16 frames. ], batch size: 45, lr: 3.88e-02, grad_scale: 4.0 +2023-03-31 21:09:59,912 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6680.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:10:09,254 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6688.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:10:23,707 INFO [train.py:903] (3/4) Epoch 1, batch 6700, loss[loss=0.4066, simple_loss=0.4352, pruned_loss=0.189, over 19676.00 frames. ], tot_loss[loss=0.3909, simple_loss=0.4167, pruned_loss=0.1825, over 3842432.06 frames. ], batch size: 58, lr: 3.87e-02, grad_scale: 4.0 +2023-03-31 21:10:35,447 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.729e+02 8.354e+02 1.101e+03 1.693e+03 1.016e+04, threshold=2.202e+03, percent-clipped=16.0 +2023-03-31 21:10:51,620 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6723.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:11:15,021 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6743.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:11:20,797 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6748.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:11:23,805 INFO [train.py:903] (3/4) Epoch 1, batch 6750, loss[loss=0.4068, simple_loss=0.4389, pruned_loss=0.1873, over 19785.00 frames. ], tot_loss[loss=0.389, simple_loss=0.4152, pruned_loss=0.1814, over 3839690.94 frames. ], batch size: 56, lr: 3.86e-02, grad_scale: 4.0 +2023-03-31 21:11:51,653 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6776.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:12:18,954 INFO [train.py:903] (3/4) Epoch 1, batch 6800, loss[loss=0.4219, simple_loss=0.4485, pruned_loss=0.1976, over 19789.00 frames. ], tot_loss[loss=0.3877, simple_loss=0.4139, pruned_loss=0.1807, over 3837967.15 frames. ], batch size: 56, lr: 3.85e-02, grad_scale: 8.0 +2023-03-31 21:12:19,295 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6801.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:12:28,859 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.487e+02 9.125e+02 1.072e+03 1.412e+03 3.162e+03, threshold=2.143e+03, percent-clipped=4.0 +2023-03-31 21:12:43,734 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6824.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:13:03,112 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-03-31 21:13:04,135 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-03-31 21:13:06,750 INFO [train.py:903] (3/4) Epoch 2, batch 0, loss[loss=0.4173, simple_loss=0.4129, pruned_loss=0.2109, over 19394.00 frames. ], tot_loss[loss=0.4173, simple_loss=0.4129, pruned_loss=0.2109, over 19394.00 frames. ], batch size: 47, lr: 3.77e-02, grad_scale: 8.0 +2023-03-31 21:13:06,751 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 21:13:16,045 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1283, 1.4364, 1.2343, 1.1810, 1.2767, 0.7777, 0.7814, 1.3436], + device='cuda:3'), covar=tensor([0.1069, 0.0516, 0.1056, 0.0815, 0.0828, 0.1405, 0.1369, 0.0629], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0165, 0.0240, 0.0248, 0.0173, 0.0275, 0.0263, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:13:18,413 INFO [train.py:937] (3/4) Epoch 2, validation: loss=0.2802, simple_loss=0.3637, pruned_loss=0.09835, over 944034.00 frames. +2023-03-31 21:13:18,415 INFO [train.py:938] (3/4) Maximum memory allocated so far is 17850MB +2023-03-31 21:13:18,572 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6829.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:13:28,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-03-31 21:14:06,591 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6868.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:14:20,893 INFO [train.py:903] (3/4) Epoch 2, batch 50, loss[loss=0.459, simple_loss=0.463, pruned_loss=0.2276, over 19672.00 frames. ], tot_loss[loss=0.3856, simple_loss=0.4121, pruned_loss=0.1796, over 866670.85 frames. ], batch size: 59, lr: 3.76e-02, grad_scale: 8.0 +2023-03-31 21:14:21,609 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-03-31 21:14:37,451 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6893.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:14:41,583 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6896.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 21:14:43,759 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.5238, 5.0981, 3.0291, 4.6307, 2.0066, 5.2525, 4.9111, 5.0963], + device='cuda:3'), covar=tensor([0.0380, 0.0774, 0.1800, 0.0492, 0.2956, 0.0602, 0.0413, 0.0453], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0242, 0.0279, 0.0236, 0.0304, 0.0235, 0.0182, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 21:14:49,464 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6903.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:14:54,417 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-03-31 21:14:57,929 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.488e+02 9.076e+02 1.150e+03 1.515e+03 2.802e+03, threshold=2.301e+03, percent-clipped=3.0 +2023-03-31 21:15:19,073 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6927.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:15:20,161 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6928.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:15:20,722 INFO [train.py:903] (3/4) Epoch 2, batch 100, loss[loss=0.4472, simple_loss=0.4695, pruned_loss=0.2125, over 19297.00 frames. ], tot_loss[loss=0.3893, simple_loss=0.4142, pruned_loss=0.1822, over 1530710.39 frames. ], batch size: 66, lr: 3.75e-02, grad_scale: 8.0 +2023-03-31 21:15:30,198 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6936.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:15:32,200 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-03-31 21:15:33,673 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6939.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:15:39,467 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6944.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:15:39,538 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6944.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:16:01,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6961.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:16:10,937 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6969.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:16:15,271 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3858, 0.9155, 1.3150, 1.2592, 2.1813, 1.0417, 1.7694, 1.9858], + device='cuda:3'), covar=tensor([0.0439, 0.2663, 0.2547, 0.1614, 0.0536, 0.1994, 0.0953, 0.0637], + device='cuda:3'), in_proj_covar=tensor([0.0189, 0.0286, 0.0273, 0.0265, 0.0224, 0.0315, 0.0242, 0.0231], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:16:23,040 INFO [train.py:903] (3/4) Epoch 2, batch 150, loss[loss=0.4502, simple_loss=0.458, pruned_loss=0.2212, over 19687.00 frames. ], tot_loss[loss=0.3877, simple_loss=0.4137, pruned_loss=0.1809, over 2037730.72 frames. ], batch size: 59, lr: 3.74e-02, grad_scale: 4.0 +2023-03-31 21:16:38,556 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=6991.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:16:49,270 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6999.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:17:03,027 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.128e+02 7.870e+02 9.855e+02 1.288e+03 4.108e+03, threshold=1.971e+03, percent-clipped=4.0 +2023-03-31 21:17:19,729 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7024.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:17:24,043 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-03-31 21:17:25,199 INFO [train.py:903] (3/4) Epoch 2, batch 200, loss[loss=0.436, simple_loss=0.4566, pruned_loss=0.2077, over 19075.00 frames. ], tot_loss[loss=0.3865, simple_loss=0.4133, pruned_loss=0.1798, over 2420734.94 frames. ], batch size: 69, lr: 3.73e-02, grad_scale: 4.0 +2023-03-31 21:17:41,272 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-31 21:18:11,622 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7066.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 21:18:29,124 INFO [train.py:903] (3/4) Epoch 2, batch 250, loss[loss=0.4169, simple_loss=0.4216, pruned_loss=0.2061, over 19476.00 frames. ], tot_loss[loss=0.3853, simple_loss=0.4125, pruned_loss=0.1791, over 2736844.66 frames. ], batch size: 49, lr: 3.72e-02, grad_scale: 4.0 +2023-03-31 21:19:03,872 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7106.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:19:09,458 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.258e+02 7.747e+02 9.740e+02 1.157e+03 2.695e+03, threshold=1.948e+03, percent-clipped=1.0 +2023-03-31 21:19:33,825 INFO [train.py:903] (3/4) Epoch 2, batch 300, loss[loss=0.4502, simple_loss=0.4594, pruned_loss=0.2205, over 19785.00 frames. ], tot_loss[loss=0.3801, simple_loss=0.4093, pruned_loss=0.1755, over 2979747.81 frames. ], batch size: 56, lr: 3.72e-02, grad_scale: 4.0 +2023-03-31 21:20:15,178 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5535, 1.4816, 1.2091, 1.5666, 1.7563, 1.7210, 1.4072, 1.7338], + device='cuda:3'), covar=tensor([0.0945, 0.1849, 0.1577, 0.1232, 0.1422, 0.0587, 0.1357, 0.0721], + device='cuda:3'), in_proj_covar=tensor([0.0237, 0.0368, 0.0283, 0.0247, 0.0314, 0.0245, 0.0271, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:20:35,441 INFO [train.py:903] (3/4) Epoch 2, batch 350, loss[loss=0.3833, simple_loss=0.4092, pruned_loss=0.1787, over 19782.00 frames. ], tot_loss[loss=0.382, simple_loss=0.4104, pruned_loss=0.1768, over 3179632.54 frames. ], batch size: 56, lr: 3.71e-02, grad_scale: 4.0 +2023-03-31 21:20:37,924 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7181.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:20:39,487 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-31 21:20:39,881 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-03-31 21:20:40,463 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-03-31 21:20:55,887 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7195.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:21:01,703 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7200.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:21:15,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.882e+02 9.781e+02 1.245e+03 1.512e+03 3.081e+03, threshold=2.489e+03, percent-clipped=8.0 +2023-03-31 21:21:26,543 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7220.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:21:32,339 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7225.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:21:37,131 INFO [train.py:903] (3/4) Epoch 2, batch 400, loss[loss=0.5256, simple_loss=0.5042, pruned_loss=0.2735, over 18174.00 frames. ], tot_loss[loss=0.3846, simple_loss=0.4122, pruned_loss=0.1785, over 3319514.48 frames. ], batch size: 83, lr: 3.70e-02, grad_scale: 8.0 +2023-03-31 21:21:37,588 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4377, 2.3143, 1.7167, 1.8049, 1.7993, 1.0248, 1.0098, 1.7947], + device='cuda:3'), covar=tensor([0.1070, 0.0383, 0.0934, 0.0654, 0.0724, 0.1316, 0.1166, 0.0618], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0164, 0.0245, 0.0239, 0.0174, 0.0270, 0.0256, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:21:51,122 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7240.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:22:30,123 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7271.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:22:40,128 INFO [train.py:903] (3/4) Epoch 2, batch 450, loss[loss=0.3501, simple_loss=0.401, pruned_loss=0.1497, over 19786.00 frames. ], tot_loss[loss=0.3831, simple_loss=0.4104, pruned_loss=0.1779, over 3438466.19 frames. ], batch size: 56, lr: 3.69e-02, grad_scale: 8.0 +2023-03-31 21:22:58,507 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7293.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:23:14,155 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-03-31 21:23:15,293 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-03-31 21:23:19,470 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.877e+02 8.763e+02 1.192e+03 1.491e+03 2.950e+03, threshold=2.384e+03, percent-clipped=4.0 +2023-03-31 21:23:43,126 INFO [train.py:903] (3/4) Epoch 2, batch 500, loss[loss=0.4153, simple_loss=0.4385, pruned_loss=0.1961, over 19326.00 frames. ], tot_loss[loss=0.3797, simple_loss=0.4078, pruned_loss=0.1758, over 3534038.61 frames. ], batch size: 66, lr: 3.68e-02, grad_scale: 8.0 +2023-03-31 21:24:14,696 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7355.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:24:23,792 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7362.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:24:37,512 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8752, 3.7589, 4.3775, 4.3129, 1.4710, 3.7663, 3.6683, 3.7523], + device='cuda:3'), covar=tensor([0.0280, 0.0561, 0.0450, 0.0228, 0.3129, 0.0226, 0.0338, 0.0923], + device='cuda:3'), in_proj_covar=tensor([0.0216, 0.0255, 0.0323, 0.0218, 0.0396, 0.0155, 0.0231, 0.0332], + device='cuda:3'), out_proj_covar=tensor([1.3372e-04, 1.5776e-04, 2.0237e-04, 1.2543e-04, 2.1244e-04, 9.9618e-05, + 1.3496e-04, 1.8547e-04], device='cuda:3') +2023-03-31 21:24:45,285 INFO [train.py:903] (3/4) Epoch 2, batch 550, loss[loss=0.317, simple_loss=0.361, pruned_loss=0.1365, over 19615.00 frames. ], tot_loss[loss=0.3807, simple_loss=0.4091, pruned_loss=0.1762, over 3594453.31 frames. ], batch size: 50, lr: 3.67e-02, grad_scale: 8.0 +2023-03-31 21:24:53,947 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7386.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:24:55,120 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7387.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:25:25,335 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7410.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 21:25:26,118 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.968e+02 9.206e+02 1.127e+03 1.377e+03 2.659e+03, threshold=2.254e+03, percent-clipped=2.0 +2023-03-31 21:25:36,563 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7799, 1.1234, 1.0493, 1.6953, 1.4071, 1.8331, 1.6925, 1.5609], + device='cuda:3'), covar=tensor([0.1003, 0.1643, 0.1816, 0.1513, 0.1922, 0.1103, 0.1621, 0.1144], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0309, 0.0304, 0.0329, 0.0378, 0.0276, 0.0343, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 21:25:43,322 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7425.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:25:47,657 INFO [train.py:903] (3/4) Epoch 2, batch 600, loss[loss=0.3452, simple_loss=0.3808, pruned_loss=0.1548, over 19835.00 frames. ], tot_loss[loss=0.3822, simple_loss=0.4106, pruned_loss=0.1769, over 3649469.69 frames. ], batch size: 52, lr: 3.66e-02, grad_scale: 8.0 +2023-03-31 21:26:29,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-03-31 21:26:50,581 INFO [train.py:903] (3/4) Epoch 2, batch 650, loss[loss=0.3196, simple_loss=0.3612, pruned_loss=0.139, over 19776.00 frames. ], tot_loss[loss=0.38, simple_loss=0.4094, pruned_loss=0.1754, over 3695484.94 frames. ], batch size: 48, lr: 3.66e-02, grad_scale: 8.0 +2023-03-31 21:27:30,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.799e+02 8.519e+02 1.041e+03 1.431e+03 3.840e+03, threshold=2.082e+03, percent-clipped=3.0 +2023-03-31 21:27:49,412 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7525.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:27:49,645 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7525.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 21:27:53,960 INFO [train.py:903] (3/4) Epoch 2, batch 700, loss[loss=0.3297, simple_loss=0.3802, pruned_loss=0.1396, over 19855.00 frames. ], tot_loss[loss=0.3814, simple_loss=0.4103, pruned_loss=0.1763, over 3725621.49 frames. ], batch size: 52, lr: 3.65e-02, grad_scale: 8.0 +2023-03-31 21:28:56,971 INFO [train.py:903] (3/4) Epoch 2, batch 750, loss[loss=0.336, simple_loss=0.3715, pruned_loss=0.1503, over 19717.00 frames. ], tot_loss[loss=0.382, simple_loss=0.4105, pruned_loss=0.1767, over 3752740.02 frames. ], batch size: 45, lr: 3.64e-02, grad_scale: 8.0 +2023-03-31 21:29:24,871 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8780, 4.3896, 5.6575, 5.3985, 1.7649, 5.1474, 4.6216, 4.7584], + device='cuda:3'), covar=tensor([0.0255, 0.0502, 0.0336, 0.0196, 0.3306, 0.0158, 0.0281, 0.0984], + device='cuda:3'), in_proj_covar=tensor([0.0226, 0.0261, 0.0330, 0.0224, 0.0398, 0.0162, 0.0234, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 21:29:35,964 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.266e+02 8.676e+02 1.032e+03 1.220e+03 3.020e+03, threshold=2.064e+03, percent-clipped=5.0 +2023-03-31 21:29:36,446 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7611.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:29:55,639 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7341, 1.5375, 1.2996, 1.9006, 1.3702, 1.6242, 1.6051, 2.0160], + device='cuda:3'), covar=tensor([0.0875, 0.1940, 0.1535, 0.1173, 0.1787, 0.0723, 0.1270, 0.0560], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0370, 0.0290, 0.0253, 0.0327, 0.0259, 0.0285, 0.0215], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:29:58,618 INFO [train.py:903] (3/4) Epoch 2, batch 800, loss[loss=0.2923, simple_loss=0.3382, pruned_loss=0.1231, over 19420.00 frames. ], tot_loss[loss=0.38, simple_loss=0.4097, pruned_loss=0.1752, over 3778506.89 frames. ], batch size: 48, lr: 3.63e-02, grad_scale: 8.0 +2023-03-31 21:30:01,397 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5842, 1.1754, 1.0079, 1.6388, 1.1035, 1.5891, 1.5799, 1.3572], + device='cuda:3'), covar=tensor([0.1099, 0.1652, 0.2028, 0.1189, 0.1867, 0.1292, 0.1553, 0.1063], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0303, 0.0311, 0.0329, 0.0378, 0.0274, 0.0340, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 21:30:08,340 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7636.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:30:09,191 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7637.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:30:11,716 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0970, 1.1889, 2.3010, 1.3001, 3.0514, 3.5116, 3.7047, 1.9009], + device='cuda:3'), covar=tensor([0.1449, 0.1951, 0.1334, 0.1439, 0.0965, 0.0660, 0.1155, 0.1856], + device='cuda:3'), in_proj_covar=tensor([0.0314, 0.0321, 0.0304, 0.0314, 0.0352, 0.0286, 0.0408, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 21:30:12,839 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7640.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:30:15,289 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7642.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:30:15,427 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7642.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:30:16,182 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-03-31 21:30:46,842 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7667.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:31:01,249 INFO [train.py:903] (3/4) Epoch 2, batch 850, loss[loss=0.3233, simple_loss=0.3818, pruned_loss=0.1324, over 19662.00 frames. ], tot_loss[loss=0.3791, simple_loss=0.4093, pruned_loss=0.1744, over 3792890.80 frames. ], batch size: 60, lr: 3.62e-02, grad_scale: 8.0 +2023-03-31 21:31:41,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.436e+02 9.016e+02 1.057e+03 1.450e+03 5.160e+03, threshold=2.114e+03, percent-clipped=6.0 +2023-03-31 21:31:56,582 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-03-31 21:32:02,491 INFO [train.py:903] (3/4) Epoch 2, batch 900, loss[loss=0.4582, simple_loss=0.4662, pruned_loss=0.2251, over 19421.00 frames. ], tot_loss[loss=0.3802, simple_loss=0.4096, pruned_loss=0.1754, over 3807434.63 frames. ], batch size: 70, lr: 3.61e-02, grad_scale: 4.0 +2023-03-31 21:32:32,765 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7752.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:32:43,849 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7761.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:32:53,091 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7769.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:33:06,313 INFO [train.py:903] (3/4) Epoch 2, batch 950, loss[loss=0.3565, simple_loss=0.3804, pruned_loss=0.1663, over 19783.00 frames. ], tot_loss[loss=0.382, simple_loss=0.4107, pruned_loss=0.1767, over 3806376.64 frames. ], batch size: 48, lr: 3.61e-02, grad_scale: 4.0 +2023-03-31 21:33:09,127 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7781.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:33:10,165 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0233, 1.2765, 1.2924, 2.1041, 1.5942, 2.0052, 2.0946, 1.7940], + device='cuda:3'), covar=tensor([0.0795, 0.1645, 0.1752, 0.1202, 0.1743, 0.1079, 0.1360, 0.0945], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0305, 0.0304, 0.0329, 0.0377, 0.0272, 0.0338, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 21:33:10,898 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-03-31 21:33:24,393 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-31 21:33:40,022 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7806.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 21:33:46,539 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.660e+02 8.698e+02 1.089e+03 1.494e+03 2.916e+03, threshold=2.178e+03, percent-clipped=6.0 +2023-03-31 21:33:58,938 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7820.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:34:07,586 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6638, 4.2300, 2.4737, 3.8680, 1.2334, 4.0494, 3.7953, 3.9796], + device='cuda:3'), covar=tensor([0.0621, 0.1190, 0.2114, 0.0653, 0.3653, 0.0865, 0.0749, 0.0605], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0263, 0.0294, 0.0244, 0.0310, 0.0249, 0.0198, 0.0216], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 21:34:09,637 INFO [train.py:903] (3/4) Epoch 2, batch 1000, loss[loss=0.3106, simple_loss=0.3475, pruned_loss=0.1368, over 17381.00 frames. ], tot_loss[loss=0.3801, simple_loss=0.4091, pruned_loss=0.1755, over 3800221.92 frames. ], batch size: 38, lr: 3.60e-02, grad_scale: 4.0 +2023-03-31 21:35:04,945 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-03-31 21:35:11,840 INFO [train.py:903] (3/4) Epoch 2, batch 1050, loss[loss=0.4215, simple_loss=0.4428, pruned_loss=0.2001, over 19590.00 frames. ], tot_loss[loss=0.3786, simple_loss=0.4082, pruned_loss=0.1745, over 3811575.54 frames. ], batch size: 61, lr: 3.59e-02, grad_scale: 4.0 +2023-03-31 21:35:18,873 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7884.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:35:31,653 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0640, 1.8100, 1.8420, 2.9521, 2.1156, 3.1795, 2.1826, 1.7119], + device='cuda:3'), covar=tensor([0.1089, 0.0865, 0.0598, 0.0450, 0.0985, 0.0243, 0.1158, 0.0988], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0233, 0.0254, 0.0322, 0.0324, 0.0171, 0.0346, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:35:33,753 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7896.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:35:46,820 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-03-31 21:35:53,415 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.580e+02 8.878e+02 9.952e+02 1.228e+03 3.126e+03, threshold=1.990e+03, percent-clipped=5.0 +2023-03-31 21:35:56,209 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4602, 0.9740, 1.3893, 1.3479, 2.2093, 1.1077, 1.8661, 2.1429], + device='cuda:3'), covar=tensor([0.0476, 0.2719, 0.2340, 0.1534, 0.0496, 0.1825, 0.0801, 0.0609], + device='cuda:3'), in_proj_covar=tensor([0.0206, 0.0294, 0.0276, 0.0270, 0.0233, 0.0311, 0.0254, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:36:05,137 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7921.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:36:14,159 INFO [train.py:903] (3/4) Epoch 2, batch 1100, loss[loss=0.3416, simple_loss=0.3868, pruned_loss=0.1482, over 19532.00 frames. ], tot_loss[loss=0.3788, simple_loss=0.4087, pruned_loss=0.1745, over 3825707.18 frames. ], batch size: 54, lr: 3.58e-02, grad_scale: 4.0 +2023-03-31 21:37:16,171 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4786, 0.7951, 1.2288, 0.7165, 2.5050, 2.5359, 2.3335, 2.7061], + device='cuda:3'), covar=tensor([0.2149, 0.5067, 0.4146, 0.3313, 0.0554, 0.0329, 0.0579, 0.0291], + device='cuda:3'), in_proj_covar=tensor([0.0298, 0.0282, 0.0325, 0.0295, 0.0197, 0.0109, 0.0185, 0.0114], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 21:37:16,834 INFO [train.py:903] (3/4) Epoch 2, batch 1150, loss[loss=0.5199, simple_loss=0.4986, pruned_loss=0.2706, over 19561.00 frames. ], tot_loss[loss=0.3808, simple_loss=0.4102, pruned_loss=0.1757, over 3822894.69 frames. ], batch size: 61, lr: 3.57e-02, grad_scale: 4.0 +2023-03-31 21:37:26,764 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=7986.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:37:54,512 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8008.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:37:58,746 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.862e+02 8.681e+02 1.035e+03 1.273e+03 2.854e+03, threshold=2.070e+03, percent-clipped=5.0 +2023-03-31 21:38:21,515 INFO [train.py:903] (3/4) Epoch 2, batch 1200, loss[loss=0.3931, simple_loss=0.4215, pruned_loss=0.1823, over 19485.00 frames. ], tot_loss[loss=0.3776, simple_loss=0.4083, pruned_loss=0.1735, over 3840310.24 frames. ], batch size: 64, lr: 3.56e-02, grad_scale: 8.0 +2023-03-31 21:38:26,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8033.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:38:35,980 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-31 21:38:52,456 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-03-31 21:39:22,161 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-03-31 21:39:22,669 INFO [train.py:903] (3/4) Epoch 2, batch 1250, loss[loss=0.3593, simple_loss=0.393, pruned_loss=0.1628, over 19681.00 frames. ], tot_loss[loss=0.3796, simple_loss=0.4098, pruned_loss=0.1747, over 3853310.41 frames. ], batch size: 53, lr: 3.56e-02, grad_scale: 8.0 +2023-03-31 21:39:35,508 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7722, 3.6852, 3.9155, 4.0320, 1.5044, 3.7296, 3.4378, 3.1589], + device='cuda:3'), covar=tensor([0.0473, 0.0764, 0.0877, 0.0490, 0.3704, 0.0425, 0.0585, 0.1580], + device='cuda:3'), in_proj_covar=tensor([0.0227, 0.0248, 0.0327, 0.0225, 0.0383, 0.0157, 0.0227, 0.0342], + device='cuda:3'), out_proj_covar=tensor([1.3801e-04, 1.5211e-04, 2.0109e-04, 1.2689e-04, 2.0383e-04, 9.8941e-05, + 1.2860e-04, 1.8633e-04], device='cuda:3') +2023-03-31 21:39:51,765 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8101.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:39:57,187 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=8105.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:40:05,139 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.516e+02 8.392e+02 1.041e+03 1.254e+03 3.427e+03, threshold=2.083e+03, percent-clipped=3.0 +2023-03-31 21:40:25,464 INFO [train.py:903] (3/4) Epoch 2, batch 1300, loss[loss=0.3172, simple_loss=0.3636, pruned_loss=0.1354, over 19763.00 frames. ], tot_loss[loss=0.377, simple_loss=0.4078, pruned_loss=0.1731, over 3847918.91 frames. ], batch size: 45, lr: 3.55e-02, grad_scale: 8.0 +2023-03-31 21:40:40,991 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8140.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:40:47,742 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-03-31 21:40:49,843 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4922, 2.3925, 1.7983, 1.8659, 2.0357, 1.0856, 0.7683, 1.5422], + device='cuda:3'), covar=tensor([0.1136, 0.0407, 0.0948, 0.0555, 0.0802, 0.1416, 0.1255, 0.0822], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0168, 0.0251, 0.0231, 0.0171, 0.0274, 0.0255, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:41:09,973 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=8164.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:41:11,430 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8165.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:41:28,089 INFO [train.py:903] (3/4) Epoch 2, batch 1350, loss[loss=0.2886, simple_loss=0.3368, pruned_loss=0.1202, over 19392.00 frames. ], tot_loss[loss=0.3798, simple_loss=0.4098, pruned_loss=0.1749, over 3827396.15 frames. ], batch size: 47, lr: 3.54e-02, grad_scale: 8.0 +2023-03-31 21:42:08,820 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.809e+02 9.088e+02 1.109e+03 1.527e+03 2.312e+03, threshold=2.218e+03, percent-clipped=6.0 +2023-03-31 21:42:20,861 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8220.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:42:30,891 INFO [train.py:903] (3/4) Epoch 2, batch 1400, loss[loss=0.393, simple_loss=0.4269, pruned_loss=0.1796, over 19073.00 frames. ], tot_loss[loss=0.3785, simple_loss=0.4087, pruned_loss=0.1742, over 3809277.00 frames. ], batch size: 69, lr: 3.53e-02, grad_scale: 8.0 +2023-03-31 21:43:32,692 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-03-31 21:43:33,840 INFO [train.py:903] (3/4) Epoch 2, batch 1450, loss[loss=0.3796, simple_loss=0.4139, pruned_loss=0.1726, over 19681.00 frames. ], tot_loss[loss=0.3765, simple_loss=0.4077, pruned_loss=0.1727, over 3820109.80 frames. ], batch size: 59, lr: 3.53e-02, grad_scale: 8.0 +2023-03-31 21:43:34,161 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8279.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:43:39,548 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8283.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:44:15,432 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.216e+02 8.724e+02 1.078e+03 1.353e+03 2.729e+03, threshold=2.156e+03, percent-clipped=3.0 +2023-03-31 21:44:34,914 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8798, 1.2502, 1.3649, 1.7278, 2.5426, 1.2137, 1.9234, 2.5856], + device='cuda:3'), covar=tensor([0.0500, 0.2747, 0.2757, 0.1674, 0.0548, 0.2205, 0.1057, 0.0525], + device='cuda:3'), in_proj_covar=tensor([0.0221, 0.0301, 0.0285, 0.0276, 0.0245, 0.0318, 0.0258, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:44:35,647 INFO [train.py:903] (3/4) Epoch 2, batch 1500, loss[loss=0.4194, simple_loss=0.4209, pruned_loss=0.209, over 19829.00 frames. ], tot_loss[loss=0.3778, simple_loss=0.4086, pruned_loss=0.1735, over 3818912.60 frames. ], batch size: 52, lr: 3.52e-02, grad_scale: 8.0 +2023-03-31 21:45:10,940 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8357.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:45:38,215 INFO [train.py:903] (3/4) Epoch 2, batch 1550, loss[loss=0.3737, simple_loss=0.399, pruned_loss=0.1742, over 19595.00 frames. ], tot_loss[loss=0.3794, simple_loss=0.4092, pruned_loss=0.1748, over 3826692.41 frames. ], batch size: 50, lr: 3.51e-02, grad_scale: 8.0 +2023-03-31 21:45:43,354 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8382.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:46:19,241 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.712e+02 9.525e+02 1.175e+03 1.582e+03 3.285e+03, threshold=2.351e+03, percent-clipped=5.0 +2023-03-31 21:46:41,074 INFO [train.py:903] (3/4) Epoch 2, batch 1600, loss[loss=0.3237, simple_loss=0.3647, pruned_loss=0.1413, over 19425.00 frames. ], tot_loss[loss=0.3763, simple_loss=0.4072, pruned_loss=0.1728, over 3836434.15 frames. ], batch size: 48, lr: 3.50e-02, grad_scale: 8.0 +2023-03-31 21:47:00,247 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1424, 2.6974, 1.8126, 2.3772, 2.0078, 1.7866, 0.3493, 2.0155], + device='cuda:3'), covar=tensor([0.0583, 0.0488, 0.0442, 0.0497, 0.0834, 0.0988, 0.1504, 0.0961], + device='cuda:3'), in_proj_covar=tensor([0.0195, 0.0202, 0.0192, 0.0230, 0.0270, 0.0239, 0.0244, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:47:02,107 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-03-31 21:47:39,685 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8476.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:47:42,847 INFO [train.py:903] (3/4) Epoch 2, batch 1650, loss[loss=0.4384, simple_loss=0.4508, pruned_loss=0.213, over 19576.00 frames. ], tot_loss[loss=0.3755, simple_loss=0.4069, pruned_loss=0.1721, over 3836091.26 frames. ], batch size: 61, lr: 3.49e-02, grad_scale: 8.0 +2023-03-31 21:48:10,082 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8501.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:48:23,196 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.971e+02 8.791e+02 1.048e+03 1.403e+03 4.696e+03, threshold=2.096e+03, percent-clipped=2.0 +2023-03-31 21:48:44,128 INFO [train.py:903] (3/4) Epoch 2, batch 1700, loss[loss=0.3149, simple_loss=0.3516, pruned_loss=0.1391, over 19070.00 frames. ], tot_loss[loss=0.376, simple_loss=0.4068, pruned_loss=0.1726, over 3841664.44 frames. ], batch size: 42, lr: 3.49e-02, grad_scale: 8.0 +2023-03-31 21:48:52,096 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8535.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:49:23,172 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8560.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:49:23,926 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-03-31 21:49:46,288 INFO [train.py:903] (3/4) Epoch 2, batch 1750, loss[loss=0.3733, simple_loss=0.4144, pruned_loss=0.1661, over 19593.00 frames. ], tot_loss[loss=0.3752, simple_loss=0.4063, pruned_loss=0.172, over 3841909.39 frames. ], batch size: 61, lr: 3.48e-02, grad_scale: 8.0 +2023-03-31 21:50:27,274 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.711e+02 8.589e+02 1.062e+03 1.367e+03 2.706e+03, threshold=2.124e+03, percent-clipped=6.0 +2023-03-31 21:50:47,026 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=8627.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:50:48,939 INFO [train.py:903] (3/4) Epoch 2, batch 1800, loss[loss=0.3533, simple_loss=0.3675, pruned_loss=0.1695, over 19711.00 frames. ], tot_loss[loss=0.3736, simple_loss=0.4054, pruned_loss=0.1709, over 3842848.58 frames. ], batch size: 46, lr: 3.47e-02, grad_scale: 8.0 +2023-03-31 21:51:48,931 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-03-31 21:51:52,354 INFO [train.py:903] (3/4) Epoch 2, batch 1850, loss[loss=0.3588, simple_loss=0.4053, pruned_loss=0.1562, over 19709.00 frames. ], tot_loss[loss=0.3745, simple_loss=0.4064, pruned_loss=0.1713, over 3830969.14 frames. ], batch size: 59, lr: 3.46e-02, grad_scale: 8.0 +2023-03-31 21:51:59,655 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7992, 1.3034, 1.1546, 1.6644, 1.5261, 2.0382, 2.1311, 2.1700], + device='cuda:3'), covar=tensor([0.0984, 0.1558, 0.1860, 0.1627, 0.1918, 0.0986, 0.1375, 0.0771], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0296, 0.0302, 0.0320, 0.0364, 0.0264, 0.0334, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 21:52:26,732 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-03-31 21:52:32,409 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.683e+02 8.545e+02 1.022e+03 1.402e+03 2.945e+03, threshold=2.044e+03, percent-clipped=4.0 +2023-03-31 21:52:53,431 INFO [train.py:903] (3/4) Epoch 2, batch 1900, loss[loss=0.3865, simple_loss=0.4006, pruned_loss=0.1862, over 19619.00 frames. ], tot_loss[loss=0.3749, simple_loss=0.4068, pruned_loss=0.1715, over 3833770.36 frames. ], batch size: 50, lr: 3.46e-02, grad_scale: 8.0 +2023-03-31 21:52:53,873 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7155, 3.6950, 2.0826, 2.5081, 3.0110, 1.2657, 1.1765, 1.6536], + device='cuda:3'), covar=tensor([0.1799, 0.0366, 0.1202, 0.0765, 0.0790, 0.1656, 0.1534, 0.1341], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0177, 0.0266, 0.0247, 0.0174, 0.0291, 0.0266, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 21:53:10,060 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8742.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:53:13,049 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-03-31 21:53:18,232 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9497, 1.7015, 1.9477, 1.9141, 2.8868, 3.7151, 3.7604, 3.9649], + device='cuda:3'), covar=tensor([0.1440, 0.2374, 0.2541, 0.1744, 0.0523, 0.0180, 0.0192, 0.0200], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0286, 0.0335, 0.0298, 0.0201, 0.0111, 0.0193, 0.0118], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 21:53:19,070 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-03-31 21:53:20,894 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-31 21:53:44,527 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-03-31 21:53:56,652 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8778.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:53:57,544 INFO [train.py:903] (3/4) Epoch 2, batch 1950, loss[loss=0.3827, simple_loss=0.411, pruned_loss=0.1772, over 19663.00 frames. ], tot_loss[loss=0.3725, simple_loss=0.405, pruned_loss=0.17, over 3833004.13 frames. ], batch size: 53, lr: 3.45e-02, grad_scale: 8.0 +2023-03-31 21:54:38,567 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.007e+02 8.441e+02 1.013e+03 1.280e+03 2.038e+03, threshold=2.026e+03, percent-clipped=0.0 +2023-03-31 21:55:00,982 INFO [train.py:903] (3/4) Epoch 2, batch 2000, loss[loss=0.3599, simple_loss=0.4056, pruned_loss=0.1571, over 19664.00 frames. ], tot_loss[loss=0.3724, simple_loss=0.4049, pruned_loss=0.1699, over 3833245.45 frames. ], batch size: 55, lr: 3.44e-02, grad_scale: 8.0 +2023-03-31 21:55:24,882 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-03-31 21:56:00,575 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-03-31 21:56:01,581 INFO [train.py:903] (3/4) Epoch 2, batch 2050, loss[loss=0.4867, simple_loss=0.474, pruned_loss=0.2497, over 18281.00 frames. ], tot_loss[loss=0.3731, simple_loss=0.405, pruned_loss=0.1706, over 3842592.34 frames. ], batch size: 83, lr: 3.43e-02, grad_scale: 8.0 +2023-03-31 21:56:19,563 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-03-31 21:56:20,280 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-31 21:56:20,771 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-03-31 21:56:41,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-03-31 21:56:44,119 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.978e+02 1.029e+03 1.194e+03 1.427e+03 4.040e+03, threshold=2.389e+03, percent-clipped=7.0 +2023-03-31 21:57:05,506 INFO [train.py:903] (3/4) Epoch 2, batch 2100, loss[loss=0.3238, simple_loss=0.3709, pruned_loss=0.1384, over 19695.00 frames. ], tot_loss[loss=0.3725, simple_loss=0.4045, pruned_loss=0.1703, over 3826005.90 frames. ], batch size: 53, lr: 3.43e-02, grad_scale: 8.0 +2023-03-31 21:57:25,762 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8945.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:57:36,777 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-03-31 21:57:59,444 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-03-31 21:58:07,606 INFO [train.py:903] (3/4) Epoch 2, batch 2150, loss[loss=0.369, simple_loss=0.4095, pruned_loss=0.1643, over 19759.00 frames. ], tot_loss[loss=0.3707, simple_loss=0.4031, pruned_loss=0.1691, over 3817407.76 frames. ], batch size: 56, lr: 3.42e-02, grad_scale: 8.0 +2023-03-31 21:58:32,426 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8998.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:58:40,186 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9004.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:58:49,979 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.755e+02 7.392e+02 9.171e+02 1.181e+03 2.165e+03, threshold=1.834e+03, percent-clipped=0.0 +2023-03-31 21:59:04,449 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9023.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 21:59:11,992 INFO [train.py:903] (3/4) Epoch 2, batch 2200, loss[loss=0.3652, simple_loss=0.4063, pruned_loss=0.162, over 17419.00 frames. ], tot_loss[loss=0.3687, simple_loss=0.402, pruned_loss=0.1677, over 3811323.24 frames. ], batch size: 101, lr: 3.41e-02, grad_scale: 8.0 +2023-03-31 21:59:14,478 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9031.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:00:13,982 INFO [train.py:903] (3/4) Epoch 2, batch 2250, loss[loss=0.3809, simple_loss=0.4151, pruned_loss=0.1734, over 19756.00 frames. ], tot_loss[loss=0.3675, simple_loss=0.4014, pruned_loss=0.1669, over 3813662.23 frames. ], batch size: 63, lr: 3.41e-02, grad_scale: 8.0 +2023-03-31 22:00:56,144 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.993e+02 8.727e+02 9.943e+02 1.293e+03 2.077e+03, threshold=1.989e+03, percent-clipped=4.0 +2023-03-31 22:01:08,152 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9122.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:01:17,886 INFO [train.py:903] (3/4) Epoch 2, batch 2300, loss[loss=0.31, simple_loss=0.3617, pruned_loss=0.1291, over 19685.00 frames. ], tot_loss[loss=0.365, simple_loss=0.3994, pruned_loss=0.1654, over 3818363.28 frames. ], batch size: 53, lr: 3.40e-02, grad_scale: 8.0 +2023-03-31 22:01:30,558 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-03-31 22:02:19,096 INFO [train.py:903] (3/4) Epoch 2, batch 2350, loss[loss=0.331, simple_loss=0.3639, pruned_loss=0.1491, over 19779.00 frames. ], tot_loss[loss=0.3652, simple_loss=0.3993, pruned_loss=0.1656, over 3801445.00 frames. ], batch size: 48, lr: 3.39e-02, grad_scale: 8.0 +2023-03-31 22:03:00,732 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.988e+02 9.117e+02 1.090e+03 1.432e+03 2.529e+03, threshold=2.180e+03, percent-clipped=5.0 +2023-03-31 22:03:00,817 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-03-31 22:03:19,484 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-03-31 22:03:19,783 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1631, 2.7046, 1.9458, 2.7214, 2.0664, 2.2421, 0.4142, 2.1833], + device='cuda:3'), covar=tensor([0.0486, 0.0474, 0.0350, 0.0431, 0.0786, 0.0708, 0.1328, 0.0860], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0213, 0.0208, 0.0252, 0.0296, 0.0261, 0.0262, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:03:22,887 INFO [train.py:903] (3/4) Epoch 2, batch 2400, loss[loss=0.35, simple_loss=0.3716, pruned_loss=0.1642, over 16040.00 frames. ], tot_loss[loss=0.3675, simple_loss=0.4009, pruned_loss=0.1671, over 3796688.28 frames. ], batch size: 35, lr: 3.38e-02, grad_scale: 8.0 +2023-03-31 22:03:32,318 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9237.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:03:44,348 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.77 vs. limit=5.0 +2023-03-31 22:04:21,301 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1131, 1.3039, 2.1806, 1.4103, 2.9046, 3.1771, 3.1437, 1.8918], + device='cuda:3'), covar=tensor([0.1372, 0.1820, 0.1140, 0.1343, 0.0919, 0.0697, 0.1263, 0.1635], + device='cuda:3'), in_proj_covar=tensor([0.0334, 0.0360, 0.0333, 0.0341, 0.0387, 0.0305, 0.0454, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:04:24,304 INFO [train.py:903] (3/4) Epoch 2, batch 2450, loss[loss=0.3151, simple_loss=0.3587, pruned_loss=0.1357, over 19750.00 frames. ], tot_loss[loss=0.3687, simple_loss=0.4017, pruned_loss=0.1679, over 3787199.90 frames. ], batch size: 51, lr: 3.38e-02, grad_scale: 8.0 +2023-03-31 22:04:38,168 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9289.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:04:38,394 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0456, 1.2629, 1.8507, 1.7122, 2.9720, 4.4236, 4.6362, 4.8636], + device='cuda:3'), covar=tensor([0.1359, 0.2494, 0.2419, 0.1952, 0.0484, 0.0124, 0.0107, 0.0091], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0284, 0.0322, 0.0293, 0.0200, 0.0109, 0.0185, 0.0115], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 22:05:06,391 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.471e+02 8.791e+02 1.096e+03 1.484e+03 3.289e+03, threshold=2.192e+03, percent-clipped=7.0 +2023-03-31 22:05:27,995 INFO [train.py:903] (3/4) Epoch 2, batch 2500, loss[loss=0.3862, simple_loss=0.4203, pruned_loss=0.176, over 19680.00 frames. ], tot_loss[loss=0.3664, simple_loss=0.4006, pruned_loss=0.1661, over 3805549.83 frames. ], batch size: 58, lr: 3.37e-02, grad_scale: 8.0 +2023-03-31 22:05:52,450 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9348.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:06:25,085 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9375.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:06:29,686 INFO [train.py:903] (3/4) Epoch 2, batch 2550, loss[loss=0.3701, simple_loss=0.4093, pruned_loss=0.1654, over 19613.00 frames. ], tot_loss[loss=0.3675, simple_loss=0.4014, pruned_loss=0.1669, over 3799750.72 frames. ], batch size: 57, lr: 3.36e-02, grad_scale: 8.0 +2023-03-31 22:06:54,757 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5416, 3.8409, 4.0885, 3.9991, 1.4371, 3.6276, 3.3019, 3.5671], + device='cuda:3'), covar=tensor([0.0324, 0.0393, 0.0397, 0.0228, 0.2751, 0.0234, 0.0315, 0.0855], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0273, 0.0363, 0.0252, 0.0410, 0.0172, 0.0247, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 22:07:01,818 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9404.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:07:11,545 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.591e+02 8.134e+02 9.492e+02 1.275e+03 2.544e+03, threshold=1.898e+03, percent-clipped=3.0 +2023-03-31 22:07:18,176 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-03-31 22:07:25,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-03-31 22:07:33,275 INFO [train.py:903] (3/4) Epoch 2, batch 2600, loss[loss=0.3499, simple_loss=0.3977, pruned_loss=0.151, over 19766.00 frames. ], tot_loss[loss=0.3665, simple_loss=0.4004, pruned_loss=0.1663, over 3799818.01 frames. ], batch size: 56, lr: 3.36e-02, grad_scale: 8.0 +2023-03-31 22:07:36,955 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9432.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:08:14,545 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9463.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:08:23,778 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.19 vs. limit=2.0 +2023-03-31 22:08:34,690 INFO [train.py:903] (3/4) Epoch 2, batch 2650, loss[loss=0.3624, simple_loss=0.4008, pruned_loss=0.162, over 19664.00 frames. ], tot_loss[loss=0.3672, simple_loss=0.4007, pruned_loss=0.1669, over 3797857.30 frames. ], batch size: 53, lr: 3.35e-02, grad_scale: 8.0 +2023-03-31 22:08:49,063 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9490.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:08:52,759 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9493.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:08:53,571 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-03-31 22:09:16,685 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.258e+02 8.887e+02 1.023e+03 1.383e+03 3.476e+03, threshold=2.047e+03, percent-clipped=7.0 +2023-03-31 22:09:24,065 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9518.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:09:36,956 INFO [train.py:903] (3/4) Epoch 2, batch 2700, loss[loss=0.3468, simple_loss=0.3697, pruned_loss=0.1619, over 19070.00 frames. ], tot_loss[loss=0.366, simple_loss=0.3998, pruned_loss=0.1661, over 3808844.48 frames. ], batch size: 42, lr: 3.34e-02, grad_scale: 8.0 +2023-03-31 22:09:43,946 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5175, 3.7606, 4.0487, 3.8905, 1.3732, 3.6004, 3.2408, 3.5529], + device='cuda:3'), covar=tensor([0.0365, 0.0503, 0.0440, 0.0285, 0.3095, 0.0236, 0.0369, 0.0871], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0273, 0.0361, 0.0256, 0.0414, 0.0175, 0.0249, 0.0367], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 22:10:39,551 INFO [train.py:903] (3/4) Epoch 2, batch 2750, loss[loss=0.3817, simple_loss=0.4162, pruned_loss=0.1736, over 17543.00 frames. ], tot_loss[loss=0.3651, simple_loss=0.3997, pruned_loss=0.1653, over 3815945.62 frames. ], batch size: 101, lr: 3.34e-02, grad_scale: 8.0 +2023-03-31 22:10:49,084 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1788, 0.8052, 1.1488, 0.4420, 2.3517, 2.0089, 1.7867, 2.1793], + device='cuda:3'), covar=tensor([0.1796, 0.3821, 0.3549, 0.3023, 0.0443, 0.0235, 0.0461, 0.0239], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0282, 0.0319, 0.0291, 0.0199, 0.0107, 0.0185, 0.0114], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 22:11:01,855 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5959, 0.9631, 1.1890, 1.8196, 1.6186, 1.5924, 2.0008, 1.4839], + device='cuda:3'), covar=tensor([0.1397, 0.2508, 0.2207, 0.1531, 0.1881, 0.1766, 0.1620, 0.1441], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0310, 0.0309, 0.0327, 0.0364, 0.0275, 0.0342, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 22:11:08,842 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9602.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:11:20,989 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.046e+02 9.047e+02 1.065e+03 1.297e+03 2.590e+03, threshold=2.130e+03, percent-clipped=3.0 +2023-03-31 22:11:43,321 INFO [train.py:903] (3/4) Epoch 2, batch 2800, loss[loss=0.4547, simple_loss=0.4562, pruned_loss=0.2266, over 13969.00 frames. ], tot_loss[loss=0.364, simple_loss=0.3992, pruned_loss=0.1644, over 3812090.15 frames. ], batch size: 137, lr: 3.33e-02, grad_scale: 8.0 +2023-03-31 22:12:20,976 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9660.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:12:41,317 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9676.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:12:44,618 INFO [train.py:903] (3/4) Epoch 2, batch 2850, loss[loss=0.3552, simple_loss=0.4002, pruned_loss=0.1551, over 19544.00 frames. ], tot_loss[loss=0.3645, simple_loss=0.3997, pruned_loss=0.1647, over 3818683.37 frames. ], batch size: 54, lr: 3.32e-02, grad_scale: 8.0 +2023-03-31 22:12:51,774 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9685.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:13:26,582 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 6.408e+02 8.898e+02 1.125e+03 1.384e+03 2.599e+03, threshold=2.251e+03, percent-clipped=6.0 +2023-03-31 22:13:35,213 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9719.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:13:43,488 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6580, 3.8134, 4.1422, 3.9959, 1.5348, 3.6484, 3.3383, 3.5947], + device='cuda:3'), covar=tensor([0.0357, 0.0463, 0.0370, 0.0271, 0.2854, 0.0209, 0.0411, 0.0932], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0281, 0.0375, 0.0263, 0.0421, 0.0183, 0.0260, 0.0379], + device='cuda:3'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 22:13:45,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-03-31 22:13:46,851 INFO [train.py:903] (3/4) Epoch 2, batch 2900, loss[loss=0.3653, simple_loss=0.4089, pruned_loss=0.1609, over 18748.00 frames. ], tot_loss[loss=0.3641, simple_loss=0.3994, pruned_loss=0.1644, over 3816457.70 frames. ], batch size: 74, lr: 3.31e-02, grad_scale: 16.0 +2023-03-31 22:14:06,836 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9744.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:14:10,305 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9746.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:14:23,242 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3018, 1.1209, 1.0406, 1.4143, 1.1669, 1.2952, 1.3406, 1.2684], + device='cuda:3'), covar=tensor([0.0984, 0.1539, 0.1788, 0.1033, 0.1442, 0.1297, 0.1215, 0.1085], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0307, 0.0303, 0.0332, 0.0361, 0.0273, 0.0331, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 22:14:27,853 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9761.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:14:39,311 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9771.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:14:45,623 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9776.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:14:49,545 INFO [train.py:903] (3/4) Epoch 2, batch 2950, loss[loss=0.3927, simple_loss=0.4124, pruned_loss=0.1865, over 19841.00 frames. ], tot_loss[loss=0.3658, simple_loss=0.4005, pruned_loss=0.1656, over 3813782.10 frames. ], batch size: 52, lr: 3.31e-02, grad_scale: 8.0 +2023-03-31 22:14:52,369 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-03-31 22:15:31,704 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.533e+02 8.475e+02 1.131e+03 1.413e+03 3.215e+03, threshold=2.262e+03, percent-clipped=4.0 +2023-03-31 22:15:53,078 INFO [train.py:903] (3/4) Epoch 2, batch 3000, loss[loss=0.3925, simple_loss=0.4184, pruned_loss=0.1833, over 18708.00 frames. ], tot_loss[loss=0.365, simple_loss=0.3997, pruned_loss=0.1651, over 3809602.00 frames. ], batch size: 74, lr: 3.30e-02, grad_scale: 4.0 +2023-03-31 22:15:53,078 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 22:16:06,232 INFO [train.py:937] (3/4) Epoch 2, validation: loss=0.2513, simple_loss=0.3423, pruned_loss=0.08019, over 944034.00 frames. +2023-03-31 22:16:06,233 INFO [train.py:938] (3/4) Maximum memory allocated so far is 17850MB +2023-03-31 22:16:12,130 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-03-31 22:17:08,084 INFO [train.py:903] (3/4) Epoch 2, batch 3050, loss[loss=0.355, simple_loss=0.3807, pruned_loss=0.1646, over 19476.00 frames. ], tot_loss[loss=0.3661, simple_loss=0.4005, pruned_loss=0.1659, over 3811836.39 frames. ], batch size: 49, lr: 3.29e-02, grad_scale: 4.0 +2023-03-31 22:17:22,810 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9891.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:17:50,736 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.558e+02 8.806e+02 1.158e+03 1.402e+03 3.282e+03, threshold=2.315e+03, percent-clipped=2.0 +2023-03-31 22:18:10,044 INFO [train.py:903] (3/4) Epoch 2, batch 3100, loss[loss=0.3902, simple_loss=0.4037, pruned_loss=0.1884, over 19630.00 frames. ], tot_loss[loss=0.3656, simple_loss=0.4, pruned_loss=0.1656, over 3812931.28 frames. ], batch size: 50, lr: 3.29e-02, grad_scale: 4.0 +2023-03-31 22:18:29,705 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=9946.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:19:11,239 INFO [train.py:903] (3/4) Epoch 2, batch 3150, loss[loss=0.2728, simple_loss=0.3283, pruned_loss=0.1087, over 19750.00 frames. ], tot_loss[loss=0.3634, simple_loss=0.3982, pruned_loss=0.1643, over 3819225.61 frames. ], batch size: 47, lr: 3.28e-02, grad_scale: 4.0 +2023-03-31 22:19:42,630 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-03-31 22:19:56,512 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.945e+02 7.345e+02 9.404e+02 1.247e+03 3.615e+03, threshold=1.881e+03, percent-clipped=3.0 +2023-03-31 22:20:03,838 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10020.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:20:15,210 INFO [train.py:903] (3/4) Epoch 2, batch 3200, loss[loss=0.3325, simple_loss=0.3616, pruned_loss=0.1517, over 19310.00 frames. ], tot_loss[loss=0.364, simple_loss=0.3987, pruned_loss=0.1647, over 3802337.15 frames. ], batch size: 44, lr: 3.27e-02, grad_scale: 8.0 +2023-03-31 22:20:50,858 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10057.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:20:55,735 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10061.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:21:19,936 INFO [train.py:903] (3/4) Epoch 2, batch 3250, loss[loss=0.4401, simple_loss=0.4604, pruned_loss=0.2099, over 19530.00 frames. ], tot_loss[loss=0.3638, simple_loss=0.3989, pruned_loss=0.1644, over 3783810.68 frames. ], batch size: 56, lr: 3.27e-02, grad_scale: 8.0 +2023-03-31 22:21:21,296 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10080.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:21:50,581 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10105.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:22:02,472 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.757e+02 8.486e+02 1.037e+03 1.291e+03 3.604e+03, threshold=2.074e+03, percent-clipped=6.0 +2023-03-31 22:22:20,716 INFO [train.py:903] (3/4) Epoch 2, batch 3300, loss[loss=0.3674, simple_loss=0.4113, pruned_loss=0.1618, over 19759.00 frames. ], tot_loss[loss=0.3614, simple_loss=0.3968, pruned_loss=0.163, over 3802563.09 frames. ], batch size: 63, lr: 3.26e-02, grad_scale: 8.0 +2023-03-31 22:22:25,215 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-03-31 22:22:27,933 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10135.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:22:42,649 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10147.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:23:15,000 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10172.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:23:22,725 INFO [train.py:903] (3/4) Epoch 2, batch 3350, loss[loss=0.5118, simple_loss=0.487, pruned_loss=0.2683, over 13098.00 frames. ], tot_loss[loss=0.3623, simple_loss=0.3975, pruned_loss=0.1635, over 3783768.60 frames. ], batch size: 136, lr: 3.26e-02, grad_scale: 8.0 +2023-03-31 22:24:07,629 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.516e+02 8.361e+02 9.909e+02 1.198e+03 2.844e+03, threshold=1.982e+03, percent-clipped=3.0 +2023-03-31 22:24:14,935 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10220.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:24:17,827 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-31 22:24:25,945 INFO [train.py:903] (3/4) Epoch 2, batch 3400, loss[loss=0.333, simple_loss=0.3853, pruned_loss=0.1404, over 19672.00 frames. ], tot_loss[loss=0.3614, simple_loss=0.3972, pruned_loss=0.1628, over 3794245.68 frames. ], batch size: 60, lr: 3.25e-02, grad_scale: 8.0 +2023-03-31 22:25:29,368 INFO [train.py:903] (3/4) Epoch 2, batch 3450, loss[loss=0.4057, simple_loss=0.4348, pruned_loss=0.1883, over 19653.00 frames. ], tot_loss[loss=0.3623, simple_loss=0.397, pruned_loss=0.1638, over 3802648.19 frames. ], batch size: 55, lr: 3.24e-02, grad_scale: 4.0 +2023-03-31 22:25:32,660 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-03-31 22:26:13,347 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.771e+02 9.409e+02 1.166e+03 1.453e+03 2.796e+03, threshold=2.333e+03, percent-clipped=9.0 +2023-03-31 22:26:17,058 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10317.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:26:31,778 INFO [train.py:903] (3/4) Epoch 2, batch 3500, loss[loss=0.361, simple_loss=0.3938, pruned_loss=0.1641, over 19767.00 frames. ], tot_loss[loss=0.362, simple_loss=0.3968, pruned_loss=0.1636, over 3810362.15 frames. ], batch size: 54, lr: 3.24e-02, grad_scale: 4.0 +2023-03-31 22:26:47,010 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10342.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:27:28,357 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4593, 2.8194, 2.1237, 2.4962, 2.4444, 1.6564, 0.7552, 2.2385], + device='cuda:3'), covar=tensor([0.0340, 0.0326, 0.0316, 0.0333, 0.0513, 0.0710, 0.0864, 0.0665], + device='cuda:3'), in_proj_covar=tensor([0.0206, 0.0216, 0.0212, 0.0241, 0.0287, 0.0256, 0.0251, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:27:33,749 INFO [train.py:903] (3/4) Epoch 2, batch 3550, loss[loss=0.3482, simple_loss=0.3884, pruned_loss=0.154, over 19664.00 frames. ], tot_loss[loss=0.3613, simple_loss=0.3963, pruned_loss=0.1631, over 3816758.41 frames. ], batch size: 53, lr: 3.23e-02, grad_scale: 4.0 +2023-03-31 22:27:49,478 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10391.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:28:04,139 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10401.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:28:20,156 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.354e+02 7.776e+02 1.013e+03 1.369e+03 3.978e+03, threshold=2.027e+03, percent-clipped=2.0 +2023-03-31 22:28:21,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10416.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:28:30,605 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10424.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 22:28:35,060 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10427.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:28:37,136 INFO [train.py:903] (3/4) Epoch 2, batch 3600, loss[loss=0.2907, simple_loss=0.3413, pruned_loss=0.1201, over 19768.00 frames. ], tot_loss[loss=0.3621, simple_loss=0.397, pruned_loss=0.1636, over 3813655.05 frames. ], batch size: 46, lr: 3.22e-02, grad_scale: 8.0 +2023-03-31 22:29:32,600 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10472.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:29:37,372 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10476.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:29:41,349 INFO [train.py:903] (3/4) Epoch 2, batch 3650, loss[loss=0.3323, simple_loss=0.3679, pruned_loss=0.1483, over 19786.00 frames. ], tot_loss[loss=0.3612, simple_loss=0.3968, pruned_loss=0.1628, over 3812569.93 frames. ], batch size: 48, lr: 3.22e-02, grad_scale: 8.0 +2023-03-31 22:29:59,891 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5040, 1.4878, 1.1027, 1.7855, 1.5035, 1.3979, 1.4497, 1.3997], + device='cuda:3'), covar=tensor([0.0992, 0.1875, 0.1555, 0.0920, 0.1337, 0.0760, 0.1115, 0.0871], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0386, 0.0291, 0.0259, 0.0320, 0.0263, 0.0286, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:30:09,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10501.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:30:26,011 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.209e+02 8.454e+02 1.072e+03 1.396e+03 2.688e+03, threshold=2.143e+03, percent-clipped=6.0 +2023-03-31 22:30:27,514 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10516.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:30:45,121 INFO [train.py:903] (3/4) Epoch 2, batch 3700, loss[loss=0.2929, simple_loss=0.3353, pruned_loss=0.1252, over 19345.00 frames. ], tot_loss[loss=0.3625, simple_loss=0.398, pruned_loss=0.1635, over 3811406.93 frames. ], batch size: 44, lr: 3.21e-02, grad_scale: 8.0 +2023-03-31 22:30:50,130 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1177, 3.7934, 2.0520, 3.4606, 1.0570, 3.3964, 3.2698, 3.6289], + device='cuda:3'), covar=tensor([0.0779, 0.1145, 0.2237, 0.0722, 0.3618, 0.1078, 0.0814, 0.0743], + device='cuda:3'), in_proj_covar=tensor([0.0288, 0.0284, 0.0312, 0.0254, 0.0324, 0.0273, 0.0214, 0.0245], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:30:57,292 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10539.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 22:31:47,571 INFO [train.py:903] (3/4) Epoch 2, batch 3750, loss[loss=0.383, simple_loss=0.4243, pruned_loss=0.1709, over 19510.00 frames. ], tot_loss[loss=0.3608, simple_loss=0.3968, pruned_loss=0.1624, over 3815087.21 frames. ], batch size: 64, lr: 3.20e-02, grad_scale: 8.0 +2023-03-31 22:32:33,530 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.274e+02 8.986e+02 1.057e+03 1.326e+03 2.585e+03, threshold=2.114e+03, percent-clipped=3.0 +2023-03-31 22:32:50,180 INFO [train.py:903] (3/4) Epoch 2, batch 3800, loss[loss=0.4662, simple_loss=0.4735, pruned_loss=0.2295, over 19608.00 frames. ], tot_loss[loss=0.36, simple_loss=0.3961, pruned_loss=0.162, over 3825111.73 frames. ], batch size: 57, lr: 3.20e-02, grad_scale: 8.0 +2023-03-31 22:33:23,964 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-03-31 22:33:51,759 INFO [train.py:903] (3/4) Epoch 2, batch 3850, loss[loss=0.404, simple_loss=0.4304, pruned_loss=0.1888, over 19792.00 frames. ], tot_loss[loss=0.359, simple_loss=0.3954, pruned_loss=0.1613, over 3819871.52 frames. ], batch size: 56, lr: 3.19e-02, grad_scale: 8.0 +2023-03-31 22:34:15,325 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-31 22:34:36,685 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5425, 4.0974, 2.3319, 3.6499, 1.2761, 3.9159, 3.7851, 3.8031], + device='cuda:3'), covar=tensor([0.0610, 0.1194, 0.2108, 0.0691, 0.3438, 0.0930, 0.0536, 0.0648], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0275, 0.0305, 0.0249, 0.0316, 0.0270, 0.0208, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:34:37,535 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.106e+02 8.604e+02 1.077e+03 1.442e+03 2.985e+03, threshold=2.155e+03, percent-clipped=6.0 +2023-03-31 22:34:56,858 INFO [train.py:903] (3/4) Epoch 2, batch 3900, loss[loss=0.3817, simple_loss=0.4192, pruned_loss=0.1721, over 19500.00 frames. ], tot_loss[loss=0.3566, simple_loss=0.3935, pruned_loss=0.1598, over 3834178.56 frames. ], batch size: 64, lr: 3.19e-02, grad_scale: 8.0 +2023-03-31 22:35:37,427 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10762.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:35:49,464 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10771.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:35:50,889 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10772.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:35:58,743 INFO [train.py:903] (3/4) Epoch 2, batch 3950, loss[loss=0.3157, simple_loss=0.3657, pruned_loss=0.1329, over 19726.00 frames. ], tot_loss[loss=0.357, simple_loss=0.3938, pruned_loss=0.1601, over 3833975.23 frames. ], batch size: 51, lr: 3.18e-02, grad_scale: 8.0 +2023-03-31 22:36:00,418 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3304, 1.2323, 1.7800, 1.3850, 2.5996, 2.5233, 2.8330, 1.1995], + device='cuda:3'), covar=tensor([0.1423, 0.2120, 0.1384, 0.1533, 0.0855, 0.0824, 0.0884, 0.1886], + device='cuda:3'), in_proj_covar=tensor([0.0351, 0.0378, 0.0352, 0.0350, 0.0410, 0.0331, 0.0483, 0.0368], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:36:04,589 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-03-31 22:36:15,286 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10792.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:36:18,966 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10795.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:36:21,148 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10797.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:36:44,827 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.316e+02 7.972e+02 1.008e+03 1.220e+03 2.629e+03, threshold=2.016e+03, percent-clipped=1.0 +2023-03-31 22:36:46,287 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=10816.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:36:51,386 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10820.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 22:37:01,758 INFO [train.py:903] (3/4) Epoch 2, batch 4000, loss[loss=0.3045, simple_loss=0.3463, pruned_loss=0.1313, over 19750.00 frames. ], tot_loss[loss=0.3564, simple_loss=0.3938, pruned_loss=0.1595, over 3808843.50 frames. ], batch size: 45, lr: 3.17e-02, grad_scale: 8.0 +2023-03-31 22:37:49,238 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-03-31 22:38:04,671 INFO [train.py:903] (3/4) Epoch 2, batch 4050, loss[loss=0.3065, simple_loss=0.3539, pruned_loss=0.1295, over 19757.00 frames. ], tot_loss[loss=0.3564, simple_loss=0.3936, pruned_loss=0.1596, over 3805946.52 frames. ], batch size: 47, lr: 3.17e-02, grad_scale: 8.0 +2023-03-31 22:38:15,786 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10886.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:38:35,390 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10903.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:38:49,943 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.194e+02 9.186e+02 1.101e+03 1.338e+03 4.215e+03, threshold=2.201e+03, percent-clipped=7.0 +2023-03-31 22:39:10,595 INFO [train.py:903] (3/4) Epoch 2, batch 4100, loss[loss=0.3647, simple_loss=0.4117, pruned_loss=0.1588, over 19549.00 frames. ], tot_loss[loss=0.355, simple_loss=0.3926, pruned_loss=0.1587, over 3806827.38 frames. ], batch size: 56, lr: 3.16e-02, grad_scale: 8.0 +2023-03-31 22:39:13,297 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10931.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:39:45,322 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-03-31 22:40:13,114 INFO [train.py:903] (3/4) Epoch 2, batch 4150, loss[loss=0.3907, simple_loss=0.4269, pruned_loss=0.1772, over 19594.00 frames. ], tot_loss[loss=0.3555, simple_loss=0.3932, pruned_loss=0.1589, over 3806801.07 frames. ], batch size: 57, lr: 3.16e-02, grad_scale: 8.0 +2023-03-31 22:40:59,345 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.653e+02 7.731e+02 1.007e+03 1.258e+03 2.097e+03, threshold=2.015e+03, percent-clipped=0.0 +2023-03-31 22:41:15,492 INFO [train.py:903] (3/4) Epoch 2, batch 4200, loss[loss=0.3795, simple_loss=0.4168, pruned_loss=0.1711, over 19558.00 frames. ], tot_loss[loss=0.3547, simple_loss=0.3924, pruned_loss=0.1585, over 3810388.26 frames. ], batch size: 61, lr: 3.15e-02, grad_scale: 8.0 +2023-03-31 22:41:18,936 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-03-31 22:42:18,014 INFO [train.py:903] (3/4) Epoch 2, batch 4250, loss[loss=0.4823, simple_loss=0.4785, pruned_loss=0.243, over 18155.00 frames. ], tot_loss[loss=0.3555, simple_loss=0.3927, pruned_loss=0.1591, over 3828809.69 frames. ], batch size: 83, lr: 3.14e-02, grad_scale: 8.0 +2023-03-31 22:42:35,133 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-03-31 22:42:45,446 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-03-31 22:42:52,703 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11106.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:43:03,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.611e+02 8.485e+02 1.107e+03 1.406e+03 3.284e+03, threshold=2.214e+03, percent-clipped=7.0 +2023-03-31 22:43:21,881 INFO [train.py:903] (3/4) Epoch 2, batch 4300, loss[loss=0.3368, simple_loss=0.3809, pruned_loss=0.1464, over 19588.00 frames. ], tot_loss[loss=0.3539, simple_loss=0.3921, pruned_loss=0.1578, over 3828420.88 frames. ], batch size: 52, lr: 3.14e-02, grad_scale: 8.0 +2023-03-31 22:43:30,408 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11136.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:43:37,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11142.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:44:09,048 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11167.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:44:16,377 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-03-31 22:44:23,252 INFO [train.py:903] (3/4) Epoch 2, batch 4350, loss[loss=0.338, simple_loss=0.3846, pruned_loss=0.1457, over 17236.00 frames. ], tot_loss[loss=0.3543, simple_loss=0.3929, pruned_loss=0.1579, over 3818565.40 frames. ], batch size: 101, lr: 3.13e-02, grad_scale: 8.0 +2023-03-31 22:44:32,898 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11187.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:44:39,318 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11192.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:45:04,764 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11212.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:45:09,090 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.029e+02 7.841e+02 9.559e+02 1.160e+03 2.939e+03, threshold=1.912e+03, percent-clipped=2.0 +2023-03-31 22:45:15,643 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11221.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:45:24,374 INFO [train.py:903] (3/4) Epoch 2, batch 4400, loss[loss=0.3465, simple_loss=0.3791, pruned_loss=0.1569, over 19831.00 frames. ], tot_loss[loss=0.3529, simple_loss=0.3916, pruned_loss=0.1571, over 3825176.46 frames. ], batch size: 52, lr: 3.13e-02, grad_scale: 8.0 +2023-03-31 22:45:46,922 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-31 22:45:48,415 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11247.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:45:50,514 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-03-31 22:45:54,212 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11251.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:46:00,706 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-03-31 22:46:04,600 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7038, 1.3937, 1.6133, 2.1751, 3.2659, 1.3287, 2.0781, 3.3400], + device='cuda:3'), covar=tensor([0.0258, 0.2119, 0.2100, 0.1306, 0.0357, 0.1933, 0.1044, 0.0319], + device='cuda:3'), in_proj_covar=tensor([0.0224, 0.0294, 0.0279, 0.0274, 0.0252, 0.0320, 0.0258, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:46:27,496 INFO [train.py:903] (3/4) Epoch 2, batch 4450, loss[loss=0.3118, simple_loss=0.3547, pruned_loss=0.1345, over 19355.00 frames. ], tot_loss[loss=0.3529, simple_loss=0.3915, pruned_loss=0.1572, over 3828704.77 frames. ], batch size: 47, lr: 3.12e-02, grad_scale: 8.0 +2023-03-31 22:46:31,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6463, 1.1375, 1.0512, 1.7204, 1.1879, 1.7166, 1.7134, 1.5250], + device='cuda:3'), covar=tensor([0.0816, 0.1419, 0.1513, 0.1126, 0.1442, 0.0976, 0.1103, 0.0857], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0305, 0.0295, 0.0330, 0.0353, 0.0267, 0.0318, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0005, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 22:47:14,190 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.163e+02 8.820e+02 1.081e+03 1.372e+03 2.333e+03, threshold=2.162e+03, percent-clipped=5.0 +2023-03-31 22:47:18,411 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.74 vs. limit=5.0 +2023-03-31 22:47:31,605 INFO [train.py:903] (3/4) Epoch 2, batch 4500, loss[loss=0.3509, simple_loss=0.3829, pruned_loss=0.1594, over 19590.00 frames. ], tot_loss[loss=0.3525, simple_loss=0.3909, pruned_loss=0.157, over 3827258.33 frames. ], batch size: 52, lr: 3.12e-02, grad_scale: 8.0 +2023-03-31 22:48:12,963 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11362.0, num_to_drop=1, layers_to_drop={1} +2023-03-31 22:48:26,005 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7782, 1.2039, 1.1351, 1.6956, 1.2881, 1.5522, 1.6593, 1.6043], + device='cuda:3'), covar=tensor([0.0775, 0.1444, 0.1603, 0.1119, 0.1497, 0.1212, 0.1387, 0.0835], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0308, 0.0297, 0.0328, 0.0354, 0.0270, 0.0324, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-03-31 22:48:26,046 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1117, 1.8001, 1.4451, 1.4938, 1.6324, 0.8961, 0.6122, 1.6010], + device='cuda:3'), covar=tensor([0.1086, 0.0647, 0.1190, 0.0602, 0.0650, 0.1593, 0.1189, 0.0566], + device='cuda:3'), in_proj_covar=tensor([0.0291, 0.0203, 0.0293, 0.0253, 0.0201, 0.0297, 0.0271, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 22:48:34,553 INFO [train.py:903] (3/4) Epoch 2, batch 4550, loss[loss=0.403, simple_loss=0.4344, pruned_loss=0.1858, over 19565.00 frames. ], tot_loss[loss=0.3534, simple_loss=0.3916, pruned_loss=0.1576, over 3832077.92 frames. ], batch size: 61, lr: 3.11e-02, grad_scale: 8.0 +2023-03-31 22:48:45,435 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-03-31 22:49:08,229 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-03-31 22:49:21,598 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.316e+02 7.788e+02 1.003e+03 1.220e+03 2.356e+03, threshold=2.005e+03, percent-clipped=1.0 +2023-03-31 22:49:37,109 INFO [train.py:903] (3/4) Epoch 2, batch 4600, loss[loss=0.3514, simple_loss=0.3937, pruned_loss=0.1546, over 19096.00 frames. ], tot_loss[loss=0.3527, simple_loss=0.3909, pruned_loss=0.1573, over 3830515.44 frames. ], batch size: 69, lr: 3.10e-02, grad_scale: 8.0 +2023-03-31 22:49:49,978 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3982, 2.2245, 1.7306, 1.9249, 1.5139, 1.6211, 0.3532, 1.2246], + device='cuda:3'), covar=tensor([0.0371, 0.0298, 0.0242, 0.0293, 0.0690, 0.0478, 0.0749, 0.0578], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0222, 0.0219, 0.0247, 0.0302, 0.0261, 0.0252, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:50:37,638 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11477.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:50:39,641 INFO [train.py:903] (3/4) Epoch 2, batch 4650, loss[loss=0.3124, simple_loss=0.3624, pruned_loss=0.1312, over 19859.00 frames. ], tot_loss[loss=0.3522, simple_loss=0.3905, pruned_loss=0.157, over 3829129.81 frames. ], batch size: 52, lr: 3.10e-02, grad_scale: 8.0 +2023-03-31 22:50:45,018 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8733, 1.7942, 1.7592, 2.6509, 1.6898, 2.5354, 1.9965, 1.6860], + device='cuda:3'), covar=tensor([0.0890, 0.0743, 0.0442, 0.0366, 0.0873, 0.0246, 0.0922, 0.0763], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0325, 0.0343, 0.0450, 0.0411, 0.0244, 0.0451, 0.0357], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0001, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:50:56,512 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-31 22:50:59,956 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-03-31 22:51:09,665 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11502.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:51:10,458 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-03-31 22:51:15,630 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11507.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:51:20,436 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5088, 2.2381, 1.5639, 1.6844, 1.8711, 0.9379, 1.1193, 1.7497], + device='cuda:3'), covar=tensor([0.1206, 0.0473, 0.1394, 0.0690, 0.0753, 0.1505, 0.1031, 0.0668], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0199, 0.0300, 0.0247, 0.0203, 0.0300, 0.0267, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 22:51:26,521 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.039e+02 8.212e+02 1.126e+03 1.409e+03 2.689e+03, threshold=2.252e+03, percent-clipped=6.0 +2023-03-31 22:51:42,941 INFO [train.py:903] (3/4) Epoch 2, batch 4700, loss[loss=0.3231, simple_loss=0.3765, pruned_loss=0.1349, over 19661.00 frames. ], tot_loss[loss=0.3515, simple_loss=0.3899, pruned_loss=0.1565, over 3825855.99 frames. ], batch size: 58, lr: 3.09e-02, grad_scale: 8.0 +2023-03-31 22:51:47,991 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11532.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:51:49,465 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-31 22:51:52,287 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11536.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:52:05,035 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-03-31 22:52:08,737 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11550.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:52:15,832 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8082, 3.5834, 2.1672, 2.6621, 2.8183, 1.4832, 1.4819, 1.5494], + device='cuda:3'), covar=tensor([0.1519, 0.0388, 0.1315, 0.0604, 0.0868, 0.1518, 0.1187, 0.1188], + device='cuda:3'), in_proj_covar=tensor([0.0286, 0.0199, 0.0298, 0.0247, 0.0200, 0.0296, 0.0266, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 22:52:46,172 INFO [train.py:903] (3/4) Epoch 2, batch 4750, loss[loss=0.3551, simple_loss=0.3989, pruned_loss=0.1557, over 18791.00 frames. ], tot_loss[loss=0.3516, simple_loss=0.3903, pruned_loss=0.1565, over 3827636.00 frames. ], batch size: 74, lr: 3.09e-02, grad_scale: 8.0 +2023-03-31 22:53:32,504 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.603e+02 7.433e+02 1.012e+03 1.332e+03 3.283e+03, threshold=2.025e+03, percent-clipped=2.0 +2023-03-31 22:53:35,213 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11618.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:53:47,427 INFO [train.py:903] (3/4) Epoch 2, batch 4800, loss[loss=0.3172, simple_loss=0.3597, pruned_loss=0.1373, over 19612.00 frames. ], tot_loss[loss=0.3511, simple_loss=0.3895, pruned_loss=0.1564, over 3811629.35 frames. ], batch size: 50, lr: 3.08e-02, grad_scale: 8.0 +2023-03-31 22:53:47,662 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0224, 5.3381, 2.8151, 4.7132, 1.5663, 5.3466, 5.2320, 5.5082], + device='cuda:3'), covar=tensor([0.0402, 0.0894, 0.2133, 0.0626, 0.3355, 0.0923, 0.0551, 0.0531], + device='cuda:3'), in_proj_covar=tensor([0.0291, 0.0287, 0.0315, 0.0260, 0.0326, 0.0284, 0.0223, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:54:04,672 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11643.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 22:54:13,893 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9767, 1.9051, 2.0184, 2.4649, 2.3062, 2.8557, 3.1191, 2.8991], + device='cuda:3'), covar=tensor([0.0620, 0.1208, 0.1370, 0.1371, 0.1367, 0.0846, 0.1164, 0.0633], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0307, 0.0291, 0.0322, 0.0347, 0.0263, 0.0320, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0005, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-03-31 22:54:15,091 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11651.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:54:19,410 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11653.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:54:49,932 INFO [train.py:903] (3/4) Epoch 2, batch 4850, loss[loss=0.3233, simple_loss=0.3539, pruned_loss=0.1463, over 19747.00 frames. ], tot_loss[loss=0.3491, simple_loss=0.3885, pruned_loss=0.1549, over 3830495.47 frames. ], batch size: 46, lr: 3.08e-02, grad_scale: 8.0 +2023-03-31 22:55:14,991 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-03-31 22:55:34,591 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-03-31 22:55:36,946 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.407e+02 7.585e+02 1.008e+03 1.288e+03 2.592e+03, threshold=2.016e+03, percent-clipped=4.0 +2023-03-31 22:55:38,504 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11717.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:55:40,628 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-03-31 22:55:41,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-03-31 22:55:51,135 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-03-31 22:55:53,217 INFO [train.py:903] (3/4) Epoch 2, batch 4900, loss[loss=0.3603, simple_loss=0.4086, pruned_loss=0.156, over 19694.00 frames. ], tot_loss[loss=0.3505, simple_loss=0.3893, pruned_loss=0.1558, over 3807663.09 frames. ], batch size: 59, lr: 3.07e-02, grad_scale: 8.0 +2023-03-31 22:56:04,446 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2424, 3.7424, 2.1802, 3.5018, 1.2645, 3.7189, 3.4459, 3.6552], + device='cuda:3'), covar=tensor([0.0625, 0.1382, 0.1997, 0.0669, 0.3418, 0.0801, 0.0562, 0.0690], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0298, 0.0316, 0.0266, 0.0332, 0.0286, 0.0225, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 22:56:13,250 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-03-31 22:56:55,835 INFO [train.py:903] (3/4) Epoch 2, batch 4950, loss[loss=0.3595, simple_loss=0.3993, pruned_loss=0.1599, over 19519.00 frames. ], tot_loss[loss=0.3506, simple_loss=0.3894, pruned_loss=0.1558, over 3815388.13 frames. ], batch size: 54, lr: 3.06e-02, grad_scale: 8.0 +2023-03-31 22:57:12,057 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-03-31 22:57:37,549 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-03-31 22:57:41,842 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.086e+02 9.257e+02 1.136e+03 1.412e+03 3.441e+03, threshold=2.272e+03, percent-clipped=4.0 +2023-03-31 22:57:52,801 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-03-31 22:57:57,810 INFO [train.py:903] (3/4) Epoch 2, batch 5000, loss[loss=0.3843, simple_loss=0.4141, pruned_loss=0.1772, over 19668.00 frames. ], tot_loss[loss=0.3505, simple_loss=0.3895, pruned_loss=0.1558, over 3826466.35 frames. ], batch size: 58, lr: 3.06e-02, grad_scale: 8.0 +2023-03-31 22:58:04,651 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11834.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:58:07,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-03-31 22:58:08,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4073, 1.2218, 1.5389, 1.8113, 3.0873, 1.1948, 2.0294, 2.9430], + device='cuda:3'), covar=tensor([0.0340, 0.2441, 0.2215, 0.1503, 0.0374, 0.2059, 0.1071, 0.0461], + device='cuda:3'), in_proj_covar=tensor([0.0229, 0.0298, 0.0281, 0.0276, 0.0258, 0.0313, 0.0257, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 22:58:16,655 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-03-31 22:58:21,712 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5594, 1.1086, 1.3739, 1.0580, 2.6118, 3.2697, 3.2690, 3.5300], + device='cuda:3'), covar=tensor([0.1514, 0.2905, 0.2849, 0.2160, 0.0506, 0.0138, 0.0227, 0.0124], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0278, 0.0315, 0.0276, 0.0195, 0.0105, 0.0192, 0.0109], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-03-31 22:58:42,527 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11864.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:59:00,551 INFO [train.py:903] (3/4) Epoch 2, batch 5050, loss[loss=0.3276, simple_loss=0.3788, pruned_loss=0.1382, over 19531.00 frames. ], tot_loss[loss=0.3522, simple_loss=0.3909, pruned_loss=0.1568, over 3828414.67 frames. ], batch size: 54, lr: 3.05e-02, grad_scale: 8.0 +2023-03-31 22:59:19,020 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11894.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:59:36,721 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11907.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 22:59:38,652 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-03-31 22:59:48,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.479e+02 8.836e+02 1.119e+03 1.460e+03 3.605e+03, threshold=2.237e+03, percent-clipped=7.0 +2023-03-31 23:00:03,346 INFO [train.py:903] (3/4) Epoch 2, batch 5100, loss[loss=0.3543, simple_loss=0.3989, pruned_loss=0.1549, over 17942.00 frames. ], tot_loss[loss=0.3529, simple_loss=0.3915, pruned_loss=0.1571, over 3825504.91 frames. ], batch size: 83, lr: 3.05e-02, grad_scale: 8.0 +2023-03-31 23:00:08,071 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11932.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:00:16,225 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11938.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:00:19,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-03-31 23:00:22,851 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-03-31 23:00:26,270 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-03-31 23:00:50,344 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11966.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:01:07,396 INFO [train.py:903] (3/4) Epoch 2, batch 5150, loss[loss=0.4319, simple_loss=0.4464, pruned_loss=0.2087, over 19506.00 frames. ], tot_loss[loss=0.3514, simple_loss=0.3902, pruned_loss=0.1563, over 3820570.76 frames. ], batch size: 64, lr: 3.04e-02, grad_scale: 8.0 +2023-03-31 23:01:20,920 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-03-31 23:01:29,902 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=11997.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:01:36,674 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12002.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:01:44,903 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12009.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:01:54,417 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.835e+02 8.254e+02 9.989e+02 1.287e+03 2.673e+03, threshold=1.998e+03, percent-clipped=4.0 +2023-03-31 23:01:56,795 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-03-31 23:02:10,377 INFO [train.py:903] (3/4) Epoch 2, batch 5200, loss[loss=0.3255, simple_loss=0.3588, pruned_loss=0.1461, over 19727.00 frames. ], tot_loss[loss=0.3517, simple_loss=0.3905, pruned_loss=0.1564, over 3820552.12 frames. ], batch size: 47, lr: 3.04e-02, grad_scale: 8.0 +2023-03-31 23:02:25,403 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-03-31 23:02:52,155 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12061.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:03:11,235 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-03-31 23:03:13,776 INFO [train.py:903] (3/4) Epoch 2, batch 5250, loss[loss=0.2782, simple_loss=0.3271, pruned_loss=0.1146, over 19761.00 frames. ], tot_loss[loss=0.3515, simple_loss=0.3902, pruned_loss=0.1563, over 3819706.90 frames. ], batch size: 47, lr: 3.03e-02, grad_scale: 8.0 +2023-03-31 23:03:32,695 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12094.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:03:55,623 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12112.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:03:59,697 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.115e+02 8.666e+02 1.057e+03 1.421e+03 4.195e+03, threshold=2.115e+03, percent-clipped=5.0 +2023-03-31 23:04:02,192 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7290, 1.3234, 1.2772, 1.8147, 1.5824, 1.5051, 1.3958, 1.6012], + device='cuda:3'), covar=tensor([0.0754, 0.1607, 0.1293, 0.0868, 0.1117, 0.0591, 0.0950, 0.0664], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0378, 0.0289, 0.0259, 0.0325, 0.0266, 0.0275, 0.0231], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:04:14,787 INFO [train.py:903] (3/4) Epoch 2, batch 5300, loss[loss=0.3498, simple_loss=0.3951, pruned_loss=0.1522, over 18770.00 frames. ], tot_loss[loss=0.3532, simple_loss=0.3915, pruned_loss=0.1575, over 3815446.59 frames. ], batch size: 74, lr: 3.03e-02, grad_scale: 8.0 +2023-03-31 23:04:35,591 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-03-31 23:05:13,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12176.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:05:15,938 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12178.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:05:16,931 INFO [train.py:903] (3/4) Epoch 2, batch 5350, loss[loss=0.3479, simple_loss=0.3951, pruned_loss=0.1503, over 19671.00 frames. ], tot_loss[loss=0.3521, simple_loss=0.3903, pruned_loss=0.1569, over 3832045.39 frames. ], batch size: 58, lr: 3.02e-02, grad_scale: 8.0 +2023-03-31 23:05:45,073 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1403, 1.1956, 1.8030, 1.3880, 2.8070, 2.7553, 2.8759, 1.3123], + device='cuda:3'), covar=tensor([0.1296, 0.2077, 0.1152, 0.1232, 0.0679, 0.0663, 0.0871, 0.1682], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0388, 0.0349, 0.0350, 0.0420, 0.0339, 0.0493, 0.0373], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:05:53,531 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-03-31 23:05:53,660 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12208.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:06:03,420 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.241e+02 9.122e+02 1.169e+03 1.506e+03 3.477e+03, threshold=2.338e+03, percent-clipped=13.0 +2023-03-31 23:06:20,773 INFO [train.py:903] (3/4) Epoch 2, batch 5400, loss[loss=0.3777, simple_loss=0.4111, pruned_loss=0.1721, over 18751.00 frames. ], tot_loss[loss=0.3519, simple_loss=0.3903, pruned_loss=0.1567, over 3824508.44 frames. ], batch size: 74, lr: 3.02e-02, grad_scale: 8.0 +2023-03-31 23:07:06,480 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12265.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:07:23,454 INFO [train.py:903] (3/4) Epoch 2, batch 5450, loss[loss=0.439, simple_loss=0.4497, pruned_loss=0.2141, over 17540.00 frames. ], tot_loss[loss=0.3507, simple_loss=0.3898, pruned_loss=0.1558, over 3821010.33 frames. ], batch size: 101, lr: 3.01e-02, grad_scale: 8.0 +2023-03-31 23:07:27,096 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12282.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:07:36,432 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12290.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:07:40,664 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12293.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:08:01,536 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12310.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:08:08,448 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.315e+02 7.539e+02 9.464e+02 1.137e+03 1.898e+03, threshold=1.893e+03, percent-clipped=0.0 +2023-03-31 23:08:17,866 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12323.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:08:23,997 INFO [train.py:903] (3/4) Epoch 2, batch 5500, loss[loss=0.369, simple_loss=0.4081, pruned_loss=0.1649, over 19533.00 frames. ], tot_loss[loss=0.3516, simple_loss=0.3908, pruned_loss=0.1562, over 3810210.99 frames. ], batch size: 64, lr: 3.01e-02, grad_scale: 8.0 +2023-03-31 23:08:47,643 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12346.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:08:49,882 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-03-31 23:09:14,235 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12368.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:09:28,529 INFO [train.py:903] (3/4) Epoch 2, batch 5550, loss[loss=0.3696, simple_loss=0.414, pruned_loss=0.1626, over 19374.00 frames. ], tot_loss[loss=0.3483, simple_loss=0.3885, pruned_loss=0.1541, over 3826612.04 frames. ], batch size: 66, lr: 3.00e-02, grad_scale: 8.0 +2023-03-31 23:09:36,189 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-03-31 23:09:46,133 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12393.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:09:50,671 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12397.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:10:14,308 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.827e+02 8.577e+02 1.018e+03 1.198e+03 2.956e+03, threshold=2.037e+03, percent-clipped=3.0 +2023-03-31 23:10:25,914 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-03-31 23:10:27,088 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12425.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:10:31,331 INFO [train.py:903] (3/4) Epoch 2, batch 5600, loss[loss=0.3203, simple_loss=0.3534, pruned_loss=0.1435, over 19498.00 frames. ], tot_loss[loss=0.3469, simple_loss=0.3876, pruned_loss=0.1531, over 3832137.14 frames. ], batch size: 49, lr: 3.00e-02, grad_scale: 8.0 +2023-03-31 23:10:35,306 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12432.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:10:41,880 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12438.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:11:06,511 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12457.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:11:10,906 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1645, 1.7732, 1.4514, 1.2075, 1.5955, 0.7977, 0.8404, 1.5643], + device='cuda:3'), covar=tensor([0.0702, 0.0429, 0.0817, 0.0507, 0.0459, 0.1106, 0.0737, 0.0422], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0210, 0.0301, 0.0249, 0.0209, 0.0297, 0.0271, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 23:11:12,054 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12461.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:11:16,301 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12464.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:11:33,650 INFO [train.py:903] (3/4) Epoch 2, batch 5650, loss[loss=0.3308, simple_loss=0.3835, pruned_loss=0.139, over 19675.00 frames. ], tot_loss[loss=0.3473, simple_loss=0.3876, pruned_loss=0.1535, over 3826696.32 frames. ], batch size: 55, lr: 2.99e-02, grad_scale: 8.0 +2023-03-31 23:12:19,594 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.393e+02 8.152e+02 1.040e+03 1.340e+03 2.595e+03, threshold=2.080e+03, percent-clipped=4.0 +2023-03-31 23:12:21,931 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-03-31 23:12:35,557 INFO [train.py:903] (3/4) Epoch 2, batch 5700, loss[loss=0.3048, simple_loss=0.3457, pruned_loss=0.132, over 19344.00 frames. ], tot_loss[loss=0.3477, simple_loss=0.3881, pruned_loss=0.1537, over 3839238.94 frames. ], batch size: 47, lr: 2.98e-02, grad_scale: 8.0 +2023-03-31 23:13:02,450 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12549.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:13:06,767 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0261, 4.8235, 5.7567, 5.5857, 1.9759, 5.1345, 4.8601, 5.1420], + device='cuda:3'), covar=tensor([0.0382, 0.0429, 0.0385, 0.0211, 0.2964, 0.0158, 0.0245, 0.0781], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0297, 0.0408, 0.0305, 0.0448, 0.0201, 0.0279, 0.0408], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 23:13:06,850 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12553.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:13:32,367 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12574.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:13:37,668 INFO [train.py:903] (3/4) Epoch 2, batch 5750, loss[loss=0.3765, simple_loss=0.4266, pruned_loss=0.1632, over 19663.00 frames. ], tot_loss[loss=0.3462, simple_loss=0.3872, pruned_loss=0.1526, over 3841287.04 frames. ], batch size: 58, lr: 2.98e-02, grad_scale: 8.0 +2023-03-31 23:13:39,172 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12579.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:13:42,220 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-03-31 23:13:50,271 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-03-31 23:13:56,874 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-03-31 23:14:10,379 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12604.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:14:23,858 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.880e+02 8.083e+02 9.516e+02 1.322e+03 3.330e+03, threshold=1.903e+03, percent-clipped=5.0 +2023-03-31 23:14:40,968 INFO [train.py:903] (3/4) Epoch 2, batch 5800, loss[loss=0.4082, simple_loss=0.4371, pruned_loss=0.1897, over 19539.00 frames. ], tot_loss[loss=0.3468, simple_loss=0.3876, pruned_loss=0.1529, over 3821426.59 frames. ], batch size: 56, lr: 2.97e-02, grad_scale: 8.0 +2023-03-31 23:15:10,932 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12653.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:15:42,714 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12678.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:15:43,517 INFO [train.py:903] (3/4) Epoch 2, batch 5850, loss[loss=0.2995, simple_loss=0.3399, pruned_loss=0.1295, over 19100.00 frames. ], tot_loss[loss=0.3475, simple_loss=0.388, pruned_loss=0.1536, over 3813977.74 frames. ], batch size: 42, lr: 2.97e-02, grad_scale: 8.0 +2023-03-31 23:15:44,004 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9726, 0.9533, 1.5998, 1.1061, 2.3548, 2.3073, 2.6035, 1.0032], + device='cuda:3'), covar=tensor([0.1681, 0.2332, 0.1317, 0.1637, 0.0920, 0.0977, 0.1018, 0.2008], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0396, 0.0363, 0.0363, 0.0438, 0.0349, 0.0510, 0.0382], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:15:47,153 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12681.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:16:18,364 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12706.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:16:30,351 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.479e+02 7.980e+02 9.827e+02 1.217e+03 2.781e+03, threshold=1.965e+03, percent-clipped=6.0 +2023-03-31 23:16:32,017 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12717.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:16:46,108 INFO [train.py:903] (3/4) Epoch 2, batch 5900, loss[loss=0.3428, simple_loss=0.3877, pruned_loss=0.149, over 19604.00 frames. ], tot_loss[loss=0.3478, simple_loss=0.3884, pruned_loss=0.1536, over 3821992.16 frames. ], batch size: 57, lr: 2.96e-02, grad_scale: 8.0 +2023-03-31 23:16:49,441 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-03-31 23:17:03,733 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12742.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:17:11,479 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-03-31 23:17:41,839 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.66 vs. limit=5.0 +2023-03-31 23:17:49,031 INFO [train.py:903] (3/4) Epoch 2, batch 5950, loss[loss=0.327, simple_loss=0.3829, pruned_loss=0.1356, over 19615.00 frames. ], tot_loss[loss=0.3465, simple_loss=0.3872, pruned_loss=0.1529, over 3823631.25 frames. ], batch size: 57, lr: 2.96e-02, grad_scale: 8.0 +2023-03-31 23:17:52,884 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12781.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:18:25,408 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=12808.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:18:26,826 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12809.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:18:34,471 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.856e+02 9.090e+02 1.139e+03 1.452e+03 3.383e+03, threshold=2.279e+03, percent-clipped=8.0 +2023-03-31 23:18:51,783 INFO [train.py:903] (3/4) Epoch 2, batch 6000, loss[loss=0.2916, simple_loss=0.3401, pruned_loss=0.1216, over 19461.00 frames. ], tot_loss[loss=0.3478, simple_loss=0.3876, pruned_loss=0.154, over 3809237.54 frames. ], batch size: 49, lr: 2.95e-02, grad_scale: 8.0 +2023-03-31 23:18:51,784 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 23:19:06,012 INFO [train.py:937] (3/4) Epoch 2, validation: loss=0.246, simple_loss=0.337, pruned_loss=0.07745, over 944034.00 frames. +2023-03-31 23:19:06,013 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18251MB +2023-03-31 23:19:13,368 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12834.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:20:08,252 INFO [train.py:903] (3/4) Epoch 2, batch 6050, loss[loss=0.3127, simple_loss=0.3696, pruned_loss=0.1279, over 19596.00 frames. ], tot_loss[loss=0.3476, simple_loss=0.387, pruned_loss=0.1541, over 3807929.15 frames. ], batch size: 52, lr: 2.95e-02, grad_scale: 4.0 +2023-03-31 23:20:52,510 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6465, 1.6517, 1.5097, 2.2826, 1.6284, 1.9499, 1.8855, 1.4222], + device='cuda:3'), covar=tensor([0.0813, 0.0649, 0.0444, 0.0333, 0.0691, 0.0272, 0.0832, 0.0724], + device='cuda:3'), in_proj_covar=tensor([0.0375, 0.0347, 0.0361, 0.0476, 0.0431, 0.0268, 0.0461, 0.0367], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:20:56,515 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.060e+02 7.840e+02 9.937e+02 1.323e+03 8.220e+03, threshold=1.987e+03, percent-clipped=9.0 +2023-03-31 23:21:03,491 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12923.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:21:10,256 INFO [train.py:903] (3/4) Epoch 2, batch 6100, loss[loss=0.2977, simple_loss=0.3485, pruned_loss=0.1235, over 19605.00 frames. ], tot_loss[loss=0.3464, simple_loss=0.3864, pruned_loss=0.1532, over 3816647.00 frames. ], batch size: 50, lr: 2.94e-02, grad_scale: 4.0 +2023-03-31 23:21:24,588 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-31 23:21:30,965 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6330, 1.5868, 1.6016, 2.2709, 1.6651, 1.9352, 1.8877, 1.5109], + device='cuda:3'), covar=tensor([0.0811, 0.0662, 0.0426, 0.0318, 0.0647, 0.0274, 0.0813, 0.0669], + device='cuda:3'), in_proj_covar=tensor([0.0377, 0.0348, 0.0362, 0.0476, 0.0431, 0.0269, 0.0463, 0.0364], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:22:11,868 INFO [train.py:903] (3/4) Epoch 2, batch 6150, loss[loss=0.3516, simple_loss=0.3796, pruned_loss=0.1619, over 19212.00 frames. ], tot_loss[loss=0.3498, simple_loss=0.3886, pruned_loss=0.1554, over 3811312.60 frames. ], batch size: 42, lr: 2.94e-02, grad_scale: 4.0 +2023-03-31 23:22:14,912 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-31 23:22:42,902 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-03-31 23:23:00,757 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.553e+02 7.749e+02 1.029e+03 1.287e+03 3.235e+03, threshold=2.059e+03, percent-clipped=7.0 +2023-03-31 23:23:13,338 INFO [train.py:903] (3/4) Epoch 2, batch 6200, loss[loss=0.3004, simple_loss=0.3463, pruned_loss=0.1273, over 18551.00 frames. ], tot_loss[loss=0.3475, simple_loss=0.3868, pruned_loss=0.1541, over 3811288.81 frames. ], batch size: 41, lr: 2.93e-02, grad_scale: 4.0 +2023-03-31 23:24:15,396 INFO [train.py:903] (3/4) Epoch 2, batch 6250, loss[loss=0.4301, simple_loss=0.4421, pruned_loss=0.209, over 13262.00 frames. ], tot_loss[loss=0.3485, simple_loss=0.3879, pruned_loss=0.1545, over 3807984.32 frames. ], batch size: 136, lr: 2.93e-02, grad_scale: 4.0 +2023-03-31 23:24:47,227 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-03-31 23:25:04,357 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.358e+02 8.552e+02 1.023e+03 1.333e+03 3.705e+03, threshold=2.046e+03, percent-clipped=2.0 +2023-03-31 23:25:13,052 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=13125.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:25:17,707 INFO [train.py:903] (3/4) Epoch 2, batch 6300, loss[loss=0.2935, simple_loss=0.3458, pruned_loss=0.1206, over 19735.00 frames. ], tot_loss[loss=0.3486, simple_loss=0.3885, pruned_loss=0.1543, over 3813204.98 frames. ], batch size: 51, lr: 2.92e-02, grad_scale: 4.0 +2023-03-31 23:26:19,879 INFO [train.py:903] (3/4) Epoch 2, batch 6350, loss[loss=0.3244, simple_loss=0.3758, pruned_loss=0.1366, over 19664.00 frames. ], tot_loss[loss=0.3474, simple_loss=0.388, pruned_loss=0.1534, over 3824768.46 frames. ], batch size: 53, lr: 2.92e-02, grad_scale: 4.0 +2023-03-31 23:26:20,353 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=13179.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:26:49,668 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=13204.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:27:06,691 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.673e+02 8.539e+02 1.071e+03 1.407e+03 4.202e+03, threshold=2.141e+03, percent-clipped=6.0 +2023-03-31 23:27:19,362 INFO [train.py:903] (3/4) Epoch 2, batch 6400, loss[loss=0.3714, simple_loss=0.4077, pruned_loss=0.1675, over 19531.00 frames. ], tot_loss[loss=0.3484, simple_loss=0.3888, pruned_loss=0.154, over 3810161.37 frames. ], batch size: 54, lr: 2.92e-02, grad_scale: 8.0 +2023-03-31 23:27:33,753 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=13240.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:28:22,169 INFO [train.py:903] (3/4) Epoch 2, batch 6450, loss[loss=0.3334, simple_loss=0.3838, pruned_loss=0.1415, over 17975.00 frames. ], tot_loss[loss=0.3481, simple_loss=0.3888, pruned_loss=0.1537, over 3808236.33 frames. ], batch size: 83, lr: 2.91e-02, grad_scale: 8.0 +2023-03-31 23:29:09,608 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-03-31 23:29:10,633 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.859e+02 7.874e+02 9.907e+02 1.203e+03 2.411e+03, threshold=1.981e+03, percent-clipped=4.0 +2023-03-31 23:29:24,102 INFO [train.py:903] (3/4) Epoch 2, batch 6500, loss[loss=0.2989, simple_loss=0.3427, pruned_loss=0.1276, over 19758.00 frames. ], tot_loss[loss=0.3469, simple_loss=0.3878, pruned_loss=0.153, over 3807694.78 frames. ], batch size: 46, lr: 2.91e-02, grad_scale: 8.0 +2023-03-31 23:29:30,885 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-03-31 23:30:26,693 INFO [train.py:903] (3/4) Epoch 2, batch 6550, loss[loss=0.3412, simple_loss=0.3886, pruned_loss=0.1468, over 19776.00 frames. ], tot_loss[loss=0.3459, simple_loss=0.387, pruned_loss=0.1524, over 3820393.86 frames. ], batch size: 56, lr: 2.90e-02, grad_scale: 8.0 +2023-03-31 23:31:14,491 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.467e+02 7.924e+02 9.507e+02 1.218e+03 2.525e+03, threshold=1.901e+03, percent-clipped=3.0 +2023-03-31 23:31:27,083 INFO [train.py:903] (3/4) Epoch 2, batch 6600, loss[loss=0.384, simple_loss=0.4257, pruned_loss=0.1711, over 18136.00 frames. ], tot_loss[loss=0.3454, simple_loss=0.3866, pruned_loss=0.1521, over 3821605.32 frames. ], batch size: 83, lr: 2.90e-02, grad_scale: 8.0 +2023-03-31 23:32:29,095 INFO [train.py:903] (3/4) Epoch 2, batch 6650, loss[loss=0.4002, simple_loss=0.434, pruned_loss=0.1832, over 19621.00 frames. ], tot_loss[loss=0.3467, simple_loss=0.3868, pruned_loss=0.1533, over 3822582.79 frames. ], batch size: 57, lr: 2.89e-02, grad_scale: 8.0 +2023-03-31 23:32:51,548 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=13496.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:33:17,438 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.558e+02 9.096e+02 1.181e+03 1.471e+03 3.411e+03, threshold=2.361e+03, percent-clipped=10.0 +2023-03-31 23:33:21,194 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=13521.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:33:29,651 INFO [train.py:903] (3/4) Epoch 2, batch 6700, loss[loss=0.4083, simple_loss=0.4347, pruned_loss=0.191, over 18153.00 frames. ], tot_loss[loss=0.3469, simple_loss=0.3866, pruned_loss=0.1536, over 3805182.59 frames. ], batch size: 83, lr: 2.89e-02, grad_scale: 8.0 +2023-03-31 23:34:27,762 INFO [train.py:903] (3/4) Epoch 2, batch 6750, loss[loss=0.3641, simple_loss=0.3846, pruned_loss=0.1718, over 19404.00 frames. ], tot_loss[loss=0.3483, simple_loss=0.3878, pruned_loss=0.1544, over 3796984.28 frames. ], batch size: 48, lr: 2.88e-02, grad_scale: 8.0 +2023-03-31 23:35:12,671 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.052e+02 8.174e+02 9.996e+02 1.293e+03 2.664e+03, threshold=1.999e+03, percent-clipped=1.0 +2023-03-31 23:35:25,230 INFO [train.py:903] (3/4) Epoch 2, batch 6800, loss[loss=0.3364, simple_loss=0.3907, pruned_loss=0.1411, over 19531.00 frames. ], tot_loss[loss=0.3469, simple_loss=0.3868, pruned_loss=0.1535, over 3799474.06 frames. ], batch size: 54, lr: 2.88e-02, grad_scale: 8.0 +2023-03-31 23:36:09,779 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-03-31 23:36:10,968 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-03-31 23:36:13,728 INFO [train.py:903] (3/4) Epoch 3, batch 0, loss[loss=0.389, simple_loss=0.4155, pruned_loss=0.1812, over 18205.00 frames. ], tot_loss[loss=0.389, simple_loss=0.4155, pruned_loss=0.1812, over 18205.00 frames. ], batch size: 83, lr: 2.73e-02, grad_scale: 8.0 +2023-03-31 23:36:13,728 INFO [train.py:928] (3/4) Computing validation loss +2023-03-31 23:36:24,494 INFO [train.py:937] (3/4) Epoch 3, validation: loss=0.241, simple_loss=0.3346, pruned_loss=0.07374, over 944034.00 frames. +2023-03-31 23:36:24,495 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18419MB +2023-03-31 23:36:37,412 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-03-31 23:37:25,523 INFO [train.py:903] (3/4) Epoch 3, batch 50, loss[loss=0.3678, simple_loss=0.4036, pruned_loss=0.166, over 19426.00 frames. ], tot_loss[loss=0.3432, simple_loss=0.3852, pruned_loss=0.1506, over 866559.39 frames. ], batch size: 70, lr: 2.73e-02, grad_scale: 8.0 +2023-03-31 23:37:38,309 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.119e+02 7.787e+02 9.326e+02 1.115e+03 3.182e+03, threshold=1.865e+03, percent-clipped=5.0 +2023-03-31 23:37:58,965 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-03-31 23:38:24,031 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=13755.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:38:25,865 INFO [train.py:903] (3/4) Epoch 3, batch 100, loss[loss=0.3872, simple_loss=0.425, pruned_loss=0.1748, over 19596.00 frames. ], tot_loss[loss=0.339, simple_loss=0.3813, pruned_loss=0.1484, over 1505485.58 frames. ], batch size: 57, lr: 2.72e-02, grad_scale: 8.0 +2023-03-31 23:38:33,262 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3416, 1.2456, 2.3287, 1.7299, 3.1306, 3.0798, 3.4312, 1.8212], + device='cuda:3'), covar=tensor([0.1365, 0.2091, 0.1108, 0.1177, 0.0930, 0.0843, 0.1097, 0.1837], + device='cuda:3'), in_proj_covar=tensor([0.0380, 0.0420, 0.0380, 0.0377, 0.0447, 0.0365, 0.0530, 0.0391], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:38:35,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-03-31 23:39:21,045 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7879, 1.7513, 2.0621, 2.3621, 1.9993, 2.1529, 2.1568, 2.7246], + device='cuda:3'), covar=tensor([0.0601, 0.2034, 0.1300, 0.0939, 0.1414, 0.0554, 0.1012, 0.0547], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0383, 0.0292, 0.0259, 0.0324, 0.0271, 0.0279, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:39:27,419 INFO [train.py:903] (3/4) Epoch 3, batch 150, loss[loss=0.3367, simple_loss=0.3952, pruned_loss=0.1391, over 19669.00 frames. ], tot_loss[loss=0.3405, simple_loss=0.3821, pruned_loss=0.1495, over 2022685.54 frames. ], batch size: 58, lr: 2.72e-02, grad_scale: 8.0 +2023-03-31 23:39:40,058 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.782e+02 7.521e+02 1.009e+03 1.351e+03 3.530e+03, threshold=2.018e+03, percent-clipped=10.0 +2023-03-31 23:40:28,871 INFO [train.py:903] (3/4) Epoch 3, batch 200, loss[loss=0.3394, simple_loss=0.3644, pruned_loss=0.1573, over 19786.00 frames. ], tot_loss[loss=0.342, simple_loss=0.384, pruned_loss=0.15, over 2433702.94 frames. ], batch size: 48, lr: 2.71e-02, grad_scale: 8.0 +2023-03-31 23:40:28,910 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-03-31 23:40:37,961 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6962, 1.3296, 1.5264, 1.8841, 1.7881, 1.6120, 1.5913, 1.9062], + device='cuda:3'), covar=tensor([0.0881, 0.1997, 0.1343, 0.0973, 0.1284, 0.0637, 0.1098, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0389, 0.0290, 0.0260, 0.0326, 0.0273, 0.0279, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:41:28,994 INFO [train.py:903] (3/4) Epoch 3, batch 250, loss[loss=0.3335, simple_loss=0.3863, pruned_loss=0.1403, over 19602.00 frames. ], tot_loss[loss=0.3405, simple_loss=0.3833, pruned_loss=0.1489, over 2745693.77 frames. ], batch size: 61, lr: 2.71e-02, grad_scale: 8.0 +2023-03-31 23:41:44,247 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.886e+02 8.729e+02 1.056e+03 1.304e+03 3.760e+03, threshold=2.113e+03, percent-clipped=6.0 +2023-03-31 23:42:19,866 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5531, 4.1379, 2.4011, 3.7672, 1.3317, 3.7743, 3.7355, 3.9879], + device='cuda:3'), covar=tensor([0.0513, 0.0960, 0.1891, 0.0601, 0.3396, 0.0922, 0.0650, 0.0735], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0284, 0.0316, 0.0260, 0.0338, 0.0277, 0.0225, 0.0256], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 23:42:33,069 INFO [train.py:903] (3/4) Epoch 3, batch 300, loss[loss=0.2857, simple_loss=0.3552, pruned_loss=0.1082, over 19661.00 frames. ], tot_loss[loss=0.3392, simple_loss=0.3826, pruned_loss=0.1478, over 2997961.16 frames. ], batch size: 58, lr: 2.71e-02, grad_scale: 8.0 +2023-03-31 23:43:34,498 INFO [train.py:903] (3/4) Epoch 3, batch 350, loss[loss=0.3459, simple_loss=0.3913, pruned_loss=0.1503, over 19674.00 frames. ], tot_loss[loss=0.338, simple_loss=0.382, pruned_loss=0.147, over 3187564.45 frames. ], batch size: 55, lr: 2.70e-02, grad_scale: 8.0 +2023-03-31 23:43:40,035 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-03-31 23:43:46,962 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.582e+02 7.628e+02 9.853e+02 1.217e+03 3.369e+03, threshold=1.971e+03, percent-clipped=3.0 +2023-03-31 23:43:50,258 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-31 23:44:22,035 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8222, 1.1241, 1.3697, 1.6294, 2.5995, 1.1625, 1.6397, 2.5744], + device='cuda:3'), covar=tensor([0.0414, 0.2530, 0.2304, 0.1355, 0.0461, 0.2007, 0.1164, 0.0499], + device='cuda:3'), in_proj_covar=tensor([0.0239, 0.0310, 0.0293, 0.0277, 0.0275, 0.0324, 0.0271, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:44:34,931 INFO [train.py:903] (3/4) Epoch 3, batch 400, loss[loss=0.3475, simple_loss=0.3951, pruned_loss=0.1499, over 19678.00 frames. ], tot_loss[loss=0.3402, simple_loss=0.3833, pruned_loss=0.1485, over 3324038.85 frames. ], batch size: 58, lr: 2.70e-02, grad_scale: 8.0 +2023-03-31 23:45:13,841 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4307, 2.2196, 1.5101, 1.9250, 1.5108, 1.5665, 0.1500, 1.2748], + device='cuda:3'), covar=tensor([0.0211, 0.0237, 0.0197, 0.0219, 0.0515, 0.0370, 0.0635, 0.0396], + device='cuda:3'), in_proj_covar=tensor([0.0229, 0.0237, 0.0231, 0.0247, 0.0313, 0.0261, 0.0254, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 23:45:17,180 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14090.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:45:27,299 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14099.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:45:36,308 INFO [train.py:903] (3/4) Epoch 3, batch 450, loss[loss=0.3064, simple_loss=0.3503, pruned_loss=0.1313, over 19392.00 frames. ], tot_loss[loss=0.3395, simple_loss=0.3828, pruned_loss=0.1481, over 3445754.60 frames. ], batch size: 47, lr: 2.69e-02, grad_scale: 8.0 +2023-03-31 23:45:52,381 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.440e+02 8.213e+02 1.015e+03 1.206e+03 3.609e+03, threshold=2.029e+03, percent-clipped=6.0 +2023-03-31 23:45:55,996 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14121.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:46:10,159 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-03-31 23:46:11,136 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-03-31 23:46:14,941 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14138.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 23:46:24,025 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1289, 1.0591, 2.0150, 1.3939, 2.6594, 2.4375, 2.9519, 1.0776], + device='cuda:3'), covar=tensor([0.1453, 0.2314, 0.1205, 0.1221, 0.0899, 0.0944, 0.1085, 0.2025], + device='cuda:3'), in_proj_covar=tensor([0.0385, 0.0423, 0.0388, 0.0373, 0.0456, 0.0369, 0.0531, 0.0391], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:46:38,893 INFO [train.py:903] (3/4) Epoch 3, batch 500, loss[loss=0.2645, simple_loss=0.3172, pruned_loss=0.1059, over 19757.00 frames. ], tot_loss[loss=0.3392, simple_loss=0.3822, pruned_loss=0.1481, over 3526096.12 frames. ], batch size: 47, lr: 2.69e-02, grad_scale: 8.0 +2023-03-31 23:47:38,987 INFO [train.py:903] (3/4) Epoch 3, batch 550, loss[loss=0.4224, simple_loss=0.4362, pruned_loss=0.2043, over 13695.00 frames. ], tot_loss[loss=0.3406, simple_loss=0.3832, pruned_loss=0.149, over 3577877.44 frames. ], batch size: 136, lr: 2.68e-02, grad_scale: 8.0 +2023-03-31 23:47:47,487 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14214.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:47:51,335 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.716e+02 8.063e+02 9.949e+02 1.307e+03 2.222e+03, threshold=1.990e+03, percent-clipped=3.0 +2023-03-31 23:48:38,811 INFO [train.py:903] (3/4) Epoch 3, batch 600, loss[loss=0.29, simple_loss=0.3408, pruned_loss=0.1196, over 19773.00 frames. ], tot_loss[loss=0.3406, simple_loss=0.3833, pruned_loss=0.149, over 3634973.91 frames. ], batch size: 47, lr: 2.68e-02, grad_scale: 8.0 +2023-03-31 23:49:16,711 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-03-31 23:49:39,184 INFO [train.py:903] (3/4) Epoch 3, batch 650, loss[loss=0.3099, simple_loss=0.3537, pruned_loss=0.133, over 19457.00 frames. ], tot_loss[loss=0.341, simple_loss=0.3835, pruned_loss=0.1492, over 3676830.84 frames. ], batch size: 49, lr: 2.68e-02, grad_scale: 8.0 +2023-03-31 23:49:54,627 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.103e+02 8.257e+02 1.098e+03 1.322e+03 3.191e+03, threshold=2.197e+03, percent-clipped=10.0 +2023-03-31 23:50:26,638 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3435, 2.5164, 1.9236, 2.3674, 2.1382, 1.5993, 0.2776, 2.1507], + device='cuda:3'), covar=tensor([0.0202, 0.0267, 0.0254, 0.0330, 0.0429, 0.0517, 0.0716, 0.0479], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0238, 0.0232, 0.0251, 0.0315, 0.0268, 0.0257, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-03-31 23:50:41,496 INFO [train.py:903] (3/4) Epoch 3, batch 700, loss[loss=0.405, simple_loss=0.4215, pruned_loss=0.1942, over 12913.00 frames. ], tot_loss[loss=0.3421, simple_loss=0.3844, pruned_loss=0.1499, over 3707227.44 frames. ], batch size: 136, lr: 2.67e-02, grad_scale: 8.0 +2023-03-31 23:51:43,778 INFO [train.py:903] (3/4) Epoch 3, batch 750, loss[loss=0.2948, simple_loss=0.3546, pruned_loss=0.1174, over 19599.00 frames. ], tot_loss[loss=0.3393, simple_loss=0.3827, pruned_loss=0.1479, over 3745059.73 frames. ], batch size: 57, lr: 2.67e-02, grad_scale: 8.0 +2023-03-31 23:51:56,450 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.299e+02 7.798e+02 9.551e+02 1.191e+03 2.807e+03, threshold=1.910e+03, percent-clipped=6.0 +2023-03-31 23:52:11,661 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14431.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:52:14,913 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14434.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:52:44,756 INFO [train.py:903] (3/4) Epoch 3, batch 800, loss[loss=0.4176, simple_loss=0.4256, pruned_loss=0.2048, over 13679.00 frames. ], tot_loss[loss=0.3398, simple_loss=0.3833, pruned_loss=0.1481, over 3764077.04 frames. ], batch size: 136, lr: 2.66e-02, grad_scale: 8.0 +2023-03-31 23:52:53,880 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14465.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:52:54,914 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-03-31 23:52:59,823 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14470.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:53:15,263 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14482.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 23:53:31,025 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14495.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:53:42,714 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14505.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:53:44,738 INFO [train.py:903] (3/4) Epoch 3, batch 850, loss[loss=0.3158, simple_loss=0.3549, pruned_loss=0.1384, over 19341.00 frames. ], tot_loss[loss=0.3411, simple_loss=0.3843, pruned_loss=0.1489, over 3783954.59 frames. ], batch size: 47, lr: 2.66e-02, grad_scale: 8.0 +2023-03-31 23:53:58,372 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.292e+02 8.633e+02 1.105e+03 1.534e+03 3.114e+03, threshold=2.210e+03, percent-clipped=11.0 +2023-03-31 23:54:32,133 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-03-31 23:54:35,856 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14549.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:54:45,560 INFO [train.py:903] (3/4) Epoch 3, batch 900, loss[loss=0.3496, simple_loss=0.4003, pruned_loss=0.1494, over 19687.00 frames. ], tot_loss[loss=0.3417, simple_loss=0.385, pruned_loss=0.1492, over 3792156.28 frames. ], batch size: 59, lr: 2.66e-02, grad_scale: 8.0 +2023-03-31 23:55:03,507 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9425, 4.2993, 4.6334, 4.5877, 1.5684, 4.1728, 3.7935, 4.0890], + device='cuda:3'), covar=tensor([0.0500, 0.0384, 0.0391, 0.0266, 0.3025, 0.0209, 0.0314, 0.0898], + device='cuda:3'), in_proj_covar=tensor([0.0345, 0.0311, 0.0436, 0.0325, 0.0467, 0.0223, 0.0289, 0.0431], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-03-31 23:55:15,260 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14580.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:55:34,857 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14597.0, num_to_drop=1, layers_to_drop={0} +2023-03-31 23:55:46,732 INFO [train.py:903] (3/4) Epoch 3, batch 950, loss[loss=0.3525, simple_loss=0.3881, pruned_loss=0.1584, over 17398.00 frames. ], tot_loss[loss=0.3389, simple_loss=0.3827, pruned_loss=0.1475, over 3810832.89 frames. ], batch size: 101, lr: 2.65e-02, grad_scale: 4.0 +2023-03-31 23:55:46,740 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-03-31 23:56:00,994 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.858e+02 7.381e+02 9.246e+02 1.263e+03 4.500e+03, threshold=1.849e+03, percent-clipped=5.0 +2023-03-31 23:56:21,988 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2155, 1.8448, 1.3932, 1.2461, 1.6839, 1.0294, 0.8658, 1.4613], + device='cuda:3'), covar=tensor([0.0675, 0.0394, 0.0918, 0.0480, 0.0358, 0.1038, 0.0659, 0.0425], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0217, 0.0313, 0.0255, 0.0210, 0.0321, 0.0277, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-03-31 23:56:46,894 INFO [train.py:903] (3/4) Epoch 3, batch 1000, loss[loss=0.3327, simple_loss=0.3873, pruned_loss=0.1391, over 19687.00 frames. ], tot_loss[loss=0.3383, simple_loss=0.3819, pruned_loss=0.1473, over 3831677.95 frames. ], batch size: 60, lr: 2.65e-02, grad_scale: 4.0 +2023-03-31 23:57:10,686 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6087, 2.5803, 2.2988, 3.6983, 2.6884, 4.3534, 3.8297, 2.4974], + device='cuda:3'), covar=tensor([0.0963, 0.0662, 0.0391, 0.0436, 0.0823, 0.0130, 0.0499, 0.0536], + device='cuda:3'), in_proj_covar=tensor([0.0415, 0.0390, 0.0397, 0.0517, 0.0468, 0.0301, 0.0489, 0.0400], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:57:13,855 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9805, 2.0380, 2.1548, 2.6730, 4.4477, 1.3307, 2.1109, 4.3592], + device='cuda:3'), covar=tensor([0.0184, 0.2092, 0.1920, 0.1223, 0.0315, 0.1979, 0.1083, 0.0305], + device='cuda:3'), in_proj_covar=tensor([0.0234, 0.0301, 0.0292, 0.0278, 0.0273, 0.0318, 0.0262, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-03-31 23:57:38,689 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-03-31 23:57:47,680 INFO [train.py:903] (3/4) Epoch 3, batch 1050, loss[loss=0.3981, simple_loss=0.4333, pruned_loss=0.1815, over 17425.00 frames. ], tot_loss[loss=0.3357, simple_loss=0.3795, pruned_loss=0.146, over 3814030.19 frames. ], batch size: 101, lr: 2.64e-02, grad_scale: 4.0 +2023-03-31 23:58:01,055 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.125e+02 7.305e+02 8.951e+02 1.118e+03 2.421e+03, threshold=1.790e+03, percent-clipped=2.0 +2023-03-31 23:58:17,603 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-03-31 23:58:48,370 INFO [train.py:903] (3/4) Epoch 3, batch 1100, loss[loss=0.2895, simple_loss=0.3366, pruned_loss=0.1212, over 19131.00 frames. ], tot_loss[loss=0.3335, simple_loss=0.3781, pruned_loss=0.1444, over 3826678.08 frames. ], batch size: 42, lr: 2.64e-02, grad_scale: 4.0 +2023-03-31 23:58:49,837 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14758.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:59:10,923 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14775.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:59:47,760 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14805.0, num_to_drop=0, layers_to_drop=set() +2023-03-31 23:59:49,456 INFO [train.py:903] (3/4) Epoch 3, batch 1150, loss[loss=0.4636, simple_loss=0.4653, pruned_loss=0.231, over 14015.00 frames. ], tot_loss[loss=0.3346, simple_loss=0.3792, pruned_loss=0.1451, over 3821628.53 frames. ], batch size: 138, lr: 2.64e-02, grad_scale: 4.0 +2023-04-01 00:00:03,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.960e+02 7.465e+02 1.021e+03 1.238e+03 3.548e+03, threshold=2.043e+03, percent-clipped=7.0 +2023-04-01 00:00:16,807 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14830.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:00:23,708 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14836.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:00:39,270 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-01 00:00:40,937 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=14849.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:00:45,734 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14853.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 00:00:49,931 INFO [train.py:903] (3/4) Epoch 3, batch 1200, loss[loss=0.2667, simple_loss=0.3326, pruned_loss=0.1005, over 19673.00 frames. ], tot_loss[loss=0.3344, simple_loss=0.3792, pruned_loss=0.1448, over 3829703.78 frames. ], batch size: 53, lr: 2.63e-02, grad_scale: 8.0 +2023-04-01 00:00:55,823 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14861.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:01:08,125 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5825, 3.9653, 4.2573, 4.2514, 1.5687, 3.8612, 3.4775, 3.7625], + device='cuda:3'), covar=tensor([0.0516, 0.0526, 0.0405, 0.0305, 0.3063, 0.0246, 0.0359, 0.0836], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0318, 0.0445, 0.0339, 0.0473, 0.0233, 0.0299, 0.0438], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 00:01:15,100 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14878.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 00:01:17,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 00:01:31,313 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14890.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:01:51,484 INFO [train.py:903] (3/4) Epoch 3, batch 1250, loss[loss=0.4484, simple_loss=0.4467, pruned_loss=0.225, over 13071.00 frames. ], tot_loss[loss=0.3354, simple_loss=0.3799, pruned_loss=0.1455, over 3826487.00 frames. ], batch size: 136, lr: 2.63e-02, grad_scale: 8.0 +2023-04-01 00:02:05,937 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.325e+02 7.701e+02 1.002e+03 1.250e+03 2.941e+03, threshold=2.004e+03, percent-clipped=3.0 +2023-04-01 00:02:13,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2426, 2.1179, 1.9276, 3.1621, 2.3510, 3.8367, 3.2058, 1.7661], + device='cuda:3'), covar=tensor([0.1079, 0.0763, 0.0486, 0.0571, 0.0996, 0.0195, 0.0735, 0.0812], + device='cuda:3'), in_proj_covar=tensor([0.0427, 0.0397, 0.0406, 0.0531, 0.0477, 0.0306, 0.0505, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:02:53,118 INFO [train.py:903] (3/4) Epoch 3, batch 1300, loss[loss=0.3147, simple_loss=0.3614, pruned_loss=0.1341, over 19717.00 frames. ], tot_loss[loss=0.3348, simple_loss=0.3791, pruned_loss=0.1452, over 3821235.72 frames. ], batch size: 51, lr: 2.62e-02, grad_scale: 8.0 +2023-04-01 00:03:02,107 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14964.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:03:54,924 INFO [train.py:903] (3/4) Epoch 3, batch 1350, loss[loss=0.3546, simple_loss=0.397, pruned_loss=0.156, over 19575.00 frames. ], tot_loss[loss=0.3364, simple_loss=0.3803, pruned_loss=0.1463, over 3824264.76 frames. ], batch size: 61, lr: 2.62e-02, grad_scale: 8.0 +2023-04-01 00:03:59,605 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0883, 1.9669, 1.7603, 3.0334, 2.0193, 3.1728, 2.6063, 1.7175], + device='cuda:3'), covar=tensor([0.0881, 0.0693, 0.0443, 0.0404, 0.0765, 0.0178, 0.0714, 0.0672], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0399, 0.0405, 0.0523, 0.0467, 0.0305, 0.0497, 0.0405], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:04:10,657 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.789e+02 8.351e+02 9.883e+02 1.225e+03 3.360e+03, threshold=1.977e+03, percent-clipped=2.0 +2023-04-01 00:04:57,772 INFO [train.py:903] (3/4) Epoch 3, batch 1400, loss[loss=0.422, simple_loss=0.4315, pruned_loss=0.2063, over 19649.00 frames. ], tot_loss[loss=0.3364, simple_loss=0.3797, pruned_loss=0.1466, over 3811495.06 frames. ], batch size: 58, lr: 2.62e-02, grad_scale: 8.0 +2023-04-01 00:05:29,115 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7299, 1.5846, 1.5892, 1.8422, 3.3510, 1.0986, 1.8805, 3.3861], + device='cuda:3'), covar=tensor([0.0279, 0.2087, 0.2093, 0.1305, 0.0354, 0.1998, 0.1158, 0.0332], + device='cuda:3'), in_proj_covar=tensor([0.0243, 0.0303, 0.0296, 0.0282, 0.0282, 0.0324, 0.0266, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:05:53,869 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=15102.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:05:59,104 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 00:06:00,254 INFO [train.py:903] (3/4) Epoch 3, batch 1450, loss[loss=0.3407, simple_loss=0.3867, pruned_loss=0.1474, over 19611.00 frames. ], tot_loss[loss=0.3339, simple_loss=0.3782, pruned_loss=0.1448, over 3828582.32 frames. ], batch size: 57, lr: 2.61e-02, grad_scale: 8.0 +2023-04-01 00:06:00,482 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8157, 3.1490, 3.3185, 3.4325, 1.3415, 3.1725, 2.9478, 2.6199], + device='cuda:3'), covar=tensor([0.1239, 0.1134, 0.1128, 0.0881, 0.4542, 0.0708, 0.0752, 0.2074], + device='cuda:3'), in_proj_covar=tensor([0.0347, 0.0312, 0.0442, 0.0333, 0.0461, 0.0228, 0.0292, 0.0425], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 00:06:13,781 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.668e+02 7.923e+02 9.351e+02 1.150e+03 2.880e+03, threshold=1.870e+03, percent-clipped=3.0 +2023-04-01 00:06:49,295 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15146.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:06:51,634 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1799, 1.1428, 2.0055, 1.3617, 2.3948, 2.2440, 2.6107, 0.9164], + device='cuda:3'), covar=tensor([0.1330, 0.2145, 0.1109, 0.1238, 0.0913, 0.0968, 0.1069, 0.1996], + device='cuda:3'), in_proj_covar=tensor([0.0401, 0.0440, 0.0406, 0.0386, 0.0475, 0.0380, 0.0557, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:07:01,544 INFO [train.py:903] (3/4) Epoch 3, batch 1500, loss[loss=0.3178, simple_loss=0.3748, pruned_loss=0.1304, over 19754.00 frames. ], tot_loss[loss=0.3331, simple_loss=0.378, pruned_loss=0.1441, over 3833796.81 frames. ], batch size: 54, lr: 2.61e-02, grad_scale: 8.0 +2023-04-01 00:07:20,447 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15171.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:08:03,673 INFO [train.py:903] (3/4) Epoch 3, batch 1550, loss[loss=0.2894, simple_loss=0.3518, pruned_loss=0.1135, over 19644.00 frames. ], tot_loss[loss=0.3316, simple_loss=0.3768, pruned_loss=0.1432, over 3838329.16 frames. ], batch size: 53, lr: 2.60e-02, grad_scale: 8.0 +2023-04-01 00:08:12,819 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=15213.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:08:18,419 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15217.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:08:20,279 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.312e+02 7.825e+02 9.432e+02 1.149e+03 3.008e+03, threshold=1.886e+03, percent-clipped=3.0 +2023-04-01 00:08:21,917 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15220.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:08:52,281 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15245.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:08:52,350 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1074, 1.0337, 1.8243, 1.1782, 2.2904, 2.1573, 2.4299, 0.8963], + device='cuda:3'), covar=tensor([0.1366, 0.2224, 0.1106, 0.1307, 0.0918, 0.0996, 0.1077, 0.2014], + device='cuda:3'), in_proj_covar=tensor([0.0398, 0.0439, 0.0402, 0.0386, 0.0475, 0.0384, 0.0560, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:09:08,812 INFO [train.py:903] (3/4) Epoch 3, batch 1600, loss[loss=0.2937, simple_loss=0.3425, pruned_loss=0.1224, over 19754.00 frames. ], tot_loss[loss=0.3308, simple_loss=0.3765, pruned_loss=0.1426, over 3844567.05 frames. ], batch size: 45, lr: 2.60e-02, grad_scale: 8.0 +2023-04-01 00:09:32,820 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 00:09:38,930 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5105, 1.0905, 1.4294, 0.9621, 2.8704, 3.4594, 3.2581, 3.6454], + device='cuda:3'), covar=tensor([0.1466, 0.2908, 0.2924, 0.2294, 0.0392, 0.0109, 0.0224, 0.0117], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0282, 0.0329, 0.0281, 0.0193, 0.0103, 0.0199, 0.0116], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 00:10:02,522 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8953, 3.5136, 2.1868, 3.2363, 1.3134, 3.3719, 3.2337, 3.3311], + device='cuda:3'), covar=tensor([0.0664, 0.1089, 0.1869, 0.0733, 0.3361, 0.0953, 0.0674, 0.0918], + device='cuda:3'), in_proj_covar=tensor([0.0311, 0.0300, 0.0329, 0.0275, 0.0352, 0.0294, 0.0241, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:10:10,177 INFO [train.py:903] (3/4) Epoch 3, batch 1650, loss[loss=0.3904, simple_loss=0.4255, pruned_loss=0.1777, over 19592.00 frames. ], tot_loss[loss=0.3295, simple_loss=0.3754, pruned_loss=0.1418, over 3849964.39 frames. ], batch size: 61, lr: 2.60e-02, grad_scale: 8.0 +2023-04-01 00:10:24,981 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.775e+02 8.096e+02 9.310e+02 1.117e+03 2.889e+03, threshold=1.862e+03, percent-clipped=6.0 +2023-04-01 00:10:25,649 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-04-01 00:11:11,888 INFO [train.py:903] (3/4) Epoch 3, batch 1700, loss[loss=0.3288, simple_loss=0.3792, pruned_loss=0.1393, over 17495.00 frames. ], tot_loss[loss=0.3299, simple_loss=0.3754, pruned_loss=0.1422, over 3839645.34 frames. ], batch size: 101, lr: 2.59e-02, grad_scale: 8.0 +2023-04-01 00:11:50,336 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 00:12:04,725 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-01 00:12:13,187 INFO [train.py:903] (3/4) Epoch 3, batch 1750, loss[loss=0.2944, simple_loss=0.358, pruned_loss=0.1154, over 19768.00 frames. ], tot_loss[loss=0.3295, simple_loss=0.3754, pruned_loss=0.1418, over 3844451.99 frames. ], batch size: 54, lr: 2.59e-02, grad_scale: 8.0 +2023-04-01 00:12:30,184 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.268e+02 8.359e+02 1.065e+03 1.276e+03 4.198e+03, threshold=2.129e+03, percent-clipped=6.0 +2023-04-01 00:13:17,309 INFO [train.py:903] (3/4) Epoch 3, batch 1800, loss[loss=0.3146, simple_loss=0.3699, pruned_loss=0.1297, over 18722.00 frames. ], tot_loss[loss=0.3289, simple_loss=0.3748, pruned_loss=0.1415, over 3839124.53 frames. ], batch size: 74, lr: 2.58e-02, grad_scale: 8.0 +2023-04-01 00:13:37,281 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15473.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:14:07,915 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15498.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:14:14,382 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 00:14:17,663 INFO [train.py:903] (3/4) Epoch 3, batch 1850, loss[loss=0.3139, simple_loss=0.3724, pruned_loss=0.1276, over 19655.00 frames. ], tot_loss[loss=0.3275, simple_loss=0.3741, pruned_loss=0.1405, over 3818533.80 frames. ], batch size: 58, lr: 2.58e-02, grad_scale: 8.0 +2023-04-01 00:14:23,557 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2482, 1.1446, 1.9589, 1.4090, 2.7250, 2.4212, 3.0551, 1.0564], + device='cuda:3'), covar=tensor([0.1379, 0.2232, 0.1160, 0.1246, 0.0839, 0.0954, 0.0864, 0.2074], + device='cuda:3'), in_proj_covar=tensor([0.0392, 0.0429, 0.0397, 0.0376, 0.0464, 0.0378, 0.0543, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:14:32,125 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.736e+02 7.448e+02 9.407e+02 1.169e+03 2.273e+03, threshold=1.881e+03, percent-clipped=1.0 +2023-04-01 00:14:50,377 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 00:15:18,059 INFO [train.py:903] (3/4) Epoch 3, batch 1900, loss[loss=0.3989, simple_loss=0.4211, pruned_loss=0.1883, over 19645.00 frames. ], tot_loss[loss=0.33, simple_loss=0.3762, pruned_loss=0.142, over 3820025.84 frames. ], batch size: 55, lr: 2.58e-02, grad_scale: 8.0 +2023-04-01 00:15:18,213 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=15557.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:15:18,436 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2387, 1.2107, 0.9631, 1.2029, 1.0939, 1.1632, 0.9850, 1.1089], + device='cuda:3'), covar=tensor([0.0815, 0.1010, 0.1299, 0.0688, 0.0876, 0.0529, 0.0933, 0.0674], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0386, 0.0293, 0.0260, 0.0322, 0.0267, 0.0281, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:15:36,177 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 00:15:41,726 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 00:16:04,626 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 00:16:19,661 INFO [train.py:903] (3/4) Epoch 3, batch 1950, loss[loss=0.3838, simple_loss=0.4162, pruned_loss=0.1757, over 19379.00 frames. ], tot_loss[loss=0.331, simple_loss=0.3769, pruned_loss=0.1426, over 3818247.03 frames. ], batch size: 70, lr: 2.57e-02, grad_scale: 8.0 +2023-04-01 00:16:36,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.560e+02 7.789e+02 9.617e+02 1.296e+03 2.448e+03, threshold=1.923e+03, percent-clipped=3.0 +2023-04-01 00:16:54,042 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5948, 1.3566, 1.2396, 1.5985, 1.3851, 1.3937, 1.2209, 1.5922], + device='cuda:3'), covar=tensor([0.0928, 0.1465, 0.1475, 0.0897, 0.1127, 0.0638, 0.1092, 0.0701], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0386, 0.0289, 0.0255, 0.0322, 0.0263, 0.0279, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:16:54,064 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1296, 1.9550, 1.4589, 1.3796, 1.7171, 0.8873, 0.8172, 1.6975], + device='cuda:3'), covar=tensor([0.0804, 0.0469, 0.0979, 0.0525, 0.0486, 0.1348, 0.0787, 0.0386], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0228, 0.0309, 0.0254, 0.0220, 0.0306, 0.0278, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:17:10,258 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-04-01 00:17:22,860 INFO [train.py:903] (3/4) Epoch 3, batch 2000, loss[loss=0.2815, simple_loss=0.3365, pruned_loss=0.1133, over 19625.00 frames. ], tot_loss[loss=0.3284, simple_loss=0.3749, pruned_loss=0.1409, over 3832165.26 frames. ], batch size: 50, lr: 2.57e-02, grad_scale: 8.0 +2023-04-01 00:17:41,421 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15672.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:18:20,116 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 00:18:23,593 INFO [train.py:903] (3/4) Epoch 3, batch 2050, loss[loss=0.3014, simple_loss=0.3647, pruned_loss=0.119, over 19272.00 frames. ], tot_loss[loss=0.3284, simple_loss=0.3749, pruned_loss=0.141, over 3823326.21 frames. ], batch size: 66, lr: 2.57e-02, grad_scale: 8.0 +2023-04-01 00:18:38,231 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.734e+02 7.410e+02 9.269e+02 1.172e+03 2.915e+03, threshold=1.854e+03, percent-clipped=8.0 +2023-04-01 00:18:38,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 00:18:39,612 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 00:18:58,658 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 00:19:06,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0397, 1.7515, 1.2724, 1.2390, 1.4962, 0.8165, 0.6894, 1.4817], + device='cuda:3'), covar=tensor([0.0590, 0.0388, 0.0944, 0.0483, 0.0412, 0.1137, 0.0767, 0.0347], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0223, 0.0312, 0.0255, 0.0214, 0.0308, 0.0276, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:19:25,007 INFO [train.py:903] (3/4) Epoch 3, batch 2100, loss[loss=0.3012, simple_loss=0.3607, pruned_loss=0.1209, over 19527.00 frames. ], tot_loss[loss=0.3303, simple_loss=0.3765, pruned_loss=0.1421, over 3807187.53 frames. ], batch size: 54, lr: 2.56e-02, grad_scale: 8.0 +2023-04-01 00:19:52,226 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 00:20:13,818 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 00:20:20,264 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5879, 2.4442, 1.5892, 1.8012, 2.0442, 1.0779, 1.1922, 1.5379], + device='cuda:3'), covar=tensor([0.0965, 0.0411, 0.0915, 0.0597, 0.0521, 0.1254, 0.0934, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0224, 0.0315, 0.0258, 0.0215, 0.0308, 0.0278, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:20:25,858 INFO [train.py:903] (3/4) Epoch 3, batch 2150, loss[loss=0.3557, simple_loss=0.3876, pruned_loss=0.1619, over 19724.00 frames. ], tot_loss[loss=0.3285, simple_loss=0.3751, pruned_loss=0.141, over 3814864.98 frames. ], batch size: 51, lr: 2.56e-02, grad_scale: 8.0 +2023-04-01 00:20:42,353 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.336e+02 7.356e+02 9.022e+02 1.284e+03 2.686e+03, threshold=1.804e+03, percent-clipped=4.0 +2023-04-01 00:21:28,857 INFO [train.py:903] (3/4) Epoch 3, batch 2200, loss[loss=0.3176, simple_loss=0.376, pruned_loss=0.1296, over 19669.00 frames. ], tot_loss[loss=0.3277, simple_loss=0.3743, pruned_loss=0.1405, over 3817072.29 frames. ], batch size: 58, lr: 2.55e-02, grad_scale: 8.0 +2023-04-01 00:22:23,678 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=15901.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:22:30,393 INFO [train.py:903] (3/4) Epoch 3, batch 2250, loss[loss=0.287, simple_loss=0.3527, pruned_loss=0.1107, over 18761.00 frames. ], tot_loss[loss=0.3284, simple_loss=0.3749, pruned_loss=0.1409, over 3808481.29 frames. ], batch size: 74, lr: 2.55e-02, grad_scale: 8.0 +2023-04-01 00:22:35,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9509, 1.4287, 1.2243, 1.8795, 1.5369, 2.0942, 2.1427, 2.0420], + device='cuda:3'), covar=tensor([0.0734, 0.1188, 0.1307, 0.1179, 0.1342, 0.0773, 0.1079, 0.0635], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0289, 0.0276, 0.0314, 0.0325, 0.0262, 0.0293, 0.0256], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 00:22:44,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.843e+02 7.012e+02 9.226e+02 1.146e+03 2.721e+03, threshold=1.845e+03, percent-clipped=4.0 +2023-04-01 00:22:51,992 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7042, 4.2067, 2.4187, 3.7307, 1.2501, 3.9065, 3.7154, 4.0549], + device='cuda:3'), covar=tensor([0.0568, 0.1408, 0.2186, 0.0735, 0.4261, 0.1015, 0.0819, 0.0785], + device='cuda:3'), in_proj_covar=tensor([0.0309, 0.0297, 0.0331, 0.0270, 0.0342, 0.0288, 0.0238, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:22:55,698 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15928.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:23:00,327 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8422, 1.7500, 1.4792, 1.7283, 1.6408, 1.6849, 1.3845, 1.7941], + device='cuda:3'), covar=tensor([0.0779, 0.1462, 0.1288, 0.1096, 0.1198, 0.0538, 0.0999, 0.0659], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0380, 0.0280, 0.0255, 0.0314, 0.0261, 0.0270, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:23:08,945 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7812, 1.8737, 1.2286, 1.3543, 1.3235, 1.3830, 0.0493, 0.5373], + device='cuda:3'), covar=tensor([0.0213, 0.0193, 0.0149, 0.0155, 0.0431, 0.0211, 0.0382, 0.0364], + device='cuda:3'), in_proj_covar=tensor([0.0238, 0.0234, 0.0237, 0.0257, 0.0314, 0.0256, 0.0249, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:23:26,961 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15953.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:23:31,839 INFO [train.py:903] (3/4) Epoch 3, batch 2300, loss[loss=0.3676, simple_loss=0.4049, pruned_loss=0.1651, over 19541.00 frames. ], tot_loss[loss=0.3283, simple_loss=0.3746, pruned_loss=0.141, over 3802829.92 frames. ], batch size: 56, lr: 2.55e-02, grad_scale: 8.0 +2023-04-01 00:23:44,531 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 00:23:44,947 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6345, 1.4856, 1.3242, 1.5120, 1.3017, 1.4374, 1.2512, 1.5399], + device='cuda:3'), covar=tensor([0.0826, 0.1180, 0.1159, 0.0847, 0.1055, 0.0556, 0.0944, 0.0625], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0385, 0.0284, 0.0255, 0.0320, 0.0261, 0.0269, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:23:47,093 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2300, 1.1736, 1.8917, 1.3616, 2.6996, 2.5788, 2.8214, 1.1297], + device='cuda:3'), covar=tensor([0.1337, 0.2235, 0.1197, 0.1177, 0.0847, 0.0881, 0.1137, 0.1999], + device='cuda:3'), in_proj_covar=tensor([0.0398, 0.0437, 0.0397, 0.0378, 0.0466, 0.0389, 0.0555, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:24:02,632 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-04-01 00:24:33,433 INFO [train.py:903] (3/4) Epoch 3, batch 2350, loss[loss=0.3813, simple_loss=0.4215, pruned_loss=0.1705, over 19683.00 frames. ], tot_loss[loss=0.3254, simple_loss=0.3725, pruned_loss=0.1392, over 3815381.29 frames. ], batch size: 60, lr: 2.54e-02, grad_scale: 8.0 +2023-04-01 00:24:48,808 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.022e+02 7.519e+02 9.138e+02 1.115e+03 3.205e+03, threshold=1.828e+03, percent-clipped=8.0 +2023-04-01 00:24:49,961 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16019.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:25:15,424 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 00:25:17,324 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-04-01 00:25:31,112 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 00:25:34,313 INFO [train.py:903] (3/4) Epoch 3, batch 2400, loss[loss=0.3302, simple_loss=0.3815, pruned_loss=0.1395, over 18817.00 frames. ], tot_loss[loss=0.3287, simple_loss=0.3752, pruned_loss=0.1411, over 3822376.52 frames. ], batch size: 74, lr: 2.54e-02, grad_scale: 8.0 +2023-04-01 00:26:33,065 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16103.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:26:37,321 INFO [train.py:903] (3/4) Epoch 3, batch 2450, loss[loss=0.3491, simple_loss=0.3967, pruned_loss=0.1508, over 19601.00 frames. ], tot_loss[loss=0.3291, simple_loss=0.3754, pruned_loss=0.1414, over 3813041.67 frames. ], batch size: 57, lr: 2.54e-02, grad_scale: 8.0 +2023-04-01 00:26:51,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.658e+02 8.374e+02 9.822e+02 1.305e+03 3.634e+03, threshold=1.964e+03, percent-clipped=9.0 +2023-04-01 00:27:07,886 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0990, 2.9966, 1.6920, 2.2140, 1.7709, 1.7454, 0.2951, 2.1255], + device='cuda:3'), covar=tensor([0.0260, 0.0220, 0.0249, 0.0274, 0.0448, 0.0439, 0.0630, 0.0440], + device='cuda:3'), in_proj_covar=tensor([0.0244, 0.0240, 0.0242, 0.0265, 0.0320, 0.0266, 0.0259, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:27:38,194 INFO [train.py:903] (3/4) Epoch 3, batch 2500, loss[loss=0.3364, simple_loss=0.3714, pruned_loss=0.1506, over 19615.00 frames. ], tot_loss[loss=0.3273, simple_loss=0.374, pruned_loss=0.1403, over 3823051.30 frames. ], batch size: 50, lr: 2.53e-02, grad_scale: 8.0 +2023-04-01 00:27:50,922 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16167.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:28:40,057 INFO [train.py:903] (3/4) Epoch 3, batch 2550, loss[loss=0.4258, simple_loss=0.443, pruned_loss=0.2043, over 13259.00 frames. ], tot_loss[loss=0.3286, simple_loss=0.3749, pruned_loss=0.1411, over 3803517.65 frames. ], batch size: 135, lr: 2.53e-02, grad_scale: 8.0 +2023-04-01 00:28:56,248 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.521e+02 7.641e+02 9.209e+02 1.283e+03 2.881e+03, threshold=1.842e+03, percent-clipped=1.0 +2023-04-01 00:29:28,552 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=16245.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:29:36,516 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 00:29:37,933 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2998, 3.8017, 3.9474, 3.9249, 1.3245, 3.5174, 3.3089, 3.5009], + device='cuda:3'), covar=tensor([0.0851, 0.0649, 0.0596, 0.0429, 0.3428, 0.0333, 0.0444, 0.1174], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0332, 0.0450, 0.0345, 0.0469, 0.0241, 0.0295, 0.0434], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 00:29:44,015 INFO [train.py:903] (3/4) Epoch 3, batch 2600, loss[loss=0.3237, simple_loss=0.376, pruned_loss=0.1357, over 19648.00 frames. ], tot_loss[loss=0.3307, simple_loss=0.3764, pruned_loss=0.1425, over 3806612.61 frames. ], batch size: 55, lr: 2.53e-02, grad_scale: 8.0 +2023-04-01 00:30:46,335 INFO [train.py:903] (3/4) Epoch 3, batch 2650, loss[loss=0.3282, simple_loss=0.3863, pruned_loss=0.1351, over 19501.00 frames. ], tot_loss[loss=0.3289, simple_loss=0.3753, pruned_loss=0.1412, over 3809879.57 frames. ], batch size: 64, lr: 2.52e-02, grad_scale: 8.0 +2023-04-01 00:31:00,229 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.667e+02 7.570e+02 8.863e+02 1.283e+03 4.568e+03, threshold=1.773e+03, percent-clipped=9.0 +2023-04-01 00:31:06,929 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 00:31:47,149 INFO [train.py:903] (3/4) Epoch 3, batch 2700, loss[loss=0.2655, simple_loss=0.3327, pruned_loss=0.0992, over 19736.00 frames. ], tot_loss[loss=0.3263, simple_loss=0.3732, pruned_loss=0.1397, over 3809467.77 frames. ], batch size: 51, lr: 2.52e-02, grad_scale: 8.0 +2023-04-01 00:31:51,958 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16360.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:31:54,894 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=16363.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:32:01,321 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-04-01 00:32:13,166 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9132, 4.1001, 4.4828, 4.4567, 1.4437, 4.0503, 3.7116, 3.9628], + device='cuda:3'), covar=tensor([0.0502, 0.0495, 0.0396, 0.0248, 0.3135, 0.0232, 0.0360, 0.0860], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0337, 0.0463, 0.0349, 0.0476, 0.0243, 0.0300, 0.0449], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 00:32:26,173 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16389.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:32:47,349 INFO [train.py:903] (3/4) Epoch 3, batch 2750, loss[loss=0.3286, simple_loss=0.3771, pruned_loss=0.14, over 19614.00 frames. ], tot_loss[loss=0.3262, simple_loss=0.3734, pruned_loss=0.1395, over 3816517.09 frames. ], batch size: 57, lr: 2.52e-02, grad_scale: 8.0 +2023-04-01 00:33:01,694 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.986e+02 7.672e+02 9.838e+02 1.209e+03 2.463e+03, threshold=1.968e+03, percent-clipped=5.0 +2023-04-01 00:33:05,406 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7757, 1.2949, 1.2740, 1.8878, 1.4695, 1.9079, 2.1325, 1.6836], + device='cuda:3'), covar=tensor([0.0843, 0.1327, 0.1437, 0.1216, 0.1350, 0.0906, 0.1020, 0.0860], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0287, 0.0273, 0.0312, 0.0317, 0.0263, 0.0292, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 00:33:35,977 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=16447.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:33:46,966 INFO [train.py:903] (3/4) Epoch 3, batch 2800, loss[loss=0.2832, simple_loss=0.3299, pruned_loss=0.1183, over 17359.00 frames. ], tot_loss[loss=0.3283, simple_loss=0.3747, pruned_loss=0.1409, over 3808891.76 frames. ], batch size: 38, lr: 2.51e-02, grad_scale: 8.0 +2023-04-01 00:34:13,570 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16478.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:34:48,738 INFO [train.py:903] (3/4) Epoch 3, batch 2850, loss[loss=0.3912, simple_loss=0.4151, pruned_loss=0.1837, over 13087.00 frames. ], tot_loss[loss=0.3288, simple_loss=0.3752, pruned_loss=0.1413, over 3805509.12 frames. ], batch size: 137, lr: 2.51e-02, grad_scale: 8.0 +2023-04-01 00:34:54,279 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=16511.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:35:03,282 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.663e+02 7.862e+02 1.093e+03 1.433e+03 3.382e+03, threshold=2.185e+03, percent-clipped=3.0 +2023-04-01 00:35:47,689 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 00:35:49,902 INFO [train.py:903] (3/4) Epoch 3, batch 2900, loss[loss=0.2779, simple_loss=0.3291, pruned_loss=0.1134, over 19756.00 frames. ], tot_loss[loss=0.3294, simple_loss=0.3755, pruned_loss=0.1416, over 3806942.06 frames. ], batch size: 47, lr: 2.51e-02, grad_scale: 8.0 +2023-04-01 00:35:52,752 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9651, 2.0299, 1.5179, 1.5618, 1.4318, 1.5839, 0.1948, 1.0670], + device='cuda:3'), covar=tensor([0.0240, 0.0201, 0.0145, 0.0185, 0.0404, 0.0249, 0.0475, 0.0357], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0244, 0.0242, 0.0259, 0.0318, 0.0275, 0.0258, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:35:55,896 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0154, 2.0259, 1.9685, 2.5118, 4.6119, 1.4557, 2.2448, 4.2817], + device='cuda:3'), covar=tensor([0.0168, 0.1981, 0.1808, 0.1124, 0.0247, 0.1816, 0.1064, 0.0311], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0307, 0.0296, 0.0278, 0.0286, 0.0320, 0.0272, 0.0283], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:35:57,090 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16562.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:36:37,491 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8830, 1.9794, 1.7321, 3.0013, 2.1555, 3.0746, 2.7270, 1.8215], + device='cuda:3'), covar=tensor([0.1166, 0.0772, 0.0508, 0.0471, 0.0920, 0.0249, 0.0762, 0.0777], + device='cuda:3'), in_proj_covar=tensor([0.0450, 0.0419, 0.0424, 0.0555, 0.0500, 0.0332, 0.0513, 0.0422], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:36:51,764 INFO [train.py:903] (3/4) Epoch 3, batch 2950, loss[loss=0.3393, simple_loss=0.3967, pruned_loss=0.141, over 19691.00 frames. ], tot_loss[loss=0.3299, simple_loss=0.376, pruned_loss=0.1419, over 3807801.67 frames. ], batch size: 59, lr: 2.50e-02, grad_scale: 16.0 +2023-04-01 00:37:02,792 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16616.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:37:07,338 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.542e+02 7.271e+02 9.405e+02 1.170e+03 2.853e+03, threshold=1.881e+03, percent-clipped=4.0 +2023-04-01 00:37:15,876 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16626.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:37:33,996 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16641.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:37:43,775 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9793, 0.8927, 1.2608, 1.0899, 1.7181, 1.5049, 1.8808, 0.8249], + device='cuda:3'), covar=tensor([0.1182, 0.1886, 0.1023, 0.1153, 0.0712, 0.0886, 0.0713, 0.1748], + device='cuda:3'), in_proj_covar=tensor([0.0403, 0.0438, 0.0402, 0.0384, 0.0484, 0.0389, 0.0562, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:37:52,200 INFO [train.py:903] (3/4) Epoch 3, batch 3000, loss[loss=0.2683, simple_loss=0.3282, pruned_loss=0.1042, over 19487.00 frames. ], tot_loss[loss=0.3274, simple_loss=0.3742, pruned_loss=0.1403, over 3822627.72 frames. ], batch size: 49, lr: 2.50e-02, grad_scale: 16.0 +2023-04-01 00:37:52,200 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 00:38:05,259 INFO [train.py:937] (3/4) Epoch 3, validation: loss=0.231, simple_loss=0.3246, pruned_loss=0.06867, over 944034.00 frames. +2023-04-01 00:38:05,260 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18419MB +2023-04-01 00:38:08,702 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 00:39:07,683 INFO [train.py:903] (3/4) Epoch 3, batch 3050, loss[loss=0.354, simple_loss=0.4009, pruned_loss=0.1536, over 19134.00 frames. ], tot_loss[loss=0.328, simple_loss=0.3746, pruned_loss=0.1408, over 3812955.80 frames. ], batch size: 69, lr: 2.49e-02, grad_scale: 16.0 +2023-04-01 00:39:22,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.950e+02 7.860e+02 1.014e+03 1.267e+03 1.851e+03, threshold=2.027e+03, percent-clipped=0.0 +2023-04-01 00:39:34,424 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5308, 1.1428, 1.6907, 1.2777, 2.6312, 3.3640, 3.2059, 3.6058], + device='cuda:3'), covar=tensor([0.1327, 0.2594, 0.2427, 0.1884, 0.0423, 0.0123, 0.0203, 0.0092], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0277, 0.0329, 0.0271, 0.0195, 0.0106, 0.0197, 0.0111], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 00:39:38,697 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=16733.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:39:40,026 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16734.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:40:08,118 INFO [train.py:903] (3/4) Epoch 3, batch 3100, loss[loss=0.3103, simple_loss=0.3541, pruned_loss=0.1333, over 19370.00 frames. ], tot_loss[loss=0.3252, simple_loss=0.3728, pruned_loss=0.1388, over 3825892.34 frames. ], batch size: 47, lr: 2.49e-02, grad_scale: 16.0 +2023-04-01 00:40:11,507 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16759.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:40:34,232 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-04-01 00:40:47,274 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-04-01 00:41:08,266 INFO [train.py:903] (3/4) Epoch 3, batch 3150, loss[loss=0.3421, simple_loss=0.3815, pruned_loss=0.1514, over 19592.00 frames. ], tot_loss[loss=0.3243, simple_loss=0.3716, pruned_loss=0.1385, over 3830013.69 frames. ], batch size: 52, lr: 2.49e-02, grad_scale: 16.0 +2023-04-01 00:41:21,856 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16818.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:41:22,529 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.096e+02 7.784e+02 9.820e+02 1.270e+03 2.923e+03, threshold=1.964e+03, percent-clipped=2.0 +2023-04-01 00:41:31,798 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 00:41:37,158 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-01 00:41:51,861 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16843.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:41:57,391 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16848.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:42:07,316 INFO [train.py:903] (3/4) Epoch 3, batch 3200, loss[loss=0.3116, simple_loss=0.3595, pruned_loss=0.1319, over 19723.00 frames. ], tot_loss[loss=0.3252, simple_loss=0.3726, pruned_loss=0.1389, over 3814103.57 frames. ], batch size: 51, lr: 2.48e-02, grad_scale: 8.0 +2023-04-01 00:42:39,216 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16882.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:42:43,621 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16886.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:43:07,061 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-04-01 00:43:08,568 INFO [train.py:903] (3/4) Epoch 3, batch 3250, loss[loss=0.3875, simple_loss=0.4208, pruned_loss=0.1771, over 18666.00 frames. ], tot_loss[loss=0.3268, simple_loss=0.3733, pruned_loss=0.1401, over 3807241.74 frames. ], batch size: 74, lr: 2.48e-02, grad_scale: 8.0 +2023-04-01 00:43:09,806 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16907.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:43:24,371 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.206e+02 8.507e+02 1.021e+03 1.288e+03 2.328e+03, threshold=2.042e+03, percent-clipped=1.0 +2023-04-01 00:43:29,325 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16924.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:44:09,582 INFO [train.py:903] (3/4) Epoch 3, batch 3300, loss[loss=0.2728, simple_loss=0.3218, pruned_loss=0.1119, over 19777.00 frames. ], tot_loss[loss=0.3253, simple_loss=0.3721, pruned_loss=0.1392, over 3810580.31 frames. ], batch size: 46, lr: 2.48e-02, grad_scale: 8.0 +2023-04-01 00:44:16,152 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 00:44:29,259 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6192, 2.9516, 2.0625, 2.3026, 2.2761, 1.9964, 0.4330, 2.1950], + device='cuda:3'), covar=tensor([0.0176, 0.0203, 0.0240, 0.0257, 0.0398, 0.0358, 0.0527, 0.0368], + device='cuda:3'), in_proj_covar=tensor([0.0244, 0.0240, 0.0233, 0.0257, 0.0311, 0.0257, 0.0245, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:45:10,101 INFO [train.py:903] (3/4) Epoch 3, batch 3350, loss[loss=0.3214, simple_loss=0.3845, pruned_loss=0.1292, over 19576.00 frames. ], tot_loss[loss=0.3256, simple_loss=0.3728, pruned_loss=0.1392, over 3812843.48 frames. ], batch size: 61, lr: 2.47e-02, grad_scale: 8.0 +2023-04-01 00:45:24,559 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.093e+02 7.847e+02 9.419e+02 1.175e+03 3.710e+03, threshold=1.884e+03, percent-clipped=5.0 +2023-04-01 00:45:47,671 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1486, 1.1189, 1.7372, 1.3152, 2.3296, 2.0683, 2.5280, 0.9566], + device='cuda:3'), covar=tensor([0.1473, 0.2380, 0.1168, 0.1310, 0.0903, 0.1115, 0.1072, 0.2103], + device='cuda:3'), in_proj_covar=tensor([0.0405, 0.0445, 0.0408, 0.0385, 0.0490, 0.0389, 0.0566, 0.0408], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:46:10,041 INFO [train.py:903] (3/4) Epoch 3, batch 3400, loss[loss=0.324, simple_loss=0.3796, pruned_loss=0.1342, over 19707.00 frames. ], tot_loss[loss=0.3269, simple_loss=0.3739, pruned_loss=0.14, over 3811140.30 frames. ], batch size: 59, lr: 2.47e-02, grad_scale: 8.0 +2023-04-01 00:47:08,457 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17104.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:47:12,304 INFO [train.py:903] (3/4) Epoch 3, batch 3450, loss[loss=0.3249, simple_loss=0.3692, pruned_loss=0.1403, over 19849.00 frames. ], tot_loss[loss=0.3265, simple_loss=0.3736, pruned_loss=0.1398, over 3808426.04 frames. ], batch size: 52, lr: 2.47e-02, grad_scale: 8.0 +2023-04-01 00:47:14,332 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 00:47:24,041 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5954, 2.3956, 2.2983, 3.3322, 2.3915, 3.9013, 3.4276, 2.1799], + device='cuda:3'), covar=tensor([0.1152, 0.0839, 0.0510, 0.0538, 0.1007, 0.0224, 0.0762, 0.0786], + device='cuda:3'), in_proj_covar=tensor([0.0459, 0.0435, 0.0433, 0.0576, 0.0512, 0.0343, 0.0536, 0.0430], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:47:28,177 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.582e+02 8.544e+02 1.014e+03 1.278e+03 1.988e+03, threshold=2.028e+03, percent-clipped=3.0 +2023-04-01 00:47:39,689 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17129.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:48:12,433 INFO [train.py:903] (3/4) Epoch 3, batch 3500, loss[loss=0.3384, simple_loss=0.3887, pruned_loss=0.1441, over 19574.00 frames. ], tot_loss[loss=0.3297, simple_loss=0.3755, pruned_loss=0.142, over 3801734.62 frames. ], batch size: 61, lr: 2.46e-02, grad_scale: 8.0 +2023-04-01 00:49:12,364 INFO [train.py:903] (3/4) Epoch 3, batch 3550, loss[loss=0.2947, simple_loss=0.3415, pruned_loss=0.124, over 19400.00 frames. ], tot_loss[loss=0.3287, simple_loss=0.3746, pruned_loss=0.1413, over 3805596.67 frames. ], batch size: 47, lr: 2.46e-02, grad_scale: 8.0 +2023-04-01 00:49:12,668 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17207.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:49:26,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.377e+02 8.429e+02 1.068e+03 1.302e+03 2.755e+03, threshold=2.137e+03, percent-clipped=4.0 +2023-04-01 00:49:40,883 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=17230.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:50:11,404 INFO [train.py:903] (3/4) Epoch 3, batch 3600, loss[loss=0.3031, simple_loss=0.344, pruned_loss=0.1311, over 19357.00 frames. ], tot_loss[loss=0.3296, simple_loss=0.3755, pruned_loss=0.1418, over 3799521.28 frames. ], batch size: 47, lr: 2.46e-02, grad_scale: 8.0 +2023-04-01 00:50:24,843 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=17268.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:50:42,786 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6118, 1.3340, 1.2124, 1.5918, 1.3929, 1.3862, 1.2308, 1.6436], + device='cuda:3'), covar=tensor([0.0938, 0.1372, 0.1476, 0.0980, 0.1168, 0.0681, 0.1175, 0.0758], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0379, 0.0286, 0.0253, 0.0320, 0.0265, 0.0277, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 00:51:11,853 INFO [train.py:903] (3/4) Epoch 3, batch 3650, loss[loss=0.2712, simple_loss=0.3235, pruned_loss=0.1095, over 19759.00 frames. ], tot_loss[loss=0.3275, simple_loss=0.3741, pruned_loss=0.1405, over 3804177.26 frames. ], batch size: 46, lr: 2.45e-02, grad_scale: 8.0 +2023-04-01 00:51:27,519 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.083e+02 7.894e+02 9.345e+02 1.119e+03 1.949e+03, threshold=1.869e+03, percent-clipped=0.0 +2023-04-01 00:51:57,303 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17345.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:52:12,710 INFO [train.py:903] (3/4) Epoch 3, batch 3700, loss[loss=0.2746, simple_loss=0.3174, pruned_loss=0.1159, over 19755.00 frames. ], tot_loss[loss=0.3269, simple_loss=0.3732, pruned_loss=0.1403, over 3813818.37 frames. ], batch size: 45, lr: 2.45e-02, grad_scale: 8.0 +2023-04-01 00:52:18,213 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5449, 4.0815, 2.4452, 3.7280, 1.1368, 3.9545, 3.7480, 3.9163], + device='cuda:3'), covar=tensor([0.0516, 0.1186, 0.1955, 0.0586, 0.3856, 0.0725, 0.0590, 0.0795], + device='cuda:3'), in_proj_covar=tensor([0.0320, 0.0299, 0.0336, 0.0276, 0.0350, 0.0298, 0.0248, 0.0291], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:52:32,171 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-04-01 00:52:39,373 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17380.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:52:41,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5201, 1.1964, 1.5777, 0.8765, 2.6221, 2.8813, 2.7373, 3.0648], + device='cuda:3'), covar=tensor([0.1128, 0.2413, 0.2325, 0.1899, 0.0326, 0.0129, 0.0231, 0.0115], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0275, 0.0324, 0.0269, 0.0191, 0.0107, 0.0199, 0.0113], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 00:52:42,863 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17383.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:53:05,594 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17401.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:53:12,847 INFO [train.py:903] (3/4) Epoch 3, batch 3750, loss[loss=0.3122, simple_loss=0.3713, pruned_loss=0.1265, over 19609.00 frames. ], tot_loss[loss=0.3272, simple_loss=0.3732, pruned_loss=0.1406, over 3832389.58 frames. ], batch size: 57, lr: 2.45e-02, grad_scale: 8.0 +2023-04-01 00:53:27,567 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.653e+02 9.192e+02 1.060e+03 1.489e+03 3.397e+03, threshold=2.120e+03, percent-clipped=7.0 +2023-04-01 00:54:12,635 INFO [train.py:903] (3/4) Epoch 3, batch 3800, loss[loss=0.2893, simple_loss=0.3492, pruned_loss=0.1147, over 19840.00 frames. ], tot_loss[loss=0.3255, simple_loss=0.3721, pruned_loss=0.1395, over 3844601.34 frames. ], batch size: 52, lr: 2.44e-02, grad_scale: 8.0 +2023-04-01 00:54:23,098 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6681, 1.3619, 2.0137, 1.7137, 2.8218, 2.5917, 2.9024, 1.7160], + device='cuda:3'), covar=tensor([0.1062, 0.2018, 0.1092, 0.0982, 0.0656, 0.0761, 0.0777, 0.1511], + device='cuda:3'), in_proj_covar=tensor([0.0409, 0.0450, 0.0411, 0.0382, 0.0485, 0.0397, 0.0573, 0.0401], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 00:54:45,633 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 00:55:11,732 INFO [train.py:903] (3/4) Epoch 3, batch 3850, loss[loss=0.3907, simple_loss=0.4231, pruned_loss=0.1791, over 13198.00 frames. ], tot_loss[loss=0.3269, simple_loss=0.3733, pruned_loss=0.1402, over 3833709.29 frames. ], batch size: 137, lr: 2.44e-02, grad_scale: 8.0 +2023-04-01 00:55:28,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.037e+02 7.944e+02 9.720e+02 1.209e+03 3.103e+03, threshold=1.944e+03, percent-clipped=2.0 +2023-04-01 00:55:40,055 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3333, 1.0664, 1.0925, 1.5900, 1.3050, 1.4384, 1.7597, 1.3538], + device='cuda:3'), covar=tensor([0.1076, 0.1456, 0.1409, 0.1089, 0.1203, 0.1169, 0.1013, 0.0933], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0280, 0.0272, 0.0313, 0.0319, 0.0258, 0.0285, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 00:56:04,059 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=17551.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:56:05,718 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.53 vs. limit=5.0 +2023-04-01 00:56:12,763 INFO [train.py:903] (3/4) Epoch 3, batch 3900, loss[loss=0.2805, simple_loss=0.3412, pruned_loss=0.1099, over 19764.00 frames. ], tot_loss[loss=0.3267, simple_loss=0.3733, pruned_loss=0.1401, over 3825355.44 frames. ], batch size: 51, lr: 2.44e-02, grad_scale: 8.0 +2023-04-01 00:57:05,251 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17601.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:57:11,566 INFO [train.py:903] (3/4) Epoch 3, batch 3950, loss[loss=0.3431, simple_loss=0.39, pruned_loss=0.1481, over 19519.00 frames. ], tot_loss[loss=0.3271, simple_loss=0.3739, pruned_loss=0.1402, over 3827621.92 frames. ], batch size: 64, lr: 2.44e-02, grad_scale: 8.0 +2023-04-01 00:57:18,165 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 00:57:27,261 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.216e+02 7.211e+02 9.089e+02 1.148e+03 2.193e+03, threshold=1.818e+03, percent-clipped=2.0 +2023-04-01 00:57:34,470 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17626.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 00:57:51,670 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17639.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:58:12,126 INFO [train.py:903] (3/4) Epoch 3, batch 4000, loss[loss=0.2803, simple_loss=0.3365, pruned_loss=0.112, over 19471.00 frames. ], tot_loss[loss=0.3233, simple_loss=0.3706, pruned_loss=0.138, over 3834079.64 frames. ], batch size: 49, lr: 2.43e-02, grad_scale: 8.0 +2023-04-01 00:58:20,416 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17664.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:58:22,675 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17666.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:58:50,785 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4209, 1.0965, 1.4468, 1.0257, 2.5639, 3.0875, 3.0636, 3.3605], + device='cuda:3'), covar=tensor([0.1326, 0.2602, 0.2685, 0.1945, 0.0410, 0.0118, 0.0210, 0.0100], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0279, 0.0323, 0.0272, 0.0195, 0.0107, 0.0201, 0.0113], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 00:58:59,484 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 00:59:11,487 INFO [train.py:903] (3/4) Epoch 3, batch 4050, loss[loss=0.3921, simple_loss=0.4201, pruned_loss=0.182, over 13854.00 frames. ], tot_loss[loss=0.3239, simple_loss=0.3715, pruned_loss=0.1381, over 3815990.96 frames. ], batch size: 136, lr: 2.43e-02, grad_scale: 8.0 +2023-04-01 00:59:25,594 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17718.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:59:28,914 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.065e+02 7.398e+02 9.459e+02 1.250e+03 4.446e+03, threshold=1.892e+03, percent-clipped=10.0 +2023-04-01 00:59:32,714 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=17724.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 00:59:37,205 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1585, 2.8607, 2.0572, 2.7279, 1.0576, 2.7863, 2.5901, 2.7340], + device='cuda:3'), covar=tensor([0.0940, 0.1429, 0.1982, 0.0859, 0.3556, 0.1080, 0.0819, 0.1085], + device='cuda:3'), in_proj_covar=tensor([0.0319, 0.0296, 0.0336, 0.0273, 0.0349, 0.0297, 0.0243, 0.0286], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 00:59:57,742 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=17745.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:00:01,032 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17748.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:00:12,214 INFO [train.py:903] (3/4) Epoch 3, batch 4100, loss[loss=0.3553, simple_loss=0.3874, pruned_loss=0.1616, over 19658.00 frames. ], tot_loss[loss=0.3254, simple_loss=0.3728, pruned_loss=0.139, over 3820766.75 frames. ], batch size: 60, lr: 2.43e-02, grad_scale: 4.0 +2023-04-01 01:00:48,445 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 01:01:11,682 INFO [train.py:903] (3/4) Epoch 3, batch 4150, loss[loss=0.2855, simple_loss=0.3288, pruned_loss=0.1211, over 19719.00 frames. ], tot_loss[loss=0.3237, simple_loss=0.3714, pruned_loss=0.138, over 3827370.19 frames. ], batch size: 45, lr: 2.42e-02, grad_scale: 4.0 +2023-04-01 01:01:28,531 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.446e+02 7.784e+02 9.706e+02 1.186e+03 3.618e+03, threshold=1.941e+03, percent-clipped=3.0 +2023-04-01 01:01:49,977 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17839.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:02:10,744 INFO [train.py:903] (3/4) Epoch 3, batch 4200, loss[loss=0.3755, simple_loss=0.4054, pruned_loss=0.1728, over 19775.00 frames. ], tot_loss[loss=0.3228, simple_loss=0.3709, pruned_loss=0.1373, over 3823372.14 frames. ], batch size: 56, lr: 2.42e-02, grad_scale: 4.0 +2023-04-01 01:02:14,260 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17860.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:02:14,914 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 01:03:09,269 INFO [train.py:903] (3/4) Epoch 3, batch 4250, loss[loss=0.3048, simple_loss=0.3428, pruned_loss=0.1334, over 19769.00 frames. ], tot_loss[loss=0.3244, simple_loss=0.3721, pruned_loss=0.1383, over 3823045.43 frames. ], batch size: 47, lr: 2.42e-02, grad_scale: 4.0 +2023-04-01 01:03:26,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.782e+02 8.334e+02 9.808e+02 1.259e+03 2.577e+03, threshold=1.962e+03, percent-clipped=5.0 +2023-04-01 01:03:26,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 01:03:28,299 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17922.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:03:37,999 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 01:03:57,696 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17947.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:04:08,380 INFO [train.py:903] (3/4) Epoch 3, batch 4300, loss[loss=0.3549, simple_loss=0.4018, pruned_loss=0.1541, over 17477.00 frames. ], tot_loss[loss=0.3244, simple_loss=0.3721, pruned_loss=0.1383, over 3830969.44 frames. ], batch size: 101, lr: 2.41e-02, grad_scale: 4.0 +2023-04-01 01:04:35,078 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17978.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:05:06,032 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 01:05:10,739 INFO [train.py:903] (3/4) Epoch 3, batch 4350, loss[loss=0.3184, simple_loss=0.3808, pruned_loss=0.128, over 19687.00 frames. ], tot_loss[loss=0.3232, simple_loss=0.3709, pruned_loss=0.1378, over 3834125.97 frames. ], batch size: 59, lr: 2.41e-02, grad_scale: 4.0 +2023-04-01 01:05:27,030 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.929e+02 7.530e+02 9.485e+02 1.282e+03 2.824e+03, threshold=1.897e+03, percent-clipped=4.0 +2023-04-01 01:06:11,119 INFO [train.py:903] (3/4) Epoch 3, batch 4400, loss[loss=0.3617, simple_loss=0.4051, pruned_loss=0.1591, over 19606.00 frames. ], tot_loss[loss=0.3238, simple_loss=0.3714, pruned_loss=0.1381, over 3821624.49 frames. ], batch size: 57, lr: 2.41e-02, grad_scale: 8.0 +2023-04-01 01:06:17,200 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18062.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:06:32,865 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 01:06:43,600 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 01:06:51,966 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.97 vs. limit=5.0 +2023-04-01 01:06:53,814 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18092.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:06:57,486 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18095.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:07:10,941 INFO [train.py:903] (3/4) Epoch 3, batch 4450, loss[loss=0.3285, simple_loss=0.3748, pruned_loss=0.1411, over 18741.00 frames. ], tot_loss[loss=0.3219, simple_loss=0.3701, pruned_loss=0.1368, over 3824753.64 frames. ], batch size: 74, lr: 2.40e-02, grad_scale: 8.0 +2023-04-01 01:07:21,294 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18116.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:07:27,544 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18120.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:07:28,238 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.891e+02 7.625e+02 9.248e+02 1.171e+03 2.408e+03, threshold=1.850e+03, percent-clipped=4.0 +2023-04-01 01:07:53,390 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18141.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:08:11,164 INFO [train.py:903] (3/4) Epoch 3, batch 4500, loss[loss=0.3336, simple_loss=0.3653, pruned_loss=0.151, over 19766.00 frames. ], tot_loss[loss=0.3234, simple_loss=0.3713, pruned_loss=0.1377, over 3829294.80 frames. ], batch size: 45, lr: 2.40e-02, grad_scale: 4.0 +2023-04-01 01:08:31,582 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0447, 1.7624, 2.0937, 2.5506, 4.4726, 1.5646, 2.3693, 4.4448], + device='cuda:3'), covar=tensor([0.0183, 0.2032, 0.1897, 0.1150, 0.0332, 0.1770, 0.1023, 0.0286], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0304, 0.0295, 0.0279, 0.0285, 0.0321, 0.0272, 0.0283], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:08:37,477 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18177.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:08:54,083 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18192.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:08:55,175 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18193.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:09:11,759 INFO [train.py:903] (3/4) Epoch 3, batch 4550, loss[loss=0.2622, simple_loss=0.3259, pruned_loss=0.09921, over 19606.00 frames. ], tot_loss[loss=0.3232, simple_loss=0.3709, pruned_loss=0.1378, over 3822761.50 frames. ], batch size: 50, lr: 2.40e-02, grad_scale: 4.0 +2023-04-01 01:09:12,106 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18207.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:09:18,481 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 01:09:25,766 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-01 01:09:29,435 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.087e+02 7.616e+02 9.596e+02 1.201e+03 2.125e+03, threshold=1.919e+03, percent-clipped=4.0 +2023-04-01 01:09:42,206 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 01:09:48,932 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2244, 1.2278, 1.8147, 1.3171, 2.3258, 2.1556, 2.5244, 0.9597], + device='cuda:3'), covar=tensor([0.1409, 0.2266, 0.1167, 0.1340, 0.0926, 0.1109, 0.0970, 0.2033], + device='cuda:3'), in_proj_covar=tensor([0.0411, 0.0459, 0.0416, 0.0389, 0.0502, 0.0404, 0.0581, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:10:07,226 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 01:10:11,980 INFO [train.py:903] (3/4) Epoch 3, batch 4600, loss[loss=0.3138, simple_loss=0.3609, pruned_loss=0.1334, over 19475.00 frames. ], tot_loss[loss=0.3255, simple_loss=0.3724, pruned_loss=0.1393, over 3813503.92 frames. ], batch size: 49, lr: 2.39e-02, grad_scale: 4.0 +2023-04-01 01:11:11,635 INFO [train.py:903] (3/4) Epoch 3, batch 4650, loss[loss=0.348, simple_loss=0.3969, pruned_loss=0.1496, over 17176.00 frames. ], tot_loss[loss=0.3236, simple_loss=0.3708, pruned_loss=0.1382, over 3822846.97 frames. ], batch size: 101, lr: 2.39e-02, grad_scale: 4.0 +2023-04-01 01:11:27,781 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 01:11:28,861 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.638e+02 7.851e+02 9.796e+02 1.308e+03 3.825e+03, threshold=1.959e+03, percent-clipped=6.0 +2023-04-01 01:11:29,088 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18322.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:11:38,751 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 01:12:10,878 INFO [train.py:903] (3/4) Epoch 3, batch 4700, loss[loss=0.2881, simple_loss=0.3338, pruned_loss=0.1212, over 19781.00 frames. ], tot_loss[loss=0.3236, simple_loss=0.3707, pruned_loss=0.1383, over 3813016.37 frames. ], batch size: 47, lr: 2.39e-02, grad_scale: 4.0 +2023-04-01 01:12:33,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 01:12:54,111 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9619, 1.9787, 1.4190, 1.3127, 1.2585, 1.5820, 0.2371, 0.8193], + device='cuda:3'), covar=tensor([0.0273, 0.0270, 0.0217, 0.0283, 0.0578, 0.0314, 0.0511, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0246, 0.0246, 0.0244, 0.0266, 0.0322, 0.0257, 0.0251, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:12:59,735 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6491, 1.2455, 1.6585, 2.0381, 3.2794, 1.3408, 1.9857, 3.3226], + device='cuda:3'), covar=tensor([0.0307, 0.2353, 0.2136, 0.1233, 0.0429, 0.1954, 0.1166, 0.0377], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0299, 0.0296, 0.0273, 0.0280, 0.0316, 0.0266, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:13:13,560 INFO [train.py:903] (3/4) Epoch 3, batch 4750, loss[loss=0.3566, simple_loss=0.4026, pruned_loss=0.1553, over 19707.00 frames. ], tot_loss[loss=0.3235, simple_loss=0.3709, pruned_loss=0.138, over 3814386.55 frames. ], batch size: 59, lr: 2.39e-02, grad_scale: 4.0 +2023-04-01 01:13:19,797 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-04-01 01:13:22,628 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1446, 1.2022, 1.9552, 1.3047, 2.3622, 2.3305, 2.6310, 0.9672], + device='cuda:3'), covar=tensor([0.1578, 0.2517, 0.1209, 0.1431, 0.1163, 0.1164, 0.1190, 0.2322], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0464, 0.0420, 0.0395, 0.0505, 0.0403, 0.0582, 0.0417], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:13:30,998 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-01 01:13:31,290 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.455e+02 7.810e+02 9.309e+02 1.222e+03 2.382e+03, threshold=1.862e+03, percent-clipped=4.0 +2023-04-01 01:13:32,812 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6853, 1.2302, 1.6265, 1.9892, 3.3119, 1.6508, 2.2416, 3.3352], + device='cuda:3'), covar=tensor([0.0335, 0.2739, 0.2287, 0.1472, 0.0462, 0.2105, 0.0976, 0.0462], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0301, 0.0297, 0.0274, 0.0281, 0.0320, 0.0271, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:13:44,297 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18433.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:13:48,013 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.83 vs. limit=2.0 +2023-04-01 01:13:48,785 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18437.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:14:15,105 INFO [train.py:903] (3/4) Epoch 3, batch 4800, loss[loss=0.28, simple_loss=0.3225, pruned_loss=0.1188, over 19760.00 frames. ], tot_loss[loss=0.3244, simple_loss=0.3721, pruned_loss=0.1383, over 3802030.85 frames. ], batch size: 47, lr: 2.38e-02, grad_scale: 8.0 +2023-04-01 01:14:16,617 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18458.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:14:22,308 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18463.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:14:53,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18488.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:15:16,099 INFO [train.py:903] (3/4) Epoch 3, batch 4850, loss[loss=0.4558, simple_loss=0.4543, pruned_loss=0.2287, over 13075.00 frames. ], tot_loss[loss=0.3227, simple_loss=0.3709, pruned_loss=0.1373, over 3797060.48 frames. ], batch size: 135, lr: 2.38e-02, grad_scale: 8.0 +2023-04-01 01:15:17,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18508.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:15:34,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.685e+02 7.295e+02 9.130e+02 1.063e+03 1.681e+03, threshold=1.826e+03, percent-clipped=0.0 +2023-04-01 01:15:38,190 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 01:15:40,625 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18526.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:15:52,403 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18536.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:15:53,582 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18537.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:16:00,114 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 01:16:04,601 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 01:16:05,774 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 01:16:14,592 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 01:16:15,641 INFO [train.py:903] (3/4) Epoch 3, batch 4900, loss[loss=0.3823, simple_loss=0.416, pruned_loss=0.1742, over 19717.00 frames. ], tot_loss[loss=0.3208, simple_loss=0.3693, pruned_loss=0.1362, over 3803579.36 frames. ], batch size: 59, lr: 2.38e-02, grad_scale: 8.0 +2023-04-01 01:16:34,644 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 01:17:14,805 INFO [train.py:903] (3/4) Epoch 3, batch 4950, loss[loss=0.3766, simple_loss=0.4076, pruned_loss=0.1728, over 18098.00 frames. ], tot_loss[loss=0.3233, simple_loss=0.3711, pruned_loss=0.1377, over 3797598.19 frames. ], batch size: 83, lr: 2.37e-02, grad_scale: 4.0 +2023-04-01 01:17:30,524 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 01:17:34,978 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.825e+02 8.688e+02 1.059e+03 1.337e+03 3.400e+03, threshold=2.119e+03, percent-clipped=10.0 +2023-04-01 01:17:52,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 01:17:54,768 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3775, 0.9254, 1.2149, 1.2443, 2.0697, 1.0854, 1.6480, 1.9926], + device='cuda:3'), covar=tensor([0.0583, 0.2721, 0.2521, 0.1492, 0.0689, 0.1955, 0.1038, 0.0714], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0303, 0.0298, 0.0276, 0.0285, 0.0319, 0.0271, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:18:09,845 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18651.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:18:10,978 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18652.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:18:17,088 INFO [train.py:903] (3/4) Epoch 3, batch 5000, loss[loss=0.3439, simple_loss=0.4003, pruned_loss=0.1437, over 19415.00 frames. ], tot_loss[loss=0.3232, simple_loss=0.3713, pruned_loss=0.1376, over 3799562.16 frames. ], batch size: 70, lr: 2.37e-02, grad_scale: 4.0 +2023-04-01 01:18:21,588 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 01:18:32,449 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 01:18:59,281 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18692.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:19:00,356 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18693.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:19:16,428 INFO [train.py:903] (3/4) Epoch 3, batch 5050, loss[loss=0.3245, simple_loss=0.3699, pruned_loss=0.1395, over 19540.00 frames. ], tot_loss[loss=0.3235, simple_loss=0.3714, pruned_loss=0.1378, over 3810924.89 frames. ], batch size: 54, lr: 2.37e-02, grad_scale: 4.0 +2023-04-01 01:19:29,230 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18718.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:19:35,293 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.108e+02 7.776e+02 1.028e+03 1.229e+03 3.550e+03, threshold=2.057e+03, percent-clipped=2.0 +2023-04-01 01:19:48,441 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 01:20:09,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0779, 5.3600, 3.1452, 4.7200, 1.4060, 5.1888, 5.2109, 5.5203], + device='cuda:3'), covar=tensor([0.0421, 0.1036, 0.1585, 0.0665, 0.3597, 0.0588, 0.0544, 0.0580], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0293, 0.0333, 0.0278, 0.0343, 0.0290, 0.0247, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:20:16,938 INFO [train.py:903] (3/4) Epoch 3, batch 5100, loss[loss=0.2851, simple_loss=0.3425, pruned_loss=0.1138, over 19609.00 frames. ], tot_loss[loss=0.3223, simple_loss=0.3706, pruned_loss=0.137, over 3815406.47 frames. ], batch size: 50, lr: 2.36e-02, grad_scale: 4.0 +2023-04-01 01:20:24,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 01:20:27,864 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 01:20:32,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 01:20:36,148 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8600, 4.1578, 4.5044, 4.4361, 1.4097, 4.0554, 3.7565, 3.9796], + device='cuda:3'), covar=tensor([0.0614, 0.0512, 0.0412, 0.0325, 0.3343, 0.0276, 0.0376, 0.0945], + device='cuda:3'), in_proj_covar=tensor([0.0391, 0.0361, 0.0495, 0.0385, 0.0496, 0.0261, 0.0324, 0.0469], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 01:21:01,142 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9173, 1.9734, 1.6854, 2.8402, 1.8999, 2.9577, 2.6254, 1.8017], + device='cuda:3'), covar=tensor([0.1209, 0.0878, 0.0628, 0.0521, 0.1040, 0.0260, 0.0898, 0.0855], + device='cuda:3'), in_proj_covar=tensor([0.0493, 0.0457, 0.0461, 0.0617, 0.0535, 0.0372, 0.0558, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:21:19,959 INFO [train.py:903] (3/4) Epoch 3, batch 5150, loss[loss=0.3329, simple_loss=0.3898, pruned_loss=0.138, over 19490.00 frames. ], tot_loss[loss=0.3197, simple_loss=0.369, pruned_loss=0.1352, over 3822929.17 frames. ], batch size: 64, lr: 2.36e-02, grad_scale: 4.0 +2023-04-01 01:21:22,518 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9073, 1.7827, 1.7303, 2.5244, 2.0275, 2.5122, 2.1716, 1.5717], + device='cuda:3'), covar=tensor([0.0915, 0.0769, 0.0501, 0.0394, 0.0730, 0.0250, 0.0725, 0.0747], + device='cuda:3'), in_proj_covar=tensor([0.0488, 0.0458, 0.0460, 0.0615, 0.0534, 0.0375, 0.0559, 0.0456], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:21:31,907 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 01:21:40,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.229e+02 6.924e+02 8.047e+02 1.080e+03 1.932e+03, threshold=1.609e+03, percent-clipped=0.0 +2023-04-01 01:22:06,986 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 01:22:14,770 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18852.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:22:20,358 INFO [train.py:903] (3/4) Epoch 3, batch 5200, loss[loss=0.4261, simple_loss=0.4341, pruned_loss=0.209, over 13480.00 frames. ], tot_loss[loss=0.3177, simple_loss=0.3678, pruned_loss=0.1338, over 3816501.27 frames. ], batch size: 136, lr: 2.36e-02, grad_scale: 8.0 +2023-04-01 01:22:37,384 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 01:22:37,518 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=18870.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:23:21,232 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 01:23:23,236 INFO [train.py:903] (3/4) Epoch 3, batch 5250, loss[loss=0.2733, simple_loss=0.3343, pruned_loss=0.1061, over 19673.00 frames. ], tot_loss[loss=0.3168, simple_loss=0.3666, pruned_loss=0.1335, over 3812761.98 frames. ], batch size: 53, lr: 2.36e-02, grad_scale: 4.0 +2023-04-01 01:23:23,700 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18907.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:23:24,877 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18908.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:23:42,381 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.727e+02 7.388e+02 9.793e+02 1.246e+03 4.620e+03, threshold=1.959e+03, percent-clipped=9.0 +2023-04-01 01:23:45,180 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-01 01:23:53,545 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18932.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:23:54,590 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18933.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:24:12,783 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18948.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:24:22,206 INFO [train.py:903] (3/4) Epoch 3, batch 5300, loss[loss=0.3663, simple_loss=0.399, pruned_loss=0.1668, over 13532.00 frames. ], tot_loss[loss=0.3189, simple_loss=0.3678, pruned_loss=0.135, over 3805104.44 frames. ], batch size: 136, lr: 2.35e-02, grad_scale: 4.0 +2023-04-01 01:24:35,242 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18967.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:24:39,531 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 01:24:56,497 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18985.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:25:22,328 INFO [train.py:903] (3/4) Epoch 3, batch 5350, loss[loss=0.2863, simple_loss=0.3404, pruned_loss=0.1161, over 19615.00 frames. ], tot_loss[loss=0.3201, simple_loss=0.3686, pruned_loss=0.1358, over 3796551.57 frames. ], batch size: 50, lr: 2.35e-02, grad_scale: 4.0 +2023-04-01 01:25:28,792 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-04-01 01:25:39,625 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8969, 2.1090, 2.0397, 2.5342, 4.3898, 1.7119, 2.1747, 4.2359], + device='cuda:3'), covar=tensor([0.0216, 0.2008, 0.2209, 0.1301, 0.0346, 0.1838, 0.1143, 0.0310], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0307, 0.0303, 0.0279, 0.0292, 0.0322, 0.0275, 0.0287], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:25:42,787 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.207e+02 7.837e+02 1.011e+03 1.314e+03 3.062e+03, threshold=2.023e+03, percent-clipped=6.0 +2023-04-01 01:25:52,475 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2894, 2.3017, 1.5042, 1.5558, 1.8154, 0.9797, 1.1289, 1.8978], + device='cuda:3'), covar=tensor([0.0957, 0.0432, 0.1023, 0.0564, 0.0659, 0.1242, 0.0833, 0.0419], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0238, 0.0311, 0.0253, 0.0219, 0.0309, 0.0272, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:25:55,459 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 01:25:56,694 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=19036.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:26:20,416 INFO [train.py:903] (3/4) Epoch 3, batch 5400, loss[loss=0.3523, simple_loss=0.395, pruned_loss=0.1548, over 19636.00 frames. ], tot_loss[loss=0.3201, simple_loss=0.3688, pruned_loss=0.1357, over 3805680.88 frames. ], batch size: 57, lr: 2.35e-02, grad_scale: 4.0 +2023-04-01 01:26:22,879 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1501, 1.1259, 1.6690, 1.3389, 2.3110, 2.1908, 2.3717, 0.8011], + device='cuda:3'), covar=tensor([0.1599, 0.2527, 0.1288, 0.1383, 0.0914, 0.1079, 0.1011, 0.2208], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0465, 0.0424, 0.0393, 0.0498, 0.0406, 0.0575, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:26:27,106 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3959, 0.9532, 1.4224, 1.3273, 2.2047, 1.1348, 1.8855, 2.0925], + device='cuda:3'), covar=tensor([0.0546, 0.2332, 0.1913, 0.1159, 0.0538, 0.1497, 0.0701, 0.0589], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0310, 0.0303, 0.0279, 0.0293, 0.0322, 0.0278, 0.0288], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:26:39,647 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-01 01:27:22,187 INFO [train.py:903] (3/4) Epoch 3, batch 5450, loss[loss=0.3479, simple_loss=0.3882, pruned_loss=0.1538, over 13683.00 frames. ], tot_loss[loss=0.3202, simple_loss=0.3687, pruned_loss=0.1359, over 3796627.33 frames. ], batch size: 137, lr: 2.34e-02, grad_scale: 4.0 +2023-04-01 01:27:39,366 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5111, 1.2193, 1.1200, 1.7411, 1.3609, 1.5630, 1.7002, 1.5190], + device='cuda:3'), covar=tensor([0.0783, 0.1147, 0.1270, 0.0902, 0.1005, 0.0902, 0.0926, 0.0725], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0280, 0.0267, 0.0307, 0.0309, 0.0257, 0.0281, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 01:27:41,224 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.092e+02 7.941e+02 9.480e+02 1.159e+03 2.761e+03, threshold=1.896e+03, percent-clipped=1.0 +2023-04-01 01:28:10,034 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19147.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:28:14,880 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19151.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:28:21,123 INFO [train.py:903] (3/4) Epoch 3, batch 5500, loss[loss=0.3351, simple_loss=0.3896, pruned_loss=0.1403, over 19528.00 frames. ], tot_loss[loss=0.3209, simple_loss=0.3696, pruned_loss=0.1361, over 3799781.48 frames. ], batch size: 56, lr: 2.34e-02, grad_scale: 4.0 +2023-04-01 01:28:31,565 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0745, 1.5659, 1.9429, 1.5991, 2.8811, 4.3002, 4.2755, 4.6899], + device='cuda:3'), covar=tensor([0.0945, 0.2236, 0.2166, 0.1532, 0.0407, 0.0090, 0.0145, 0.0071], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0279, 0.0323, 0.0263, 0.0197, 0.0108, 0.0200, 0.0117], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 01:28:33,448 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2641, 1.3542, 1.0756, 1.0323, 0.9726, 1.1409, 0.2473, 0.6170], + device='cuda:3'), covar=tensor([0.0185, 0.0155, 0.0109, 0.0132, 0.0282, 0.0176, 0.0348, 0.0282], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0246, 0.0250, 0.0268, 0.0328, 0.0262, 0.0260, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:28:34,650 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9410, 1.8392, 1.7034, 2.7300, 1.7423, 2.7251, 2.5347, 1.5742], + device='cuda:3'), covar=tensor([0.1037, 0.0820, 0.0490, 0.0499, 0.0991, 0.0287, 0.0793, 0.0821], + device='cuda:3'), in_proj_covar=tensor([0.0498, 0.0467, 0.0469, 0.0624, 0.0543, 0.0385, 0.0567, 0.0463], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:28:45,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 01:29:20,073 INFO [train.py:903] (3/4) Epoch 3, batch 5550, loss[loss=0.3576, simple_loss=0.3935, pruned_loss=0.1608, over 17502.00 frames. ], tot_loss[loss=0.3216, simple_loss=0.37, pruned_loss=0.1366, over 3784847.07 frames. ], batch size: 101, lr: 2.34e-02, grad_scale: 4.0 +2023-04-01 01:29:27,432 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 01:29:41,395 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19223.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:29:42,151 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.073e+02 7.947e+02 9.800e+02 1.342e+03 2.993e+03, threshold=1.960e+03, percent-clipped=6.0 +2023-04-01 01:29:55,073 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6387, 3.9270, 2.6060, 3.3227, 3.2997, 2.2893, 2.1634, 2.3561], + device='cuda:3'), covar=tensor([0.0874, 0.0265, 0.0729, 0.0445, 0.0553, 0.0902, 0.0669, 0.0602], + device='cuda:3'), in_proj_covar=tensor([0.0284, 0.0240, 0.0313, 0.0261, 0.0221, 0.0312, 0.0279, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:30:02,031 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19241.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:30:10,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19248.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:30:15,626 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 01:30:21,457 INFO [train.py:903] (3/4) Epoch 3, batch 5600, loss[loss=0.2865, simple_loss=0.3479, pruned_loss=0.1125, over 19764.00 frames. ], tot_loss[loss=0.3189, simple_loss=0.3676, pruned_loss=0.1351, over 3785396.90 frames. ], batch size: 54, lr: 2.34e-02, grad_scale: 8.0 +2023-04-01 01:30:26,196 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-04-01 01:30:33,704 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19266.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:31:04,003 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=19292.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:31:22,075 INFO [train.py:903] (3/4) Epoch 3, batch 5650, loss[loss=0.3706, simple_loss=0.411, pruned_loss=0.1651, over 19670.00 frames. ], tot_loss[loss=0.3197, simple_loss=0.3684, pruned_loss=0.1354, over 3797997.39 frames. ], batch size: 58, lr: 2.33e-02, grad_scale: 8.0 +2023-04-01 01:31:29,265 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6218, 1.8659, 1.8749, 2.9306, 2.3806, 2.3478, 2.1137, 2.8037], + device='cuda:3'), covar=tensor([0.0621, 0.1589, 0.1089, 0.0590, 0.0990, 0.0404, 0.0749, 0.0472], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0375, 0.0282, 0.0250, 0.0308, 0.0261, 0.0267, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:31:41,039 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.993e+02 7.403e+02 9.102e+02 1.185e+03 3.385e+03, threshold=1.820e+03, percent-clipped=3.0 +2023-04-01 01:32:09,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 01:32:21,815 INFO [train.py:903] (3/4) Epoch 3, batch 5700, loss[loss=0.3269, simple_loss=0.3651, pruned_loss=0.1444, over 19749.00 frames. ], tot_loss[loss=0.3208, simple_loss=0.3692, pruned_loss=0.1363, over 3798756.65 frames. ], batch size: 47, lr: 2.33e-02, grad_scale: 8.0 +2023-04-01 01:33:22,522 INFO [train.py:903] (3/4) Epoch 3, batch 5750, loss[loss=0.298, simple_loss=0.356, pruned_loss=0.12, over 19771.00 frames. ], tot_loss[loss=0.3173, simple_loss=0.3667, pruned_loss=0.1339, over 3815195.32 frames. ], batch size: 54, lr: 2.33e-02, grad_scale: 8.0 +2023-04-01 01:33:22,853 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19407.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:33:22,920 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19407.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:33:25,751 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 01:33:30,125 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19412.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:33:34,052 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 01:33:39,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 01:33:43,561 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.926e+02 7.525e+02 9.538e+02 1.231e+03 3.556e+03, threshold=1.908e+03, percent-clipped=6.0 +2023-04-01 01:33:49,419 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1588, 1.1278, 1.5498, 1.3242, 2.2519, 1.8533, 2.2771, 0.9070], + device='cuda:3'), covar=tensor([0.1496, 0.2487, 0.1322, 0.1292, 0.0911, 0.1235, 0.0974, 0.2239], + device='cuda:3'), in_proj_covar=tensor([0.0409, 0.0462, 0.0416, 0.0388, 0.0502, 0.0406, 0.0569, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 01:33:53,867 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19432.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:34:22,893 INFO [train.py:903] (3/4) Epoch 3, batch 5800, loss[loss=0.3407, simple_loss=0.3886, pruned_loss=0.1464, over 18254.00 frames. ], tot_loss[loss=0.3185, simple_loss=0.3677, pruned_loss=0.1347, over 3808640.25 frames. ], batch size: 83, lr: 2.32e-02, grad_scale: 4.0 +2023-04-01 01:35:03,997 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=19491.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:35:07,671 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.73 vs. limit=5.0 +2023-04-01 01:35:23,142 INFO [train.py:903] (3/4) Epoch 3, batch 5850, loss[loss=0.301, simple_loss=0.3564, pruned_loss=0.1228, over 19747.00 frames. ], tot_loss[loss=0.32, simple_loss=0.369, pruned_loss=0.1355, over 3801862.26 frames. ], batch size: 63, lr: 2.32e-02, grad_scale: 4.0 +2023-04-01 01:35:43,379 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.468e+02 8.354e+02 1.035e+03 1.319e+03 5.609e+03, threshold=2.070e+03, percent-clipped=8.0 +2023-04-01 01:35:54,724 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1008, 2.1113, 1.3616, 1.5656, 1.3774, 1.7471, 0.1936, 0.8077], + device='cuda:3'), covar=tensor([0.0198, 0.0196, 0.0166, 0.0195, 0.0469, 0.0211, 0.0407, 0.0351], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0250, 0.0251, 0.0278, 0.0332, 0.0264, 0.0257, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:36:23,588 INFO [train.py:903] (3/4) Epoch 3, batch 5900, loss[loss=0.3061, simple_loss=0.3706, pruned_loss=0.1208, over 19526.00 frames. ], tot_loss[loss=0.3188, simple_loss=0.3683, pruned_loss=0.1347, over 3811999.38 frames. ], batch size: 56, lr: 2.32e-02, grad_scale: 4.0 +2023-04-01 01:36:26,968 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 01:36:46,206 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 01:37:21,765 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19606.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:37:22,517 INFO [train.py:903] (3/4) Epoch 3, batch 5950, loss[loss=0.3539, simple_loss=0.398, pruned_loss=0.1549, over 18742.00 frames. ], tot_loss[loss=0.3202, simple_loss=0.3696, pruned_loss=0.1354, over 3812894.43 frames. ], batch size: 74, lr: 2.32e-02, grad_scale: 4.0 +2023-04-01 01:37:45,644 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.616e+02 6.894e+02 9.030e+02 1.163e+03 3.004e+03, threshold=1.806e+03, percent-clipped=5.0 +2023-04-01 01:37:59,467 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19636.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:38:24,423 INFO [train.py:903] (3/4) Epoch 3, batch 6000, loss[loss=0.2871, simple_loss=0.3321, pruned_loss=0.121, over 19304.00 frames. ], tot_loss[loss=0.3177, simple_loss=0.3677, pruned_loss=0.1339, over 3821622.16 frames. ], batch size: 44, lr: 2.31e-02, grad_scale: 8.0 +2023-04-01 01:38:24,424 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 01:38:34,251 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3772, 1.1855, 1.4498, 1.0201, 2.3253, 2.6771, 2.6395, 2.8815], + device='cuda:3'), covar=tensor([0.1264, 0.2651, 0.2632, 0.2101, 0.0436, 0.0185, 0.0257, 0.0149], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0274, 0.0316, 0.0264, 0.0196, 0.0107, 0.0197, 0.0119], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 01:38:37,335 INFO [train.py:937] (3/4) Epoch 3, validation: loss=0.2218, simple_loss=0.3182, pruned_loss=0.06273, over 944034.00 frames. +2023-04-01 01:38:37,336 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18419MB +2023-04-01 01:38:42,421 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-04-01 01:38:45,556 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19663.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:39:15,740 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19688.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:39:37,223 INFO [train.py:903] (3/4) Epoch 3, batch 6050, loss[loss=0.2937, simple_loss=0.3457, pruned_loss=0.1209, over 19300.00 frames. ], tot_loss[loss=0.3194, simple_loss=0.3688, pruned_loss=0.135, over 3818755.54 frames. ], batch size: 44, lr: 2.31e-02, grad_scale: 8.0 +2023-04-01 01:39:59,307 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.085e+02 7.768e+02 9.451e+02 1.306e+03 2.772e+03, threshold=1.890e+03, percent-clipped=6.0 +2023-04-01 01:40:37,800 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=19756.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:40:38,857 INFO [train.py:903] (3/4) Epoch 3, batch 6100, loss[loss=0.2756, simple_loss=0.3442, pruned_loss=0.1035, over 19677.00 frames. ], tot_loss[loss=0.3182, simple_loss=0.3677, pruned_loss=0.1343, over 3812058.12 frames. ], batch size: 59, lr: 2.31e-02, grad_scale: 8.0 +2023-04-01 01:41:38,885 INFO [train.py:903] (3/4) Epoch 3, batch 6150, loss[loss=0.3051, simple_loss=0.3625, pruned_loss=0.1239, over 19598.00 frames. ], tot_loss[loss=0.3186, simple_loss=0.3682, pruned_loss=0.1345, over 3830043.41 frames. ], batch size: 57, lr: 2.30e-02, grad_scale: 8.0 +2023-04-01 01:42:01,436 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.829e+02 8.602e+02 1.123e+03 1.510e+03 2.312e+03, threshold=2.246e+03, percent-clipped=7.0 +2023-04-01 01:42:04,691 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 01:42:38,570 INFO [train.py:903] (3/4) Epoch 3, batch 6200, loss[loss=0.3639, simple_loss=0.4063, pruned_loss=0.1607, over 19320.00 frames. ], tot_loss[loss=0.3189, simple_loss=0.3689, pruned_loss=0.1344, over 3832979.30 frames. ], batch size: 66, lr: 2.30e-02, grad_scale: 8.0 +2023-04-01 01:42:46,198 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19862.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 01:42:57,096 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19871.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:43:10,824 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-01 01:43:15,884 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19887.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 01:43:39,796 INFO [train.py:903] (3/4) Epoch 3, batch 6250, loss[loss=0.3666, simple_loss=0.4101, pruned_loss=0.1615, over 19117.00 frames. ], tot_loss[loss=0.3176, simple_loss=0.3676, pruned_loss=0.1338, over 3838759.11 frames. ], batch size: 69, lr: 2.30e-02, grad_scale: 8.0 +2023-04-01 01:44:01,769 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.496e+02 7.100e+02 9.208e+02 1.133e+03 3.490e+03, threshold=1.842e+03, percent-clipped=3.0 +2023-04-01 01:44:09,646 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 01:44:19,060 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9392, 5.2134, 3.1485, 4.7897, 1.4494, 5.2321, 5.1455, 5.5171], + device='cuda:3'), covar=tensor([0.0396, 0.0899, 0.1715, 0.0509, 0.3689, 0.0692, 0.0502, 0.0656], + device='cuda:3'), in_proj_covar=tensor([0.0320, 0.0293, 0.0339, 0.0270, 0.0339, 0.0288, 0.0244, 0.0288], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:44:38,572 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9374, 4.9416, 5.8169, 5.7450, 1.6526, 5.3940, 4.6260, 5.2065], + device='cuda:3'), covar=tensor([0.0570, 0.0446, 0.0379, 0.0227, 0.3801, 0.0199, 0.0411, 0.0806], + device='cuda:3'), in_proj_covar=tensor([0.0401, 0.0365, 0.0494, 0.0375, 0.0501, 0.0266, 0.0332, 0.0469], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 01:44:40,500 INFO [train.py:903] (3/4) Epoch 3, batch 6300, loss[loss=0.3066, simple_loss=0.3597, pruned_loss=0.1267, over 19851.00 frames. ], tot_loss[loss=0.3164, simple_loss=0.3665, pruned_loss=0.1332, over 3843576.24 frames. ], batch size: 52, lr: 2.30e-02, grad_scale: 8.0 +2023-04-01 01:44:47,590 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9227, 1.5707, 1.7279, 2.1372, 1.8376, 2.0178, 2.2044, 2.0158], + device='cuda:3'), covar=tensor([0.0591, 0.0884, 0.0862, 0.0738, 0.0873, 0.0683, 0.0788, 0.0515], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0280, 0.0268, 0.0308, 0.0308, 0.0252, 0.0274, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 01:45:07,280 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=19980.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:45:42,109 INFO [train.py:903] (3/4) Epoch 3, batch 6350, loss[loss=0.3183, simple_loss=0.3765, pruned_loss=0.13, over 19570.00 frames. ], tot_loss[loss=0.3176, simple_loss=0.3671, pruned_loss=0.134, over 3819543.24 frames. ], batch size: 61, lr: 2.29e-02, grad_scale: 8.0 +2023-04-01 01:46:03,312 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.926e+02 7.424e+02 9.293e+02 1.158e+03 2.968e+03, threshold=1.859e+03, percent-clipped=5.0 +2023-04-01 01:46:20,204 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20038.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:46:42,149 INFO [train.py:903] (3/4) Epoch 3, batch 6400, loss[loss=0.2842, simple_loss=0.3428, pruned_loss=0.1128, over 19532.00 frames. ], tot_loss[loss=0.3193, simple_loss=0.3686, pruned_loss=0.1349, over 3818120.40 frames. ], batch size: 54, lr: 2.29e-02, grad_scale: 8.0 +2023-04-01 01:47:09,212 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3571, 2.1931, 1.4470, 1.7094, 2.0533, 1.1124, 1.0199, 1.5866], + device='cuda:3'), covar=tensor([0.0815, 0.0515, 0.1042, 0.0438, 0.0407, 0.1089, 0.0806, 0.0497], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0245, 0.0320, 0.0251, 0.0218, 0.0312, 0.0282, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 01:47:29,327 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20095.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:47:43,491 INFO [train.py:903] (3/4) Epoch 3, batch 6450, loss[loss=0.3127, simple_loss=0.3627, pruned_loss=0.1314, over 19776.00 frames. ], tot_loss[loss=0.3196, simple_loss=0.3691, pruned_loss=0.1351, over 3820725.73 frames. ], batch size: 54, lr: 2.29e-02, grad_scale: 8.0 +2023-04-01 01:47:43,814 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20107.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:48:05,562 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.028e+02 7.833e+02 9.380e+02 1.145e+03 2.427e+03, threshold=1.876e+03, percent-clipped=3.0 +2023-04-01 01:48:08,323 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20127.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:48:12,605 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20130.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:48:29,095 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 01:48:39,377 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20152.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:48:44,548 INFO [train.py:903] (3/4) Epoch 3, batch 6500, loss[loss=0.3153, simple_loss=0.3606, pruned_loss=0.135, over 19582.00 frames. ], tot_loss[loss=0.3186, simple_loss=0.3681, pruned_loss=0.1345, over 3824309.29 frames. ], batch size: 52, lr: 2.29e-02, grad_scale: 8.0 +2023-04-01 01:48:52,306 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 01:49:45,349 INFO [train.py:903] (3/4) Epoch 3, batch 6550, loss[loss=0.3496, simple_loss=0.3944, pruned_loss=0.1524, over 19671.00 frames. ], tot_loss[loss=0.3178, simple_loss=0.368, pruned_loss=0.1337, over 3825586.05 frames. ], batch size: 55, lr: 2.28e-02, grad_scale: 8.0 +2023-04-01 01:50:03,604 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20222.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:50:07,447 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.125e+02 7.085e+02 9.596e+02 1.224e+03 2.377e+03, threshold=1.919e+03, percent-clipped=5.0 +2023-04-01 01:50:28,031 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2208, 2.7747, 1.8963, 2.1461, 1.6335, 2.2101, 0.4576, 2.1986], + device='cuda:3'), covar=tensor([0.0275, 0.0217, 0.0239, 0.0354, 0.0501, 0.0323, 0.0630, 0.0370], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0255, 0.0254, 0.0277, 0.0337, 0.0270, 0.0259, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:50:46,685 INFO [train.py:903] (3/4) Epoch 3, batch 6600, loss[loss=0.3448, simple_loss=0.3913, pruned_loss=0.1491, over 19482.00 frames. ], tot_loss[loss=0.3161, simple_loss=0.3668, pruned_loss=0.1327, over 3837507.37 frames. ], batch size: 64, lr: 2.28e-02, grad_scale: 8.0 +2023-04-01 01:51:47,896 INFO [train.py:903] (3/4) Epoch 3, batch 6650, loss[loss=0.3002, simple_loss=0.3538, pruned_loss=0.1233, over 17172.00 frames. ], tot_loss[loss=0.3174, simple_loss=0.3678, pruned_loss=0.1336, over 3831018.85 frames. ], batch size: 101, lr: 2.28e-02, grad_scale: 8.0 +2023-04-01 01:52:10,192 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.507e+02 7.672e+02 8.992e+02 1.144e+03 3.524e+03, threshold=1.798e+03, percent-clipped=4.0 +2023-04-01 01:52:42,306 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20351.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:52:47,437 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7212, 3.8258, 4.2404, 4.1914, 2.4314, 3.6850, 3.5102, 3.8697], + device='cuda:3'), covar=tensor([0.0679, 0.1053, 0.0476, 0.0344, 0.2539, 0.0319, 0.0391, 0.0857], + device='cuda:3'), in_proj_covar=tensor([0.0416, 0.0368, 0.0504, 0.0388, 0.0514, 0.0274, 0.0330, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 01:52:49,541 INFO [train.py:903] (3/4) Epoch 3, batch 6700, loss[loss=0.2292, simple_loss=0.2991, pruned_loss=0.07963, over 19505.00 frames. ], tot_loss[loss=0.3161, simple_loss=0.3666, pruned_loss=0.1328, over 3834984.42 frames. ], batch size: 49, lr: 2.28e-02, grad_scale: 8.0 +2023-04-01 01:53:11,826 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20376.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:53:17,990 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20382.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:53:45,269 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20406.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:53:46,032 INFO [train.py:903] (3/4) Epoch 3, batch 6750, loss[loss=0.3543, simple_loss=0.3932, pruned_loss=0.1577, over 13457.00 frames. ], tot_loss[loss=0.3166, simple_loss=0.3666, pruned_loss=0.1333, over 3830594.14 frames. ], batch size: 136, lr: 2.27e-02, grad_scale: 8.0 +2023-04-01 01:54:05,360 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.786e+02 7.667e+02 1.002e+03 1.269e+03 2.908e+03, threshold=2.004e+03, percent-clipped=6.0 +2023-04-01 01:54:17,342 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20436.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:54:34,793 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20451.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:54:41,604 INFO [train.py:903] (3/4) Epoch 3, batch 6800, loss[loss=0.2975, simple_loss=0.3552, pruned_loss=0.1199, over 19625.00 frames. ], tot_loss[loss=0.3156, simple_loss=0.3661, pruned_loss=0.1326, over 3818789.69 frames. ], batch size: 50, lr: 2.27e-02, grad_scale: 8.0 +2023-04-01 01:55:00,655 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20474.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:55:25,479 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 01:55:26,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 01:55:29,039 INFO [train.py:903] (3/4) Epoch 4, batch 0, loss[loss=0.3367, simple_loss=0.3956, pruned_loss=0.1389, over 19389.00 frames. ], tot_loss[loss=0.3367, simple_loss=0.3956, pruned_loss=0.1389, over 19389.00 frames. ], batch size: 70, lr: 2.12e-02, grad_scale: 8.0 +2023-04-01 01:55:29,039 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 01:55:40,519 INFO [train.py:937] (3/4) Epoch 4, validation: loss=0.2245, simple_loss=0.3205, pruned_loss=0.06426, over 944034.00 frames. +2023-04-01 01:55:40,520 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18419MB +2023-04-01 01:55:53,638 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 01:55:55,196 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20497.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:56:27,871 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.490e+02 6.855e+02 8.790e+02 1.166e+03 2.960e+03, threshold=1.758e+03, percent-clipped=3.0 +2023-04-01 01:56:41,015 INFO [train.py:903] (3/4) Epoch 4, batch 50, loss[loss=0.3065, simple_loss=0.37, pruned_loss=0.1215, over 19642.00 frames. ], tot_loss[loss=0.321, simple_loss=0.3701, pruned_loss=0.1359, over 865177.47 frames. ], batch size: 60, lr: 2.12e-02, grad_scale: 8.0 +2023-04-01 01:57:14,499 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 01:57:15,816 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20566.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:57:16,031 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20566.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:57:32,391 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0670, 1.0032, 1.3953, 0.5335, 2.4354, 2.4276, 2.2485, 2.5262], + device='cuda:3'), covar=tensor([0.1231, 0.2599, 0.2544, 0.2001, 0.0285, 0.0133, 0.0283, 0.0165], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0273, 0.0316, 0.0261, 0.0187, 0.0104, 0.0194, 0.0116], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 01:57:40,112 INFO [train.py:903] (3/4) Epoch 4, batch 100, loss[loss=0.3389, simple_loss=0.3888, pruned_loss=0.1445, over 19741.00 frames. ], tot_loss[loss=0.3173, simple_loss=0.3659, pruned_loss=0.1343, over 1520714.54 frames. ], batch size: 63, lr: 2.12e-02, grad_scale: 8.0 +2023-04-01 01:57:46,163 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20589.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:57:52,779 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 01:58:05,929 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:58:16,462 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4353, 4.0296, 2.4359, 3.6570, 1.2790, 3.5030, 3.5203, 3.6321], + device='cuda:3'), covar=tensor([0.0595, 0.1036, 0.1994, 0.0669, 0.3783, 0.1023, 0.0756, 0.1136], + device='cuda:3'), in_proj_covar=tensor([0.0323, 0.0297, 0.0336, 0.0267, 0.0345, 0.0290, 0.0248, 0.0288], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 01:58:29,320 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.104e+02 7.763e+02 9.275e+02 1.175e+03 2.763e+03, threshold=1.855e+03, percent-clipped=7.0 +2023-04-01 01:58:41,720 INFO [train.py:903] (3/4) Epoch 4, batch 150, loss[loss=0.2743, simple_loss=0.3372, pruned_loss=0.1057, over 19593.00 frames. ], tot_loss[loss=0.3124, simple_loss=0.3634, pruned_loss=0.1307, over 2035614.70 frames. ], batch size: 52, lr: 2.11e-02, grad_scale: 8.0 +2023-04-01 01:59:03,928 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.73 vs. limit=5.0 +2023-04-01 01:59:35,941 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20681.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 01:59:40,165 INFO [train.py:903] (3/4) Epoch 4, batch 200, loss[loss=0.2877, simple_loss=0.3375, pruned_loss=0.1189, over 19782.00 frames. ], tot_loss[loss=0.3116, simple_loss=0.3628, pruned_loss=0.1302, over 2437284.51 frames. ], batch size: 48, lr: 2.11e-02, grad_scale: 8.0 +2023-04-01 01:59:41,292 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 01:59:54,697 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.77 vs. limit=5.0 +2023-04-01 02:00:20,392 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 02:00:28,682 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.823e+02 7.097e+02 9.184e+02 1.257e+03 2.857e+03, threshold=1.837e+03, percent-clipped=5.0 +2023-04-01 02:00:39,432 INFO [train.py:903] (3/4) Epoch 4, batch 250, loss[loss=0.2642, simple_loss=0.3317, pruned_loss=0.09833, over 19661.00 frames. ], tot_loss[loss=0.3081, simple_loss=0.3599, pruned_loss=0.1281, over 2766927.36 frames. ], batch size: 55, lr: 2.11e-02, grad_scale: 8.0 +2023-04-01 02:00:45,040 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1981, 2.1399, 1.8552, 3.5259, 2.1595, 3.5578, 3.4263, 1.8818], + device='cuda:3'), covar=tensor([0.1406, 0.1000, 0.0568, 0.0633, 0.1296, 0.0309, 0.0848, 0.0957], + device='cuda:3'), in_proj_covar=tensor([0.0513, 0.0483, 0.0478, 0.0645, 0.0550, 0.0399, 0.0578, 0.0481], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:00:57,788 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:01:02,449 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20753.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:01:20,560 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7577, 1.4884, 1.9317, 1.6760, 3.3260, 4.5521, 4.6539, 5.1701], + device='cuda:3'), covar=tensor([0.1367, 0.2810, 0.2594, 0.1846, 0.0354, 0.0115, 0.0138, 0.0055], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0279, 0.0323, 0.0265, 0.0194, 0.0109, 0.0200, 0.0120], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 02:01:24,172 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1214, 1.3356, 2.1225, 1.5535, 2.9294, 2.6727, 3.1955, 1.3949], + device='cuda:3'), covar=tensor([0.1789, 0.2693, 0.1439, 0.1470, 0.1256, 0.1243, 0.1573, 0.2609], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0469, 0.0436, 0.0397, 0.0513, 0.0414, 0.0591, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:01:32,693 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20778.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:01:34,839 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20780.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:01:41,110 INFO [train.py:903] (3/4) Epoch 4, batch 300, loss[loss=0.3697, simple_loss=0.403, pruned_loss=0.1682, over 19315.00 frames. ], tot_loss[loss=0.3066, simple_loss=0.3593, pruned_loss=0.127, over 3005273.85 frames. ], batch size: 66, lr: 2.11e-02, grad_scale: 8.0 +2023-04-01 02:02:25,225 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20822.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:02:29,140 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.541e+02 7.559e+02 9.012e+02 1.206e+03 2.235e+03, threshold=1.802e+03, percent-clipped=6.0 +2023-04-01 02:02:30,545 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7344, 1.5558, 2.0794, 2.5322, 2.3110, 2.6907, 2.2281, 3.0355], + device='cuda:3'), covar=tensor([0.0697, 0.2008, 0.1188, 0.0818, 0.1228, 0.0374, 0.0824, 0.0418], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0381, 0.0278, 0.0243, 0.0312, 0.0259, 0.0267, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:02:40,183 INFO [train.py:903] (3/4) Epoch 4, batch 350, loss[loss=0.2889, simple_loss=0.3516, pruned_loss=0.1131, over 19668.00 frames. ], tot_loss[loss=0.3098, simple_loss=0.3618, pruned_loss=0.1289, over 3178525.23 frames. ], batch size: 58, lr: 2.10e-02, grad_scale: 8.0 +2023-04-01 02:02:45,656 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 02:02:52,661 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20845.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:02:54,938 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20847.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:03:00,487 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0611, 4.8359, 5.7660, 5.7624, 1.7306, 5.2606, 4.6550, 5.2110], + device='cuda:3'), covar=tensor([0.0620, 0.0457, 0.0342, 0.0237, 0.3478, 0.0176, 0.0330, 0.0747], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0374, 0.0508, 0.0390, 0.0513, 0.0278, 0.0333, 0.0483], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:03:03,958 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0957, 5.4210, 3.1831, 4.7296, 1.3024, 5.1491, 5.1417, 5.4605], + device='cuda:3'), covar=tensor([0.0404, 0.0900, 0.1828, 0.0617, 0.4001, 0.0724, 0.0493, 0.0545], + device='cuda:3'), in_proj_covar=tensor([0.0334, 0.0304, 0.0346, 0.0280, 0.0351, 0.0295, 0.0252, 0.0296], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:03:18,431 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20865.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:03:23,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20870.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:03:40,718 INFO [train.py:903] (3/4) Epoch 4, batch 400, loss[loss=0.2942, simple_loss=0.3578, pruned_loss=0.1153, over 19779.00 frames. ], tot_loss[loss=0.3097, simple_loss=0.362, pruned_loss=0.1287, over 3330595.31 frames. ], batch size: 56, lr: 2.10e-02, grad_scale: 8.0 +2023-04-01 02:03:52,334 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20895.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:04:27,878 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.312e+02 7.529e+02 9.870e+02 1.265e+03 2.610e+03, threshold=1.974e+03, percent-clipped=3.0 +2023-04-01 02:04:36,624 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.85 vs. limit=5.0 +2023-04-01 02:04:39,083 INFO [train.py:903] (3/4) Epoch 4, batch 450, loss[loss=0.2979, simple_loss=0.3612, pruned_loss=0.1172, over 19787.00 frames. ], tot_loss[loss=0.3095, simple_loss=0.3618, pruned_loss=0.1286, over 3452323.14 frames. ], batch size: 56, lr: 2.10e-02, grad_scale: 8.0 +2023-04-01 02:04:41,821 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20937.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:04:58,391 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=20950.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:05:12,735 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 02:05:13,080 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20962.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:05:13,941 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 02:05:40,645 INFO [train.py:903] (3/4) Epoch 4, batch 500, loss[loss=0.2934, simple_loss=0.3391, pruned_loss=0.1238, over 19488.00 frames. ], tot_loss[loss=0.3094, simple_loss=0.3617, pruned_loss=0.1286, over 3547205.25 frames. ], batch size: 49, lr: 2.10e-02, grad_scale: 8.0 +2023-04-01 02:06:27,515 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.944e+02 7.370e+02 9.144e+02 1.191e+03 3.185e+03, threshold=1.829e+03, percent-clipped=4.0 +2023-04-01 02:06:40,264 INFO [train.py:903] (3/4) Epoch 4, batch 550, loss[loss=0.3418, simple_loss=0.3819, pruned_loss=0.1508, over 19519.00 frames. ], tot_loss[loss=0.3104, simple_loss=0.3625, pruned_loss=0.1292, over 3610763.03 frames. ], batch size: 54, lr: 2.10e-02, grad_scale: 8.0 +2023-04-01 02:07:17,121 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21065.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:07:23,287 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21070.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:07:37,589 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9492, 3.6499, 2.2313, 3.3632, 1.2685, 3.3208, 3.1438, 3.3868], + device='cuda:3'), covar=tensor([0.0611, 0.1024, 0.2024, 0.0695, 0.3342, 0.0927, 0.0770, 0.0897], + device='cuda:3'), in_proj_covar=tensor([0.0321, 0.0295, 0.0338, 0.0273, 0.0339, 0.0292, 0.0251, 0.0288], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:07:39,554 INFO [train.py:903] (3/4) Epoch 4, batch 600, loss[loss=0.3104, simple_loss=0.3628, pruned_loss=0.129, over 17499.00 frames. ], tot_loss[loss=0.3117, simple_loss=0.3635, pruned_loss=0.13, over 3655005.95 frames. ], batch size: 101, lr: 2.09e-02, grad_scale: 8.0 +2023-04-01 02:08:15,013 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2470, 1.5412, 1.4642, 2.5025, 1.9320, 2.2652, 2.6083, 2.2636], + device='cuda:3'), covar=tensor([0.0775, 0.1186, 0.1389, 0.1071, 0.1158, 0.0777, 0.0967, 0.0721], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0272, 0.0267, 0.0296, 0.0300, 0.0249, 0.0263, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 02:08:19,258 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 02:08:23,067 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21121.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:08:27,139 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.199e+02 8.023e+02 1.001e+03 1.305e+03 2.804e+03, threshold=2.003e+03, percent-clipped=3.0 +2023-04-01 02:08:39,444 INFO [train.py:903] (3/4) Epoch 4, batch 650, loss[loss=0.2643, simple_loss=0.3195, pruned_loss=0.1046, over 19761.00 frames. ], tot_loss[loss=0.3107, simple_loss=0.3626, pruned_loss=0.1294, over 3691067.52 frames. ], batch size: 46, lr: 2.09e-02, grad_scale: 8.0 +2023-04-01 02:08:40,635 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21136.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:08:53,046 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21146.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:09:00,264 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21151.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:09:16,937 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5657, 1.5748, 1.9949, 2.4444, 4.2225, 1.1267, 2.2395, 4.2863], + device='cuda:3'), covar=tensor([0.0292, 0.2277, 0.1918, 0.1240, 0.0405, 0.2236, 0.1225, 0.0288], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0303, 0.0299, 0.0277, 0.0294, 0.0318, 0.0282, 0.0284], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:09:29,119 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21176.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:09:38,634 INFO [train.py:903] (3/4) Epoch 4, batch 700, loss[loss=0.3994, simple_loss=0.4211, pruned_loss=0.1888, over 17492.00 frames. ], tot_loss[loss=0.3107, simple_loss=0.3625, pruned_loss=0.1295, over 3723998.89 frames. ], batch size: 101, lr: 2.09e-02, grad_scale: 8.0 +2023-04-01 02:10:26,834 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.035e+02 7.118e+02 8.929e+02 1.162e+03 2.438e+03, threshold=1.786e+03, percent-clipped=3.0 +2023-04-01 02:10:40,520 INFO [train.py:903] (3/4) Epoch 4, batch 750, loss[loss=0.3089, simple_loss=0.3713, pruned_loss=0.1233, over 19747.00 frames. ], tot_loss[loss=0.3099, simple_loss=0.362, pruned_loss=0.1289, over 3756354.65 frames. ], batch size: 63, lr: 2.09e-02, grad_scale: 8.0 +2023-04-01 02:10:44,254 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7439, 1.3854, 1.2277, 1.5292, 1.4562, 1.4468, 1.2743, 1.5467], + device='cuda:3'), covar=tensor([0.0786, 0.1346, 0.1285, 0.0797, 0.0979, 0.0537, 0.0915, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0374, 0.0279, 0.0247, 0.0313, 0.0259, 0.0262, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:10:48,899 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4085, 1.3830, 2.2184, 1.6902, 3.0719, 3.0144, 3.5205, 1.4042], + device='cuda:3'), covar=tensor([0.1490, 0.2478, 0.1357, 0.1233, 0.1033, 0.1007, 0.1220, 0.2339], + device='cuda:3'), in_proj_covar=tensor([0.0424, 0.0476, 0.0438, 0.0396, 0.0517, 0.0414, 0.0597, 0.0425], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:11:40,042 INFO [train.py:903] (3/4) Epoch 4, batch 800, loss[loss=0.3172, simple_loss=0.3664, pruned_loss=0.134, over 19599.00 frames. ], tot_loss[loss=0.3089, simple_loss=0.3613, pruned_loss=0.1282, over 3782244.90 frames. ], batch size: 50, lr: 2.08e-02, grad_scale: 8.0 +2023-04-01 02:11:56,118 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 02:12:25,403 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21321.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:12:30,558 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.379e+02 8.113e+02 1.001e+03 1.207e+03 2.017e+03, threshold=2.002e+03, percent-clipped=2.0 +2023-04-01 02:12:35,470 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21330.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:12:41,604 INFO [train.py:903] (3/4) Epoch 4, batch 850, loss[loss=0.2654, simple_loss=0.3305, pruned_loss=0.1002, over 19663.00 frames. ], tot_loss[loss=0.3095, simple_loss=0.362, pruned_loss=0.1284, over 3788890.09 frames. ], batch size: 53, lr: 2.08e-02, grad_scale: 8.0 +2023-04-01 02:12:47,672 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0985, 1.9846, 1.6300, 1.6402, 1.4655, 1.6836, 0.2024, 0.8337], + device='cuda:3'), covar=tensor([0.0199, 0.0182, 0.0140, 0.0180, 0.0430, 0.0199, 0.0413, 0.0380], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0256, 0.0259, 0.0288, 0.0342, 0.0273, 0.0264, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:12:54,289 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21346.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:12:54,393 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21346.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:13:30,946 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 02:13:39,981 INFO [train.py:903] (3/4) Epoch 4, batch 900, loss[loss=0.2951, simple_loss=0.3497, pruned_loss=0.1202, over 19607.00 frames. ], tot_loss[loss=0.3097, simple_loss=0.3623, pruned_loss=0.1286, over 3798534.92 frames. ], batch size: 52, lr: 2.08e-02, grad_scale: 8.0 +2023-04-01 02:14:14,506 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21414.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:14:28,710 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.494e+02 7.656e+02 9.192e+02 1.106e+03 2.022e+03, threshold=1.838e+03, percent-clipped=1.0 +2023-04-01 02:14:40,644 INFO [train.py:903] (3/4) Epoch 4, batch 950, loss[loss=0.2954, simple_loss=0.3572, pruned_loss=0.1168, over 18594.00 frames. ], tot_loss[loss=0.3096, simple_loss=0.3621, pruned_loss=0.1286, over 3804484.22 frames. ], batch size: 74, lr: 2.08e-02, grad_scale: 8.0 +2023-04-01 02:14:43,013 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 02:14:43,648 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-04-01 02:15:35,418 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21480.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:15:37,652 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6076, 4.1540, 2.3446, 3.7474, 1.2009, 3.6983, 3.7719, 3.8275], + device='cuda:3'), covar=tensor([0.0553, 0.0917, 0.1939, 0.0681, 0.3901, 0.0969, 0.0666, 0.0857], + device='cuda:3'), in_proj_covar=tensor([0.0320, 0.0290, 0.0337, 0.0269, 0.0339, 0.0293, 0.0253, 0.0283], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:15:40,656 INFO [train.py:903] (3/4) Epoch 4, batch 1000, loss[loss=0.328, simple_loss=0.3788, pruned_loss=0.1386, over 19671.00 frames. ], tot_loss[loss=0.3098, simple_loss=0.3621, pruned_loss=0.1287, over 3811609.43 frames. ], batch size: 58, lr: 2.07e-02, grad_scale: 8.0 +2023-04-01 02:16:28,019 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4533, 1.3742, 1.0902, 1.2988, 1.2203, 1.2850, 1.0513, 1.3313], + device='cuda:3'), covar=tensor([0.0797, 0.0959, 0.1177, 0.0814, 0.0939, 0.0493, 0.1033, 0.0653], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0370, 0.0277, 0.0249, 0.0311, 0.0258, 0.0265, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:16:29,779 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.128e+02 7.436e+02 9.242e+02 1.292e+03 2.692e+03, threshold=1.848e+03, percent-clipped=7.0 +2023-04-01 02:16:33,304 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 02:16:33,634 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21529.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:16:35,886 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.6568, 0.8676, 0.6668, 0.6805, 0.8169, 0.6309, 0.4117, 0.8467], + device='cuda:3'), covar=tensor([0.0382, 0.0410, 0.0629, 0.0339, 0.0271, 0.0698, 0.0498, 0.0297], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0252, 0.0315, 0.0248, 0.0217, 0.0314, 0.0279, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:16:40,202 INFO [train.py:903] (3/4) Epoch 4, batch 1050, loss[loss=0.3298, simple_loss=0.376, pruned_loss=0.1418, over 19794.00 frames. ], tot_loss[loss=0.3106, simple_loss=0.3624, pruned_loss=0.1294, over 3810504.97 frames. ], batch size: 56, lr: 2.07e-02, grad_scale: 8.0 +2023-04-01 02:16:57,650 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4454, 2.4279, 1.7227, 1.6702, 2.2181, 1.0425, 1.1838, 1.7710], + device='cuda:3'), covar=tensor([0.0889, 0.0516, 0.0952, 0.0557, 0.0498, 0.1265, 0.0779, 0.0515], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0249, 0.0313, 0.0244, 0.0216, 0.0311, 0.0275, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:17:14,226 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 02:17:30,946 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.48 vs. limit=5.0 +2023-04-01 02:17:34,814 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9965, 2.0950, 1.8732, 2.9135, 1.9355, 3.1233, 2.6416, 1.7880], + device='cuda:3'), covar=tensor([0.1356, 0.0954, 0.0604, 0.0645, 0.1191, 0.0308, 0.1014, 0.0972], + device='cuda:3'), in_proj_covar=tensor([0.0531, 0.0503, 0.0492, 0.0666, 0.0569, 0.0418, 0.0590, 0.0490], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:17:40,184 INFO [train.py:903] (3/4) Epoch 4, batch 1100, loss[loss=0.25, simple_loss=0.3048, pruned_loss=0.09759, over 19288.00 frames. ], tot_loss[loss=0.3096, simple_loss=0.3615, pruned_loss=0.1288, over 3826648.16 frames. ], batch size: 44, lr: 2.07e-02, grad_scale: 8.0 +2023-04-01 02:17:52,925 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21595.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:18:15,016 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21613.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:18:30,558 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.767e+02 7.921e+02 9.622e+02 1.275e+03 2.981e+03, threshold=1.924e+03, percent-clipped=6.0 +2023-04-01 02:18:36,432 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21631.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:18:43,701 INFO [train.py:903] (3/4) Epoch 4, batch 1150, loss[loss=0.2695, simple_loss=0.3431, pruned_loss=0.09798, over 19668.00 frames. ], tot_loss[loss=0.3081, simple_loss=0.3601, pruned_loss=0.128, over 3837053.00 frames. ], batch size: 53, lr: 2.07e-02, grad_scale: 8.0 +2023-04-01 02:19:31,573 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21674.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:19:45,549 INFO [train.py:903] (3/4) Epoch 4, batch 1200, loss[loss=0.3406, simple_loss=0.3857, pruned_loss=0.1478, over 17520.00 frames. ], tot_loss[loss=0.3086, simple_loss=0.3608, pruned_loss=0.1282, over 3829487.50 frames. ], batch size: 100, lr: 2.07e-02, grad_scale: 8.0 +2023-04-01 02:19:51,428 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21690.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:20:14,904 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 02:20:18,749 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.78 vs. limit=5.0 +2023-04-01 02:20:35,528 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.245e+02 7.515e+02 9.038e+02 1.111e+03 2.454e+03, threshold=1.808e+03, percent-clipped=2.0 +2023-04-01 02:20:45,432 INFO [train.py:903] (3/4) Epoch 4, batch 1250, loss[loss=0.3471, simple_loss=0.391, pruned_loss=0.1516, over 19692.00 frames. ], tot_loss[loss=0.3069, simple_loss=0.3597, pruned_loss=0.127, over 3836354.46 frames. ], batch size: 60, lr: 2.06e-02, grad_scale: 8.0 +2023-04-01 02:21:39,848 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21781.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:21:44,775 INFO [train.py:903] (3/4) Epoch 4, batch 1300, loss[loss=0.277, simple_loss=0.344, pruned_loss=0.105, over 19697.00 frames. ], tot_loss[loss=0.3077, simple_loss=0.3607, pruned_loss=0.1273, over 3838910.80 frames. ], batch size: 60, lr: 2.06e-02, grad_scale: 8.0 +2023-04-01 02:21:45,180 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21785.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:21:49,643 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21789.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:22:10,090 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21805.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:22:15,829 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21810.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:22:34,649 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.842e+02 7.410e+02 9.034e+02 1.121e+03 1.849e+03, threshold=1.807e+03, percent-clipped=1.0 +2023-04-01 02:22:45,117 INFO [train.py:903] (3/4) Epoch 4, batch 1350, loss[loss=0.3459, simple_loss=0.3677, pruned_loss=0.1621, over 18599.00 frames. ], tot_loss[loss=0.308, simple_loss=0.3606, pruned_loss=0.1277, over 3836196.67 frames. ], batch size: 41, lr: 2.06e-02, grad_scale: 8.0 +2023-04-01 02:23:04,096 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21851.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:23:05,052 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4433, 2.7979, 3.0221, 2.9892, 1.0293, 2.7124, 2.6303, 2.4401], + device='cuda:3'), covar=tensor([0.1635, 0.1115, 0.0978, 0.0986, 0.4697, 0.0813, 0.0768, 0.1850], + device='cuda:3'), in_proj_covar=tensor([0.0432, 0.0377, 0.0512, 0.0397, 0.0525, 0.0284, 0.0338, 0.0485], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:23:23,912 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21868.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:23:34,825 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21876.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:23:45,417 INFO [train.py:903] (3/4) Epoch 4, batch 1400, loss[loss=0.3026, simple_loss=0.3581, pruned_loss=0.1236, over 19588.00 frames. ], tot_loss[loss=0.3096, simple_loss=0.3617, pruned_loss=0.1288, over 3816158.90 frames. ], batch size: 52, lr: 2.06e-02, grad_scale: 8.0 +2023-04-01 02:24:35,708 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.358e+02 7.364e+02 9.517e+02 1.310e+03 2.254e+03, threshold=1.903e+03, percent-clipped=6.0 +2023-04-01 02:24:42,742 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 02:24:46,194 INFO [train.py:903] (3/4) Epoch 4, batch 1450, loss[loss=0.3347, simple_loss=0.3821, pruned_loss=0.1437, over 19784.00 frames. ], tot_loss[loss=0.3092, simple_loss=0.3616, pruned_loss=0.1284, over 3807640.54 frames. ], batch size: 56, lr: 2.05e-02, grad_scale: 8.0 +2023-04-01 02:24:48,760 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7476, 4.7784, 5.5635, 5.4754, 1.7872, 5.0362, 4.5306, 4.8778], + device='cuda:3'), covar=tensor([0.0685, 0.0534, 0.0349, 0.0235, 0.3591, 0.0247, 0.0333, 0.0840], + device='cuda:3'), in_proj_covar=tensor([0.0436, 0.0383, 0.0513, 0.0398, 0.0529, 0.0288, 0.0339, 0.0496], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:25:13,225 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21957.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:25:34,247 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=21975.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:25:45,920 INFO [train.py:903] (3/4) Epoch 4, batch 1500, loss[loss=0.3979, simple_loss=0.424, pruned_loss=0.1859, over 18712.00 frames. ], tot_loss[loss=0.3109, simple_loss=0.3627, pruned_loss=0.1295, over 3815130.97 frames. ], batch size: 74, lr: 2.05e-02, grad_scale: 8.0 +2023-04-01 02:25:46,716 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-01 02:25:49,651 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21988.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:26:18,127 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2050, 1.2028, 1.9439, 1.4139, 2.6991, 2.4386, 2.7856, 1.1553], + device='cuda:3'), covar=tensor([0.1717, 0.2908, 0.1459, 0.1404, 0.1081, 0.1207, 0.1274, 0.2601], + device='cuda:3'), in_proj_covar=tensor([0.0424, 0.0477, 0.0444, 0.0404, 0.0521, 0.0417, 0.0604, 0.0429], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:26:36,549 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.570e+02 6.805e+02 9.241e+02 1.170e+03 2.581e+03, threshold=1.848e+03, percent-clipped=2.0 +2023-04-01 02:26:47,114 INFO [train.py:903] (3/4) Epoch 4, batch 1550, loss[loss=0.2902, simple_loss=0.3507, pruned_loss=0.1149, over 19532.00 frames. ], tot_loss[loss=0.3083, simple_loss=0.361, pruned_loss=0.1278, over 3805740.83 frames. ], batch size: 54, lr: 2.05e-02, grad_scale: 8.0 +2023-04-01 02:27:01,639 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22045.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:27:19,413 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22061.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:27:30,471 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22070.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:27:32,796 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22072.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:27:42,174 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3047, 2.1399, 2.0274, 3.3133, 2.1338, 3.7200, 3.1460, 1.8765], + device='cuda:3'), covar=tensor([0.1623, 0.1238, 0.0658, 0.0793, 0.1536, 0.0340, 0.1108, 0.1126], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0506, 0.0495, 0.0670, 0.0574, 0.0425, 0.0594, 0.0501], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:27:49,225 INFO [train.py:903] (3/4) Epoch 4, batch 1600, loss[loss=0.3121, simple_loss=0.3675, pruned_loss=0.1284, over 18168.00 frames. ], tot_loss[loss=0.309, simple_loss=0.3618, pruned_loss=0.1281, over 3819451.33 frames. ], batch size: 83, lr: 2.05e-02, grad_scale: 8.0 +2023-04-01 02:27:50,800 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22086.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:27:55,240 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22090.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:28:10,563 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 02:28:37,657 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22125.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:28:39,603 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.648e+02 8.565e+02 1.081e+03 1.346e+03 3.673e+03, threshold=2.162e+03, percent-clipped=6.0 +2023-04-01 02:28:44,656 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0280, 1.5615, 1.5133, 1.8685, 1.7937, 1.7821, 1.6044, 1.8774], + device='cuda:3'), covar=tensor([0.0724, 0.1544, 0.1181, 0.0927, 0.1036, 0.0448, 0.0891, 0.0565], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0368, 0.0280, 0.0245, 0.0301, 0.0251, 0.0261, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:28:48,850 INFO [train.py:903] (3/4) Epoch 4, batch 1650, loss[loss=0.294, simple_loss=0.3557, pruned_loss=0.1161, over 19771.00 frames. ], tot_loss[loss=0.3084, simple_loss=0.3611, pruned_loss=0.1278, over 3822343.82 frames. ], batch size: 56, lr: 2.05e-02, grad_scale: 4.0 +2023-04-01 02:29:04,269 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22148.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:29:40,675 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7953, 3.1116, 3.2184, 3.2160, 1.1330, 2.9410, 2.7909, 2.8696], + device='cuda:3'), covar=tensor([0.0928, 0.0582, 0.0626, 0.0526, 0.3397, 0.0406, 0.0511, 0.1187], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0372, 0.0510, 0.0395, 0.0513, 0.0289, 0.0335, 0.0484], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:29:49,263 INFO [train.py:903] (3/4) Epoch 4, batch 1700, loss[loss=0.3003, simple_loss=0.3596, pruned_loss=0.1205, over 19613.00 frames. ], tot_loss[loss=0.3061, simple_loss=0.3595, pruned_loss=0.1264, over 3825363.01 frames. ], batch size: 57, lr: 2.04e-02, grad_scale: 4.0 +2023-04-01 02:29:52,348 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-01 02:30:00,422 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22194.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:30:23,334 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22212.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:30:27,873 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 02:30:40,334 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.920e+02 6.276e+02 7.753e+02 9.050e+02 1.909e+03, threshold=1.551e+03, percent-clipped=1.0 +2023-04-01 02:30:46,032 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9934, 1.1119, 1.3949, 0.5116, 2.2719, 2.3895, 2.2047, 2.4948], + device='cuda:3'), covar=tensor([0.1145, 0.2587, 0.2549, 0.1909, 0.0370, 0.0147, 0.0349, 0.0157], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0276, 0.0320, 0.0259, 0.0192, 0.0109, 0.0198, 0.0121], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 02:30:48,089 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22233.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:30:49,931 INFO [train.py:903] (3/4) Epoch 4, batch 1750, loss[loss=0.3218, simple_loss=0.3677, pruned_loss=0.1379, over 19746.00 frames. ], tot_loss[loss=0.3071, simple_loss=0.3604, pruned_loss=0.1269, over 3819949.42 frames. ], batch size: 51, lr: 2.04e-02, grad_scale: 4.0 +2023-04-01 02:30:57,742 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22240.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:31:15,614 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22255.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:31:52,437 INFO [train.py:903] (3/4) Epoch 4, batch 1800, loss[loss=0.3264, simple_loss=0.3718, pruned_loss=0.1406, over 19624.00 frames. ], tot_loss[loss=0.3052, simple_loss=0.3589, pruned_loss=0.1258, over 3816028.68 frames. ], batch size: 50, lr: 2.04e-02, grad_scale: 4.0 +2023-04-01 02:32:43,155 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.733e+02 7.313e+02 8.961e+02 1.128e+03 3.443e+03, threshold=1.792e+03, percent-clipped=8.0 +2023-04-01 02:32:43,403 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22327.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:32:44,448 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22328.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:32:46,336 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 02:32:48,651 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22332.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:32:52,008 INFO [train.py:903] (3/4) Epoch 4, batch 1850, loss[loss=0.264, simple_loss=0.3114, pruned_loss=0.1083, over 19720.00 frames. ], tot_loss[loss=0.3072, simple_loss=0.3604, pruned_loss=0.127, over 3823160.77 frames. ], batch size: 45, lr: 2.04e-02, grad_scale: 4.0 +2023-04-01 02:33:04,854 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22346.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:33:13,922 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22353.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:33:24,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 02:33:36,207 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22371.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:33:51,852 INFO [train.py:903] (3/4) Epoch 4, batch 1900, loss[loss=0.3079, simple_loss=0.3707, pruned_loss=0.1226, over 19588.00 frames. ], tot_loss[loss=0.3057, simple_loss=0.3594, pruned_loss=0.126, over 3834765.94 frames. ], batch size: 52, lr: 2.03e-02, grad_scale: 4.0 +2023-04-01 02:34:09,608 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 02:34:14,790 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 02:34:15,067 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22403.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:34:39,427 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 02:34:42,858 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.044e+02 7.549e+02 9.520e+02 1.192e+03 3.384e+03, threshold=1.904e+03, percent-clipped=5.0 +2023-04-01 02:34:51,934 INFO [train.py:903] (3/4) Epoch 4, batch 1950, loss[loss=0.3009, simple_loss=0.3622, pruned_loss=0.1198, over 18173.00 frames. ], tot_loss[loss=0.3072, simple_loss=0.3605, pruned_loss=0.1269, over 3812129.60 frames. ], batch size: 83, lr: 2.03e-02, grad_scale: 4.0 +2023-04-01 02:35:08,586 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22447.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:35:53,486 INFO [train.py:903] (3/4) Epoch 4, batch 2000, loss[loss=0.2953, simple_loss=0.3577, pruned_loss=0.1164, over 19663.00 frames. ], tot_loss[loss=0.3061, simple_loss=0.3598, pruned_loss=0.1263, over 3812959.25 frames. ], batch size: 55, lr: 2.03e-02, grad_scale: 8.0 +2023-04-01 02:36:01,833 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22492.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:36:06,665 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22496.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:36:38,284 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22521.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:36:45,511 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.709e+02 6.735e+02 8.799e+02 1.102e+03 2.294e+03, threshold=1.760e+03, percent-clipped=1.0 +2023-04-01 02:36:46,717 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 02:36:47,091 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1699, 1.1168, 1.4352, 0.4501, 2.4596, 2.4588, 2.2006, 2.5305], + device='cuda:3'), covar=tensor([0.1032, 0.2636, 0.2441, 0.1854, 0.0287, 0.0131, 0.0299, 0.0156], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0280, 0.0323, 0.0260, 0.0192, 0.0110, 0.0201, 0.0124], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 02:36:54,495 INFO [train.py:903] (3/4) Epoch 4, batch 2050, loss[loss=0.3192, simple_loss=0.3666, pruned_loss=0.1359, over 19579.00 frames. ], tot_loss[loss=0.3042, simple_loss=0.3583, pruned_loss=0.1251, over 3814050.74 frames. ], batch size: 52, lr: 2.03e-02, grad_scale: 8.0 +2023-04-01 02:36:58,174 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22538.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:37:04,886 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 02:37:06,034 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 02:37:20,839 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2198, 1.2858, 1.4254, 1.5607, 2.7408, 1.1917, 1.9091, 2.9324], + device='cuda:3'), covar=tensor([0.0416, 0.2450, 0.2336, 0.1503, 0.0526, 0.2061, 0.1132, 0.0418], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0308, 0.0304, 0.0278, 0.0292, 0.0320, 0.0283, 0.0289], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:37:27,446 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 02:37:45,380 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22577.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:37:52,339 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22583.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:37:54,200 INFO [train.py:903] (3/4) Epoch 4, batch 2100, loss[loss=0.2692, simple_loss=0.3228, pruned_loss=0.1078, over 19493.00 frames. ], tot_loss[loss=0.304, simple_loss=0.3579, pruned_loss=0.125, over 3818887.11 frames. ], batch size: 49, lr: 2.03e-02, grad_scale: 8.0 +2023-04-01 02:38:10,692 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22599.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:38:21,647 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 02:38:21,979 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22607.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:38:23,363 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22608.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:38:42,454 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 02:38:44,803 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.772e+02 6.954e+02 8.861e+02 1.126e+03 2.028e+03, threshold=1.772e+03, percent-clipped=3.0 +2023-04-01 02:38:53,763 INFO [train.py:903] (3/4) Epoch 4, batch 2150, loss[loss=0.2566, simple_loss=0.3121, pruned_loss=0.1006, over 19729.00 frames. ], tot_loss[loss=0.3032, simple_loss=0.3571, pruned_loss=0.1247, over 3830836.17 frames. ], batch size: 46, lr: 2.02e-02, grad_scale: 8.0 +2023-04-01 02:39:17,774 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22653.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:39:22,236 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22657.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:39:56,339 INFO [train.py:903] (3/4) Epoch 4, batch 2200, loss[loss=0.2972, simple_loss=0.3649, pruned_loss=0.1147, over 19795.00 frames. ], tot_loss[loss=0.3031, simple_loss=0.3576, pruned_loss=0.1244, over 3846166.97 frames. ], batch size: 56, lr: 2.02e-02, grad_scale: 8.0 +2023-04-01 02:40:05,453 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22692.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:40:17,997 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22703.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:40:30,190 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22714.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:40:47,466 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.077e+02 6.866e+02 8.417e+02 1.098e+03 2.160e+03, threshold=1.683e+03, percent-clipped=4.0 +2023-04-01 02:40:49,085 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22728.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:40:56,702 INFO [train.py:903] (3/4) Epoch 4, batch 2250, loss[loss=0.277, simple_loss=0.3443, pruned_loss=0.1049, over 19697.00 frames. ], tot_loss[loss=0.3034, simple_loss=0.3573, pruned_loss=0.1247, over 3837547.45 frames. ], batch size: 59, lr: 2.02e-02, grad_scale: 8.0 +2023-04-01 02:41:10,273 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=22747.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:41:44,873 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6414, 4.0979, 4.3248, 4.2489, 1.3763, 3.9337, 3.6037, 3.8416], + device='cuda:3'), covar=tensor([0.0819, 0.0527, 0.0482, 0.0385, 0.3853, 0.0306, 0.0466, 0.0972], + device='cuda:3'), in_proj_covar=tensor([0.0430, 0.0386, 0.0520, 0.0407, 0.0525, 0.0291, 0.0350, 0.0492], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:41:46,038 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3205, 1.8816, 2.0394, 2.3812, 2.1589, 2.1562, 2.1429, 2.3331], + device='cuda:3'), covar=tensor([0.0580, 0.1517, 0.0915, 0.0610, 0.0865, 0.0356, 0.0624, 0.0443], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0377, 0.0280, 0.0248, 0.0309, 0.0256, 0.0271, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:41:56,362 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3923, 1.2145, 1.8465, 1.5215, 3.0820, 2.7349, 3.5039, 1.4000], + device='cuda:3'), covar=tensor([0.1553, 0.2602, 0.1564, 0.1291, 0.0965, 0.1156, 0.1111, 0.2373], + device='cuda:3'), in_proj_covar=tensor([0.0431, 0.0481, 0.0451, 0.0402, 0.0524, 0.0423, 0.0606, 0.0431], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 02:41:57,018 INFO [train.py:903] (3/4) Epoch 4, batch 2300, loss[loss=0.2758, simple_loss=0.3259, pruned_loss=0.1128, over 19732.00 frames. ], tot_loss[loss=0.3022, simple_loss=0.3562, pruned_loss=0.1241, over 3841780.56 frames. ], batch size: 51, lr: 2.02e-02, grad_scale: 8.0 +2023-04-01 02:42:09,274 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 02:42:31,195 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22812.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:42:47,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.632e+02 7.223e+02 9.387e+02 1.200e+03 1.860e+03, threshold=1.877e+03, percent-clipped=8.0 +2023-04-01 02:42:56,885 INFO [train.py:903] (3/4) Epoch 4, batch 2350, loss[loss=0.3182, simple_loss=0.3679, pruned_loss=0.1342, over 19538.00 frames. ], tot_loss[loss=0.3016, simple_loss=0.3558, pruned_loss=0.1237, over 3838228.45 frames. ], batch size: 56, lr: 2.01e-02, grad_scale: 8.0 +2023-04-01 02:43:06,749 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22843.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:43:31,058 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22862.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:43:32,288 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22863.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:43:38,745 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 02:43:54,383 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 02:43:57,673 INFO [train.py:903] (3/4) Epoch 4, batch 2400, loss[loss=0.4174, simple_loss=0.4381, pruned_loss=0.1983, over 17644.00 frames. ], tot_loss[loss=0.3008, simple_loss=0.3552, pruned_loss=0.1232, over 3836857.68 frames. ], batch size: 101, lr: 2.01e-02, grad_scale: 8.0 +2023-04-01 02:44:03,331 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22888.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:44:27,700 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22909.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:44:37,903 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9229, 4.2716, 4.5821, 4.5013, 1.4790, 4.1078, 3.7190, 4.1326], + device='cuda:3'), covar=tensor([0.0706, 0.0549, 0.0438, 0.0328, 0.3712, 0.0278, 0.0431, 0.0816], + device='cuda:3'), in_proj_covar=tensor([0.0430, 0.0386, 0.0515, 0.0404, 0.0522, 0.0290, 0.0341, 0.0487], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 02:44:49,611 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.703e+02 7.425e+02 9.466e+02 1.182e+03 3.064e+03, threshold=1.893e+03, percent-clipped=2.0 +2023-04-01 02:44:58,891 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22934.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:44:59,531 INFO [train.py:903] (3/4) Epoch 4, batch 2450, loss[loss=0.3476, simple_loss=0.385, pruned_loss=0.1551, over 19538.00 frames. ], tot_loss[loss=0.3018, simple_loss=0.356, pruned_loss=0.1238, over 3824141.82 frames. ], batch size: 54, lr: 2.01e-02, grad_scale: 8.0 +2023-04-01 02:45:14,097 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22948.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:45:28,847 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-04-01 02:45:40,912 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22970.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:45:44,367 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22973.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:45:57,746 INFO [train.py:903] (3/4) Epoch 4, batch 2500, loss[loss=0.3847, simple_loss=0.4133, pruned_loss=0.178, over 19363.00 frames. ], tot_loss[loss=0.3033, simple_loss=0.3572, pruned_loss=0.1247, over 3822868.02 frames. ], batch size: 66, lr: 2.01e-02, grad_scale: 8.0 +2023-04-01 02:46:09,341 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22995.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:46:16,256 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23001.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:46:48,542 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.107e+02 7.308e+02 8.619e+02 1.083e+03 2.930e+03, threshold=1.724e+03, percent-clipped=3.0 +2023-04-01 02:46:57,732 INFO [train.py:903] (3/4) Epoch 4, batch 2550, loss[loss=0.3103, simple_loss=0.3681, pruned_loss=0.1263, over 17790.00 frames. ], tot_loss[loss=0.304, simple_loss=0.3579, pruned_loss=0.125, over 3821812.11 frames. ], batch size: 101, lr: 2.01e-02, grad_scale: 8.0 +2023-04-01 02:47:49,710 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 02:47:57,385 INFO [train.py:903] (3/4) Epoch 4, batch 2600, loss[loss=0.3202, simple_loss=0.3778, pruned_loss=0.1313, over 19776.00 frames. ], tot_loss[loss=0.3029, simple_loss=0.3567, pruned_loss=0.1245, over 3814123.94 frames. ], batch size: 56, lr: 2.00e-02, grad_scale: 8.0 +2023-04-01 02:48:33,782 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23116.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:48:36,146 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23118.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:48:46,614 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23126.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:48:48,299 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.619e+02 6.946e+02 8.555e+02 1.074e+03 2.756e+03, threshold=1.711e+03, percent-clipped=6.0 +2023-04-01 02:48:58,591 INFO [train.py:903] (3/4) Epoch 4, batch 2650, loss[loss=0.256, simple_loss=0.3126, pruned_loss=0.09965, over 19054.00 frames. ], tot_loss[loss=0.3023, simple_loss=0.3561, pruned_loss=0.1242, over 3805326.33 frames. ], batch size: 42, lr: 2.00e-02, grad_scale: 8.0 +2023-04-01 02:49:08,318 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23143.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:49:17,083 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 02:49:22,944 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23156.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:49:58,046 INFO [train.py:903] (3/4) Epoch 4, batch 2700, loss[loss=0.2963, simple_loss=0.3541, pruned_loss=0.1193, over 19535.00 frames. ], tot_loss[loss=0.302, simple_loss=0.3563, pruned_loss=0.1239, over 3822939.65 frames. ], batch size: 54, lr: 2.00e-02, grad_scale: 8.0 +2023-04-01 02:50:00,479 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23187.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:50:48,020 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.628e+02 7.002e+02 8.947e+02 1.091e+03 2.361e+03, threshold=1.789e+03, percent-clipped=7.0 +2023-04-01 02:50:57,147 INFO [train.py:903] (3/4) Epoch 4, batch 2750, loss[loss=0.3285, simple_loss=0.3788, pruned_loss=0.139, over 17421.00 frames. ], tot_loss[loss=0.3034, simple_loss=0.3572, pruned_loss=0.1248, over 3824808.66 frames. ], batch size: 101, lr: 2.00e-02, grad_scale: 8.0 +2023-04-01 02:50:57,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23235.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:51:11,506 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23247.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:51:41,063 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23271.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:51:56,453 INFO [train.py:903] (3/4) Epoch 4, batch 2800, loss[loss=0.3096, simple_loss=0.3737, pruned_loss=0.1228, over 19293.00 frames. ], tot_loss[loss=0.3059, simple_loss=0.3595, pruned_loss=0.1262, over 3804264.37 frames. ], batch size: 66, lr: 2.00e-02, grad_scale: 8.0 +2023-04-01 02:52:17,021 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:52:45,202 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.435e+02 7.909e+02 1.044e+03 1.347e+03 2.323e+03, threshold=2.087e+03, percent-clipped=7.0 +2023-04-01 02:52:56,810 INFO [train.py:903] (3/4) Epoch 4, batch 2850, loss[loss=0.3342, simple_loss=0.39, pruned_loss=0.1392, over 19643.00 frames. ], tot_loss[loss=0.3071, simple_loss=0.3606, pruned_loss=0.1268, over 3818153.22 frames. ], batch size: 55, lr: 1.99e-02, grad_scale: 8.0 +2023-04-01 02:53:03,168 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 02:53:41,875 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23372.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:53:44,370 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-01 02:53:56,271 INFO [train.py:903] (3/4) Epoch 4, batch 2900, loss[loss=0.3058, simple_loss=0.3695, pruned_loss=0.1211, over 19604.00 frames. ], tot_loss[loss=0.306, simple_loss=0.3598, pruned_loss=0.1261, over 3833611.97 frames. ], batch size: 57, lr: 1.99e-02, grad_scale: 4.0 +2023-04-01 02:53:56,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 02:54:10,019 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23397.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:54:45,629 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.212e+02 7.852e+02 1.023e+03 1.284e+03 2.319e+03, threshold=2.047e+03, percent-clipped=2.0 +2023-04-01 02:54:53,631 INFO [train.py:903] (3/4) Epoch 4, batch 2950, loss[loss=0.3373, simple_loss=0.3801, pruned_loss=0.1472, over 19669.00 frames. ], tot_loss[loss=0.3062, simple_loss=0.3596, pruned_loss=0.1264, over 3820793.53 frames. ], batch size: 60, lr: 1.99e-02, grad_scale: 4.0 +2023-04-01 02:55:20,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8899, 2.0572, 2.0365, 2.4251, 4.3340, 1.4444, 2.4284, 4.4050], + device='cuda:3'), covar=tensor([0.0243, 0.2241, 0.2159, 0.1265, 0.0433, 0.2096, 0.1212, 0.0277], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0304, 0.0304, 0.0276, 0.0297, 0.0319, 0.0280, 0.0286], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 02:55:35,117 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23470.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:55:45,360 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4512, 4.0285, 2.4766, 3.5803, 1.3211, 3.5274, 3.5855, 3.8627], + device='cuda:3'), covar=tensor([0.0612, 0.0934, 0.1888, 0.0793, 0.3496, 0.1095, 0.0758, 0.0820], + device='cuda:3'), in_proj_covar=tensor([0.0338, 0.0301, 0.0350, 0.0277, 0.0347, 0.0302, 0.0255, 0.0294], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:55:52,230 INFO [train.py:903] (3/4) Epoch 4, batch 3000, loss[loss=0.3254, simple_loss=0.3637, pruned_loss=0.1436, over 19832.00 frames. ], tot_loss[loss=0.3044, simple_loss=0.3587, pruned_loss=0.1251, over 3821538.32 frames. ], batch size: 52, lr: 1.99e-02, grad_scale: 4.0 +2023-04-01 02:55:52,230 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 02:56:05,134 INFO [train.py:937] (3/4) Epoch 4, validation: loss=0.2145, simple_loss=0.3118, pruned_loss=0.05862, over 944034.00 frames. +2023-04-01 02:56:05,135 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 02:56:09,824 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 02:56:32,408 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:56:56,227 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23527.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:56:56,888 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.281e+02 6.234e+02 7.977e+02 1.046e+03 2.333e+03, threshold=1.595e+03, percent-clipped=2.0 +2023-04-01 02:57:05,045 INFO [train.py:903] (3/4) Epoch 4, batch 3050, loss[loss=0.2301, simple_loss=0.293, pruned_loss=0.08365, over 19388.00 frames. ], tot_loss[loss=0.3027, simple_loss=0.3572, pruned_loss=0.1241, over 3829272.95 frames. ], batch size: 48, lr: 1.99e-02, grad_scale: 4.0 +2023-04-01 02:57:26,952 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23552.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:57:33,702 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23558.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:57:58,299 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23579.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:58:04,099 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23583.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:58:06,131 INFO [train.py:903] (3/4) Epoch 4, batch 3100, loss[loss=0.2917, simple_loss=0.3489, pruned_loss=0.1173, over 19577.00 frames. ], tot_loss[loss=0.3035, simple_loss=0.3576, pruned_loss=0.1247, over 3815884.24 frames. ], batch size: 52, lr: 1.98e-02, grad_scale: 4.0 +2023-04-01 02:58:06,491 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23585.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:58:13,283 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23591.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 02:58:55,944 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.118e+02 6.915e+02 8.546e+02 1.092e+03 2.878e+03, threshold=1.709e+03, percent-clipped=7.0 +2023-04-01 02:59:01,876 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8444, 4.3996, 2.6657, 3.9773, 1.2317, 4.1592, 4.0264, 4.1937], + device='cuda:3'), covar=tensor([0.0523, 0.1016, 0.1963, 0.0686, 0.4039, 0.0888, 0.0714, 0.0774], + device='cuda:3'), in_proj_covar=tensor([0.0337, 0.0297, 0.0346, 0.0276, 0.0344, 0.0294, 0.0257, 0.0291], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 02:59:03,926 INFO [train.py:903] (3/4) Epoch 4, batch 3150, loss[loss=0.2641, simple_loss=0.3277, pruned_loss=0.1002, over 19781.00 frames. ], tot_loss[loss=0.3036, simple_loss=0.3578, pruned_loss=0.1247, over 3828367.44 frames. ], batch size: 47, lr: 1.98e-02, grad_scale: 4.0 +2023-04-01 02:59:28,014 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 03:00:02,540 INFO [train.py:903] (3/4) Epoch 4, batch 3200, loss[loss=0.2423, simple_loss=0.3036, pruned_loss=0.09051, over 19776.00 frames. ], tot_loss[loss=0.3047, simple_loss=0.3585, pruned_loss=0.1254, over 3810065.15 frames. ], batch size: 46, lr: 1.98e-02, grad_scale: 8.0 +2023-04-01 03:00:13,229 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23694.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:00:29,411 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23706.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:00:44,988 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4595, 1.1501, 1.0413, 1.3078, 1.2387, 1.2822, 1.1052, 1.2334], + device='cuda:3'), covar=tensor([0.0911, 0.1153, 0.1270, 0.0828, 0.0841, 0.0528, 0.0987, 0.0721], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0376, 0.0279, 0.0246, 0.0308, 0.0259, 0.0270, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:00:53,667 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.799e+02 7.379e+02 9.197e+02 1.143e+03 1.957e+03, threshold=1.839e+03, percent-clipped=5.0 +2023-04-01 03:01:02,112 INFO [train.py:903] (3/4) Epoch 4, batch 3250, loss[loss=0.2651, simple_loss=0.3406, pruned_loss=0.09482, over 19689.00 frames. ], tot_loss[loss=0.3057, simple_loss=0.3595, pruned_loss=0.126, over 3795494.75 frames. ], batch size: 59, lr: 1.98e-02, grad_scale: 8.0 +2023-04-01 03:01:51,726 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5120, 1.2486, 1.9129, 1.4813, 2.7216, 4.4415, 4.4156, 4.8673], + device='cuda:3'), covar=tensor([0.1408, 0.2810, 0.2704, 0.1801, 0.0483, 0.0077, 0.0124, 0.0063], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0278, 0.0318, 0.0258, 0.0188, 0.0105, 0.0201, 0.0125], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 03:02:01,902 INFO [train.py:903] (3/4) Epoch 4, batch 3300, loss[loss=0.2521, simple_loss=0.3176, pruned_loss=0.09335, over 19359.00 frames. ], tot_loss[loss=0.3046, simple_loss=0.3582, pruned_loss=0.1255, over 3802681.57 frames. ], batch size: 47, lr: 1.98e-02, grad_scale: 8.0 +2023-04-01 03:02:04,051 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 03:02:30,682 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8738, 4.2393, 4.6443, 4.4889, 1.4524, 4.1931, 3.6379, 4.1970], + device='cuda:3'), covar=tensor([0.0894, 0.0546, 0.0455, 0.0373, 0.4303, 0.0363, 0.0457, 0.0930], + device='cuda:3'), in_proj_covar=tensor([0.0449, 0.0394, 0.0537, 0.0419, 0.0539, 0.0304, 0.0347, 0.0500], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 03:02:54,105 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.057e+02 7.772e+02 9.614e+02 1.210e+03 2.492e+03, threshold=1.923e+03, percent-clipped=5.0 +2023-04-01 03:03:02,121 INFO [train.py:903] (3/4) Epoch 4, batch 3350, loss[loss=0.3382, simple_loss=0.3829, pruned_loss=0.1467, over 19486.00 frames. ], tot_loss[loss=0.3045, simple_loss=0.3579, pruned_loss=0.1255, over 3809547.09 frames. ], batch size: 64, lr: 1.97e-02, grad_scale: 8.0 +2023-04-01 03:03:09,290 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:03:19,211 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=23850.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:03:40,414 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23866.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:04:01,844 INFO [train.py:903] (3/4) Epoch 4, batch 3400, loss[loss=0.3817, simple_loss=0.4122, pruned_loss=0.1756, over 17557.00 frames. ], tot_loss[loss=0.3023, simple_loss=0.3562, pruned_loss=0.1242, over 3816026.91 frames. ], batch size: 101, lr: 1.97e-02, grad_scale: 8.0 +2023-04-01 03:04:51,950 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6352, 1.4357, 1.2128, 1.6800, 1.4022, 1.5425, 1.4391, 1.4192], + device='cuda:3'), covar=tensor([0.0850, 0.1260, 0.1244, 0.0712, 0.0948, 0.0481, 0.0865, 0.0702], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0372, 0.0281, 0.0245, 0.0313, 0.0259, 0.0272, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:04:53,667 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.081e+02 7.350e+02 9.318e+02 1.202e+03 2.145e+03, threshold=1.864e+03, percent-clipped=3.0 +2023-04-01 03:05:01,736 INFO [train.py:903] (3/4) Epoch 4, batch 3450, loss[loss=0.2684, simple_loss=0.3289, pruned_loss=0.104, over 19845.00 frames. ], tot_loss[loss=0.3013, simple_loss=0.3555, pruned_loss=0.1236, over 3814549.27 frames. ], batch size: 52, lr: 1.97e-02, grad_scale: 8.0 +2023-04-01 03:05:01,760 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 03:05:22,667 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23950.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:05:36,157 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23962.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:05:39,449 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23965.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:05:50,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23975.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:06:04,981 INFO [train.py:903] (3/4) Epoch 4, batch 3500, loss[loss=0.2361, simple_loss=0.3023, pruned_loss=0.08497, over 19293.00 frames. ], tot_loss[loss=0.3032, simple_loss=0.3568, pruned_loss=0.1248, over 3802875.96 frames. ], batch size: 44, lr: 1.97e-02, grad_scale: 8.0 +2023-04-01 03:06:07,801 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23987.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:06:58,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.373e+02 7.196e+02 8.612e+02 1.120e+03 2.630e+03, threshold=1.722e+03, percent-clipped=5.0 +2023-04-01 03:07:06,277 INFO [train.py:903] (3/4) Epoch 4, batch 3550, loss[loss=0.262, simple_loss=0.3182, pruned_loss=0.1029, over 19740.00 frames. ], tot_loss[loss=0.3006, simple_loss=0.355, pruned_loss=0.1231, over 3820313.38 frames. ], batch size: 46, lr: 1.97e-02, grad_scale: 8.0 +2023-04-01 03:07:28,455 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24055.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:07:45,474 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1514, 5.4942, 3.0248, 4.9617, 1.7182, 5.4989, 5.3405, 5.5102], + device='cuda:3'), covar=tensor([0.0419, 0.0913, 0.1715, 0.0499, 0.3247, 0.0640, 0.0612, 0.0546], + device='cuda:3'), in_proj_covar=tensor([0.0333, 0.0297, 0.0345, 0.0280, 0.0350, 0.0299, 0.0263, 0.0295], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:07:57,066 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.41 vs. limit=5.0 +2023-04-01 03:08:05,263 INFO [train.py:903] (3/4) Epoch 4, batch 3600, loss[loss=0.3715, simple_loss=0.412, pruned_loss=0.1655, over 19284.00 frames. ], tot_loss[loss=0.3026, simple_loss=0.3568, pruned_loss=0.1242, over 3810966.37 frames. ], batch size: 66, lr: 1.96e-02, grad_scale: 8.0 +2023-04-01 03:08:56,916 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.624e+02 7.148e+02 8.733e+02 1.077e+03 2.339e+03, threshold=1.747e+03, percent-clipped=4.0 +2023-04-01 03:09:04,858 INFO [train.py:903] (3/4) Epoch 4, batch 3650, loss[loss=0.2346, simple_loss=0.3128, pruned_loss=0.0782, over 19583.00 frames. ], tot_loss[loss=0.3024, simple_loss=0.3566, pruned_loss=0.1241, over 3817114.97 frames. ], batch size: 52, lr: 1.96e-02, grad_scale: 8.0 +2023-04-01 03:09:34,931 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2852, 2.2763, 2.0995, 3.4663, 2.3497, 3.7158, 3.5521, 2.0571], + device='cuda:3'), covar=tensor([0.1688, 0.1210, 0.0677, 0.0770, 0.1454, 0.0351, 0.0910, 0.1024], + device='cuda:3'), in_proj_covar=tensor([0.0563, 0.0541, 0.0512, 0.0707, 0.0601, 0.0456, 0.0611, 0.0522], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:10:05,498 INFO [train.py:903] (3/4) Epoch 4, batch 3700, loss[loss=0.2961, simple_loss=0.3319, pruned_loss=0.1302, over 19764.00 frames. ], tot_loss[loss=0.3046, simple_loss=0.3583, pruned_loss=0.1254, over 3816880.50 frames. ], batch size: 45, lr: 1.96e-02, grad_scale: 4.0 +2023-04-01 03:10:16,642 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2177, 1.2556, 1.8326, 1.3609, 2.6091, 2.1830, 2.7232, 1.0085], + device='cuda:3'), covar=tensor([0.1515, 0.2367, 0.1256, 0.1270, 0.0916, 0.1143, 0.1046, 0.2221], + device='cuda:3'), in_proj_covar=tensor([0.0434, 0.0481, 0.0451, 0.0398, 0.0525, 0.0423, 0.0602, 0.0434], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:10:38,062 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9403, 1.9232, 1.7720, 2.8645, 1.9235, 2.7304, 2.5885, 1.7883], + device='cuda:3'), covar=tensor([0.1525, 0.1151, 0.0702, 0.0652, 0.1352, 0.0427, 0.1048, 0.1069], + device='cuda:3'), in_proj_covar=tensor([0.0567, 0.0536, 0.0512, 0.0700, 0.0601, 0.0457, 0.0610, 0.0521], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:10:48,459 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24221.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:10:58,950 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.206e+02 7.207e+02 9.640e+02 1.134e+03 2.323e+03, threshold=1.928e+03, percent-clipped=6.0 +2023-04-01 03:11:07,277 INFO [train.py:903] (3/4) Epoch 4, batch 3750, loss[loss=0.3177, simple_loss=0.3734, pruned_loss=0.1311, over 19789.00 frames. ], tot_loss[loss=0.3047, simple_loss=0.3583, pruned_loss=0.1256, over 3816205.74 frames. ], batch size: 56, lr: 1.96e-02, grad_scale: 4.0 +2023-04-01 03:11:20,277 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24246.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:12:07,328 INFO [train.py:903] (3/4) Epoch 4, batch 3800, loss[loss=0.2836, simple_loss=0.3552, pruned_loss=0.106, over 19603.00 frames. ], tot_loss[loss=0.3038, simple_loss=0.3575, pruned_loss=0.1251, over 3818088.15 frames. ], batch size: 57, lr: 1.96e-02, grad_scale: 4.0 +2023-04-01 03:12:38,499 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 03:12:47,446 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2607, 1.1973, 1.5514, 1.2562, 2.1041, 1.9485, 2.2193, 0.8886], + device='cuda:3'), covar=tensor([0.1570, 0.2604, 0.1426, 0.1421, 0.1058, 0.1324, 0.1073, 0.2335], + device='cuda:3'), in_proj_covar=tensor([0.0433, 0.0484, 0.0454, 0.0401, 0.0525, 0.0426, 0.0608, 0.0435], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:12:53,008 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.73 vs. limit=5.0 +2023-04-01 03:13:00,201 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.239e+02 7.507e+02 9.076e+02 1.248e+03 3.254e+03, threshold=1.815e+03, percent-clipped=3.0 +2023-04-01 03:13:07,208 INFO [train.py:903] (3/4) Epoch 4, batch 3850, loss[loss=0.2486, simple_loss=0.3261, pruned_loss=0.08556, over 19662.00 frames. ], tot_loss[loss=0.3027, simple_loss=0.3568, pruned_loss=0.1242, over 3816102.80 frames. ], batch size: 58, lr: 1.95e-02, grad_scale: 4.0 +2023-04-01 03:13:38,878 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6728, 1.4275, 1.4702, 1.9232, 1.7215, 1.7196, 1.5103, 1.7159], + device='cuda:3'), covar=tensor([0.0849, 0.1411, 0.1176, 0.0758, 0.0957, 0.0427, 0.0882, 0.0606], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0373, 0.0282, 0.0249, 0.0309, 0.0254, 0.0272, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:14:06,605 INFO [train.py:903] (3/4) Epoch 4, batch 3900, loss[loss=0.2991, simple_loss=0.3609, pruned_loss=0.1187, over 19681.00 frames. ], tot_loss[loss=0.3025, simple_loss=0.3572, pruned_loss=0.1239, over 3824854.38 frames. ], batch size: 58, lr: 1.95e-02, grad_scale: 4.0 +2023-04-01 03:14:11,569 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.71 vs. limit=2.0 +2023-04-01 03:14:25,952 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=24399.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:14:27,418 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1213, 1.1327, 1.4005, 1.2321, 1.6875, 1.7297, 1.8480, 0.4388], + device='cuda:3'), covar=tensor([0.1565, 0.2442, 0.1322, 0.1403, 0.1045, 0.1382, 0.1019, 0.2373], + device='cuda:3'), in_proj_covar=tensor([0.0439, 0.0487, 0.0460, 0.0407, 0.0531, 0.0428, 0.0617, 0.0437], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:15:00,806 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.463e+02 8.238e+02 9.729e+02 1.230e+03 4.971e+03, threshold=1.946e+03, percent-clipped=9.0 +2023-04-01 03:15:09,421 INFO [train.py:903] (3/4) Epoch 4, batch 3950, loss[loss=0.2707, simple_loss=0.3315, pruned_loss=0.1049, over 19750.00 frames. ], tot_loss[loss=0.302, simple_loss=0.3568, pruned_loss=0.1236, over 3835884.19 frames. ], batch size: 51, lr: 1.95e-02, grad_scale: 4.0 +2023-04-01 03:15:17,175 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 03:16:10,710 INFO [train.py:903] (3/4) Epoch 4, batch 4000, loss[loss=0.2929, simple_loss=0.3493, pruned_loss=0.1183, over 19576.00 frames. ], tot_loss[loss=0.302, simple_loss=0.3571, pruned_loss=0.1234, over 3837695.12 frames. ], batch size: 52, lr: 1.95e-02, grad_scale: 8.0 +2023-04-01 03:16:45,854 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=24514.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:16:58,859 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 03:17:03,419 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.288e+02 6.850e+02 8.525e+02 1.041e+03 2.187e+03, threshold=1.705e+03, percent-clipped=1.0 +2023-04-01 03:17:08,329 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.7039, 0.9048, 0.6919, 0.6885, 0.8466, 0.5338, 0.5053, 0.8500], + device='cuda:3'), covar=tensor([0.0319, 0.0362, 0.0602, 0.0304, 0.0284, 0.0715, 0.0429, 0.0291], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0249, 0.0312, 0.0241, 0.0217, 0.0308, 0.0278, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:17:09,918 INFO [train.py:903] (3/4) Epoch 4, batch 4050, loss[loss=0.2806, simple_loss=0.3343, pruned_loss=0.1135, over 19383.00 frames. ], tot_loss[loss=0.3019, simple_loss=0.3566, pruned_loss=0.1235, over 3826264.06 frames. ], batch size: 48, lr: 1.95e-02, grad_scale: 8.0 +2023-04-01 03:17:47,527 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24565.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:17:47,602 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5544, 1.4370, 1.7169, 1.6776, 3.2019, 4.4343, 4.5057, 4.8537], + device='cuda:3'), covar=tensor([0.1326, 0.2636, 0.2749, 0.1603, 0.0352, 0.0119, 0.0131, 0.0054], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0283, 0.0325, 0.0260, 0.0192, 0.0110, 0.0205, 0.0126], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 03:18:10,059 INFO [train.py:903] (3/4) Epoch 4, batch 4100, loss[loss=0.3281, simple_loss=0.3844, pruned_loss=0.1359, over 19513.00 frames. ], tot_loss[loss=0.3014, simple_loss=0.3566, pruned_loss=0.1231, over 3822502.68 frames. ], batch size: 70, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:18:12,590 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24587.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:18:49,628 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 03:19:04,404 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.283e+02 6.813e+02 8.989e+02 1.047e+03 2.179e+03, threshold=1.798e+03, percent-clipped=2.0 +2023-04-01 03:19:12,697 INFO [train.py:903] (3/4) Epoch 4, batch 4150, loss[loss=0.3519, simple_loss=0.4026, pruned_loss=0.1506, over 19683.00 frames. ], tot_loss[loss=0.298, simple_loss=0.3536, pruned_loss=0.1212, over 3832170.56 frames. ], batch size: 59, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:19:19,920 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9611, 1.9764, 1.5270, 1.4927, 1.3547, 1.4887, 0.3673, 0.8930], + device='cuda:3'), covar=tensor([0.0213, 0.0195, 0.0163, 0.0215, 0.0453, 0.0266, 0.0431, 0.0389], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0270, 0.0263, 0.0288, 0.0354, 0.0277, 0.0271, 0.0282], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:20:13,433 INFO [train.py:903] (3/4) Epoch 4, batch 4200, loss[loss=0.2929, simple_loss=0.3519, pruned_loss=0.117, over 19768.00 frames. ], tot_loss[loss=0.2981, simple_loss=0.3538, pruned_loss=0.1211, over 3831094.04 frames. ], batch size: 63, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:20:19,920 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 03:20:31,296 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8543, 1.8452, 1.7785, 2.7998, 1.6220, 2.3803, 2.3866, 1.7054], + device='cuda:3'), covar=tensor([0.1478, 0.1108, 0.0672, 0.0622, 0.1465, 0.0494, 0.1118, 0.1110], + device='cuda:3'), in_proj_covar=tensor([0.0566, 0.0540, 0.0514, 0.0707, 0.0607, 0.0462, 0.0615, 0.0527], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:20:33,582 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2585, 2.2383, 1.5684, 1.6996, 2.0919, 0.9791, 1.0392, 1.6468], + device='cuda:3'), covar=tensor([0.0783, 0.0459, 0.0860, 0.0415, 0.0412, 0.1066, 0.0670, 0.0387], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0258, 0.0321, 0.0249, 0.0219, 0.0315, 0.0286, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:21:05,885 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.764e+02 7.220e+02 8.850e+02 1.090e+03 2.101e+03, threshold=1.770e+03, percent-clipped=3.0 +2023-04-01 03:21:12,814 INFO [train.py:903] (3/4) Epoch 4, batch 4250, loss[loss=0.2928, simple_loss=0.3459, pruned_loss=0.1198, over 19616.00 frames. ], tot_loss[loss=0.299, simple_loss=0.3546, pruned_loss=0.1217, over 3814275.75 frames. ], batch size: 50, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:21:29,767 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 03:21:41,556 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 03:21:56,492 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24770.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:22:10,842 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24783.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:22:12,642 INFO [train.py:903] (3/4) Epoch 4, batch 4300, loss[loss=0.3007, simple_loss=0.3658, pruned_loss=0.1178, over 19658.00 frames. ], tot_loss[loss=0.2988, simple_loss=0.3546, pruned_loss=0.1215, over 3836278.37 frames. ], batch size: 60, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:22:26,945 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24795.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:22:31,625 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1432, 2.6841, 1.8862, 2.2823, 1.8777, 2.0723, 0.6980, 2.0484], + device='cuda:3'), covar=tensor([0.0245, 0.0227, 0.0247, 0.0371, 0.0464, 0.0429, 0.0576, 0.0438], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0272, 0.0264, 0.0288, 0.0353, 0.0278, 0.0269, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:22:54,057 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4107, 2.3576, 2.1135, 3.6933, 2.3278, 3.8404, 3.3756, 2.0582], + device='cuda:3'), covar=tensor([0.1673, 0.1203, 0.0684, 0.0724, 0.1527, 0.0328, 0.1031, 0.1079], + device='cuda:3'), in_proj_covar=tensor([0.0568, 0.0544, 0.0512, 0.0706, 0.0605, 0.0458, 0.0618, 0.0525], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:23:06,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.009e+02 7.093e+02 8.869e+02 1.163e+03 2.104e+03, threshold=1.774e+03, percent-clipped=1.0 +2023-04-01 03:23:08,952 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 03:23:14,538 INFO [train.py:903] (3/4) Epoch 4, batch 4350, loss[loss=0.2651, simple_loss=0.3231, pruned_loss=0.1035, over 19727.00 frames. ], tot_loss[loss=0.2981, simple_loss=0.3538, pruned_loss=0.1212, over 3831375.45 frames. ], batch size: 45, lr: 1.94e-02, grad_scale: 8.0 +2023-04-01 03:23:21,138 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0951, 1.0823, 1.3486, 1.2057, 1.6931, 1.7061, 1.9149, 0.4450], + device='cuda:3'), covar=tensor([0.1691, 0.2748, 0.1470, 0.1466, 0.1115, 0.1477, 0.1042, 0.2641], + device='cuda:3'), in_proj_covar=tensor([0.0433, 0.0481, 0.0452, 0.0405, 0.0525, 0.0421, 0.0603, 0.0430], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:23:52,227 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24866.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:24:15,086 INFO [train.py:903] (3/4) Epoch 4, batch 4400, loss[loss=0.2731, simple_loss=0.3327, pruned_loss=0.1068, over 19384.00 frames. ], tot_loss[loss=0.2989, simple_loss=0.3546, pruned_loss=0.1216, over 3806477.08 frames. ], batch size: 48, lr: 1.93e-02, grad_scale: 8.0 +2023-04-01 03:24:21,307 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-01 03:24:40,901 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 03:24:44,304 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=24909.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:24:50,761 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 03:25:09,651 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.331e+02 8.078e+02 9.889e+02 1.280e+03 3.768e+03, threshold=1.978e+03, percent-clipped=10.0 +2023-04-01 03:25:12,026 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=24931.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:25:16,506 INFO [train.py:903] (3/4) Epoch 4, batch 4450, loss[loss=0.2636, simple_loss=0.334, pruned_loss=0.09656, over 19775.00 frames. ], tot_loss[loss=0.2993, simple_loss=0.3546, pruned_loss=0.122, over 3806643.90 frames. ], batch size: 54, lr: 1.93e-02, grad_scale: 8.0 +2023-04-01 03:26:17,094 INFO [train.py:903] (3/4) Epoch 4, batch 4500, loss[loss=0.2536, simple_loss=0.3225, pruned_loss=0.09241, over 19593.00 frames. ], tot_loss[loss=0.298, simple_loss=0.3538, pruned_loss=0.1211, over 3814941.64 frames. ], batch size: 50, lr: 1.93e-02, grad_scale: 8.0 +2023-04-01 03:26:21,826 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0472, 1.9995, 1.8970, 2.8723, 2.0511, 2.8043, 2.5514, 1.8795], + device='cuda:3'), covar=tensor([0.1420, 0.1126, 0.0677, 0.0718, 0.1278, 0.0435, 0.1095, 0.1059], + device='cuda:3'), in_proj_covar=tensor([0.0573, 0.0552, 0.0518, 0.0716, 0.0608, 0.0462, 0.0625, 0.0533], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:26:52,456 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25013.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:27:04,658 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25024.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:27:11,028 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.968e+02 6.473e+02 7.865e+02 1.057e+03 2.211e+03, threshold=1.573e+03, percent-clipped=1.0 +2023-04-01 03:27:18,537 INFO [train.py:903] (3/4) Epoch 4, batch 4550, loss[loss=0.2527, simple_loss=0.3117, pruned_loss=0.09682, over 18635.00 frames. ], tot_loss[loss=0.2989, simple_loss=0.3543, pruned_loss=0.1218, over 3812666.55 frames. ], batch size: 41, lr: 1.93e-02, grad_scale: 8.0 +2023-04-01 03:27:27,108 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 03:27:31,925 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25046.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:27:50,568 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 03:27:54,650 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 03:28:18,970 INFO [train.py:903] (3/4) Epoch 4, batch 4600, loss[loss=0.2816, simple_loss=0.3339, pruned_loss=0.1147, over 19414.00 frames. ], tot_loss[loss=0.2992, simple_loss=0.3545, pruned_loss=0.1219, over 3809579.02 frames. ], batch size: 48, lr: 1.93e-02, grad_scale: 8.0 +2023-04-01 03:29:10,701 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=25127.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:29:12,893 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.682e+02 7.418e+02 9.211e+02 1.176e+03 2.853e+03, threshold=1.842e+03, percent-clipped=7.0 +2023-04-01 03:29:20,334 INFO [train.py:903] (3/4) Epoch 4, batch 4650, loss[loss=0.3158, simple_loss=0.3664, pruned_loss=0.1326, over 19772.00 frames. ], tot_loss[loss=0.3001, simple_loss=0.355, pruned_loss=0.1226, over 3815759.28 frames. ], batch size: 54, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:29:37,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 03:29:46,700 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 03:30:19,358 INFO [train.py:903] (3/4) Epoch 4, batch 4700, loss[loss=0.3913, simple_loss=0.425, pruned_loss=0.1788, over 19631.00 frames. ], tot_loss[loss=0.3007, simple_loss=0.3552, pruned_loss=0.1231, over 3823825.88 frames. ], batch size: 61, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:30:42,828 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 03:30:47,461 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7536, 4.1339, 4.4268, 4.3959, 1.4303, 3.8488, 3.5023, 4.0085], + device='cuda:3'), covar=tensor([0.0964, 0.0581, 0.0521, 0.0411, 0.4100, 0.0380, 0.0521, 0.1017], + device='cuda:3'), in_proj_covar=tensor([0.0456, 0.0397, 0.0533, 0.0423, 0.0528, 0.0307, 0.0342, 0.0502], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 03:30:50,633 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=25210.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:31:13,782 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.903e+02 7.587e+02 9.394e+02 1.259e+03 3.233e+03, threshold=1.879e+03, percent-clipped=11.0 +2023-04-01 03:31:21,436 INFO [train.py:903] (3/4) Epoch 4, batch 4750, loss[loss=0.2938, simple_loss=0.3481, pruned_loss=0.1198, over 19762.00 frames. ], tot_loss[loss=0.3013, simple_loss=0.3557, pruned_loss=0.1234, over 3827811.71 frames. ], batch size: 54, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:31:24,252 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2023, 1.3120, 1.8768, 1.5009, 2.5644, 2.2283, 2.7604, 0.9556], + device='cuda:3'), covar=tensor([0.1690, 0.2655, 0.1450, 0.1309, 0.1165, 0.1262, 0.1247, 0.2458], + device='cuda:3'), in_proj_covar=tensor([0.0437, 0.0480, 0.0454, 0.0404, 0.0529, 0.0426, 0.0604, 0.0427], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:31:30,726 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25242.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:31:39,441 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6954, 1.2196, 1.2329, 1.6059, 1.4208, 1.5253, 1.3821, 1.4859], + device='cuda:3'), covar=tensor([0.0825, 0.1294, 0.1228, 0.0843, 0.1004, 0.0485, 0.0867, 0.0644], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0376, 0.0284, 0.0250, 0.0312, 0.0262, 0.0275, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:31:44,150 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-01 03:32:16,226 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25280.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:32:21,434 INFO [train.py:903] (3/4) Epoch 4, batch 4800, loss[loss=0.3215, simple_loss=0.3727, pruned_loss=0.1352, over 13418.00 frames. ], tot_loss[loss=0.2993, simple_loss=0.3541, pruned_loss=0.1222, over 3812809.87 frames. ], batch size: 136, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:32:27,815 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-01 03:32:41,282 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:32:44,310 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25305.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:33:09,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25325.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:33:11,914 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25327.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:33:13,791 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.433e+02 7.095e+02 8.715e+02 1.261e+03 2.828e+03, threshold=1.743e+03, percent-clipped=4.0 +2023-04-01 03:33:21,567 INFO [train.py:903] (3/4) Epoch 4, batch 4850, loss[loss=0.2577, simple_loss=0.3194, pruned_loss=0.09797, over 19843.00 frames. ], tot_loss[loss=0.2985, simple_loss=0.3535, pruned_loss=0.1218, over 3819278.80 frames. ], batch size: 52, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:33:28,356 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3716, 2.2370, 1.6579, 1.6998, 1.6483, 1.7068, 0.3467, 1.2135], + device='cuda:3'), covar=tensor([0.0235, 0.0229, 0.0202, 0.0290, 0.0498, 0.0327, 0.0545, 0.0416], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0271, 0.0266, 0.0289, 0.0355, 0.0285, 0.0272, 0.0280], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:33:45,829 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 03:33:48,283 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=25357.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:34:04,854 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 03:34:11,327 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 03:34:12,496 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 03:34:21,579 INFO [train.py:903] (3/4) Epoch 4, batch 4900, loss[loss=0.3273, simple_loss=0.3757, pruned_loss=0.1395, over 18273.00 frames. ], tot_loss[loss=0.2984, simple_loss=0.3535, pruned_loss=0.1217, over 3821983.42 frames. ], batch size: 84, lr: 1.92e-02, grad_scale: 8.0 +2023-04-01 03:34:21,587 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 03:34:34,900 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8135, 1.2960, 2.1410, 1.6756, 2.9250, 4.4755, 4.5352, 4.9119], + device='cuda:3'), covar=tensor([0.1168, 0.2631, 0.2293, 0.1578, 0.0416, 0.0117, 0.0116, 0.0058], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0278, 0.0317, 0.0257, 0.0195, 0.0113, 0.0201, 0.0126], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 03:34:41,705 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 03:34:43,939 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4731, 4.0384, 2.3502, 3.6294, 0.9874, 3.5610, 3.6122, 3.8409], + device='cuda:3'), covar=tensor([0.0582, 0.1162, 0.2110, 0.0704, 0.3956, 0.1043, 0.0723, 0.0875], + device='cuda:3'), in_proj_covar=tensor([0.0334, 0.0294, 0.0354, 0.0278, 0.0349, 0.0301, 0.0269, 0.0300], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:35:16,121 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.171e+02 7.094e+02 8.660e+02 1.069e+03 1.655e+03, threshold=1.732e+03, percent-clipped=0.0 +2023-04-01 03:35:23,611 INFO [train.py:903] (3/4) Epoch 4, batch 4950, loss[loss=0.2675, simple_loss=0.3264, pruned_loss=0.1043, over 19612.00 frames. ], tot_loss[loss=0.299, simple_loss=0.3546, pruned_loss=0.1217, over 3815342.88 frames. ], batch size: 50, lr: 1.91e-02, grad_scale: 8.0 +2023-04-01 03:35:37,195 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25446.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:35:41,246 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 03:36:04,441 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 03:36:04,788 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7982, 1.3223, 1.2661, 2.0964, 1.6521, 1.8926, 2.0549, 1.7320], + device='cuda:3'), covar=tensor([0.0680, 0.0999, 0.1081, 0.0758, 0.0833, 0.0686, 0.0827, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0244, 0.0262, 0.0254, 0.0288, 0.0286, 0.0244, 0.0257, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 03:36:09,084 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25472.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:36:24,197 INFO [train.py:903] (3/4) Epoch 4, batch 5000, loss[loss=0.3126, simple_loss=0.3451, pruned_loss=0.14, over 19778.00 frames. ], tot_loss[loss=0.3002, simple_loss=0.3557, pruned_loss=0.1224, over 3801132.07 frames. ], batch size: 48, lr: 1.91e-02, grad_scale: 8.0 +2023-04-01 03:36:33,161 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 03:36:40,138 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25498.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:36:44,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 03:37:11,000 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25523.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:37:17,559 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.118e+02 6.823e+02 8.824e+02 1.078e+03 2.588e+03, threshold=1.765e+03, percent-clipped=9.0 +2023-04-01 03:37:24,344 INFO [train.py:903] (3/4) Epoch 4, batch 5050, loss[loss=0.3032, simple_loss=0.3674, pruned_loss=0.1195, over 19547.00 frames. ], tot_loss[loss=0.3001, simple_loss=0.3555, pruned_loss=0.1224, over 3809851.28 frames. ], batch size: 56, lr: 1.91e-02, grad_scale: 8.0 +2023-04-01 03:38:02,119 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 03:38:06,290 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.40 vs. limit=5.0 +2023-04-01 03:38:21,775 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25581.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:38:25,844 INFO [train.py:903] (3/4) Epoch 4, batch 5100, loss[loss=0.3394, simple_loss=0.3915, pruned_loss=0.1436, over 19073.00 frames. ], tot_loss[loss=0.2988, simple_loss=0.3543, pruned_loss=0.1216, over 3809714.35 frames. ], batch size: 69, lr: 1.91e-02, grad_scale: 8.0 +2023-04-01 03:38:37,692 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 03:38:40,928 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 03:38:44,278 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 03:38:52,402 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:39:19,438 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.903e+02 6.354e+02 8.688e+02 1.168e+03 2.387e+03, threshold=1.738e+03, percent-clipped=4.0 +2023-04-01 03:39:27,220 INFO [train.py:903] (3/4) Epoch 4, batch 5150, loss[loss=0.2419, simple_loss=0.3, pruned_loss=0.09194, over 19783.00 frames. ], tot_loss[loss=0.299, simple_loss=0.3538, pruned_loss=0.1221, over 3802487.39 frames. ], batch size: 48, lr: 1.91e-02, grad_scale: 8.0 +2023-04-01 03:39:39,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 03:39:55,239 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8187, 1.8700, 1.5862, 2.8382, 2.0200, 2.7306, 2.3133, 1.2187], + device='cuda:3'), covar=tensor([0.2067, 0.1664, 0.1229, 0.0956, 0.1671, 0.0625, 0.1871, 0.2192], + device='cuda:3'), in_proj_covar=tensor([0.0567, 0.0549, 0.0515, 0.0708, 0.0613, 0.0469, 0.0622, 0.0525], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:39:58,724 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3077, 2.8767, 2.1141, 2.3169, 2.0246, 2.2768, 0.6485, 2.3965], + device='cuda:3'), covar=tensor([0.0240, 0.0244, 0.0233, 0.0315, 0.0442, 0.0338, 0.0535, 0.0359], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0274, 0.0265, 0.0295, 0.0355, 0.0283, 0.0274, 0.0280], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:39:59,816 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9473, 1.1998, 1.3522, 1.4357, 2.5333, 1.0216, 1.7939, 2.6008], + device='cuda:3'), covar=tensor([0.0481, 0.2444, 0.2243, 0.1472, 0.0604, 0.2097, 0.1101, 0.0460], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0308, 0.0306, 0.0280, 0.0304, 0.0316, 0.0285, 0.0291], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 03:40:13,785 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 03:40:28,111 INFO [train.py:903] (3/4) Epoch 4, batch 5200, loss[loss=0.267, simple_loss=0.3286, pruned_loss=0.1027, over 19678.00 frames. ], tot_loss[loss=0.2986, simple_loss=0.3538, pruned_loss=0.1216, over 3805857.45 frames. ], batch size: 53, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:40:29,757 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2143, 1.1819, 1.9398, 1.4485, 2.8978, 2.5333, 3.1455, 1.2491], + device='cuda:3'), covar=tensor([0.1976, 0.3019, 0.1587, 0.1492, 0.1252, 0.1434, 0.1535, 0.2846], + device='cuda:3'), in_proj_covar=tensor([0.0453, 0.0499, 0.0462, 0.0408, 0.0539, 0.0443, 0.0620, 0.0437], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:40:42,797 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 03:41:20,992 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25728.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:41:21,773 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.810e+02 7.304e+02 9.145e+02 1.165e+03 2.884e+03, threshold=1.829e+03, percent-clipped=6.0 +2023-04-01 03:41:25,403 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 03:41:28,710 INFO [train.py:903] (3/4) Epoch 4, batch 5250, loss[loss=0.3306, simple_loss=0.3784, pruned_loss=0.1414, over 19472.00 frames. ], tot_loss[loss=0.2984, simple_loss=0.354, pruned_loss=0.1214, over 3815596.44 frames. ], batch size: 49, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:41:44,112 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-01 03:41:52,257 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25753.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:42:30,028 INFO [train.py:903] (3/4) Epoch 4, batch 5300, loss[loss=0.359, simple_loss=0.4001, pruned_loss=0.159, over 19728.00 frames. ], tot_loss[loss=0.2974, simple_loss=0.3533, pruned_loss=0.1207, over 3811048.99 frames. ], batch size: 63, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:42:36,786 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=25790.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 03:42:48,922 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 03:43:23,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.781e+02 7.857e+02 9.756e+02 1.201e+03 3.803e+03, threshold=1.951e+03, percent-clipped=8.0 +2023-04-01 03:43:31,862 INFO [train.py:903] (3/4) Epoch 4, batch 5350, loss[loss=0.3617, simple_loss=0.4106, pruned_loss=0.1565, over 19646.00 frames. ], tot_loss[loss=0.3003, simple_loss=0.3553, pruned_loss=0.1227, over 3816268.65 frames. ], batch size: 60, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:44:04,176 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 03:44:20,736 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0245, 4.9108, 5.8592, 5.7113, 1.8610, 5.4013, 4.7567, 5.2407], + device='cuda:3'), covar=tensor([0.0746, 0.0516, 0.0366, 0.0307, 0.3562, 0.0216, 0.0365, 0.0824], + device='cuda:3'), in_proj_covar=tensor([0.0461, 0.0404, 0.0540, 0.0438, 0.0533, 0.0317, 0.0359, 0.0514], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 03:44:21,925 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25876.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:44:32,191 INFO [train.py:903] (3/4) Epoch 4, batch 5400, loss[loss=0.2578, simple_loss=0.333, pruned_loss=0.09128, over 19691.00 frames. ], tot_loss[loss=0.3016, simple_loss=0.3562, pruned_loss=0.1234, over 3817803.99 frames. ], batch size: 59, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:44:56,679 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25905.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:45:26,555 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.180e+02 6.939e+02 8.636e+02 1.056e+03 2.577e+03, threshold=1.727e+03, percent-clipped=2.0 +2023-04-01 03:45:33,338 INFO [train.py:903] (3/4) Epoch 4, batch 5450, loss[loss=0.3451, simple_loss=0.3936, pruned_loss=0.1483, over 19682.00 frames. ], tot_loss[loss=0.3001, simple_loss=0.355, pruned_loss=0.1226, over 3815171.73 frames. ], batch size: 59, lr: 1.90e-02, grad_scale: 8.0 +2023-04-01 03:46:34,635 INFO [train.py:903] (3/4) Epoch 4, batch 5500, loss[loss=0.2991, simple_loss=0.3618, pruned_loss=0.1182, over 19475.00 frames. ], tot_loss[loss=0.2995, simple_loss=0.3548, pruned_loss=0.1221, over 3815880.12 frames. ], batch size: 64, lr: 1.89e-02, grad_scale: 4.0 +2023-04-01 03:46:58,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 03:47:31,760 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.407e+02 7.568e+02 9.013e+02 1.115e+03 1.816e+03, threshold=1.803e+03, percent-clipped=1.0 +2023-04-01 03:47:37,485 INFO [train.py:903] (3/4) Epoch 4, batch 5550, loss[loss=0.287, simple_loss=0.3384, pruned_loss=0.1178, over 19479.00 frames. ], tot_loss[loss=0.2982, simple_loss=0.3538, pruned_loss=0.1213, over 3815811.37 frames. ], batch size: 49, lr: 1.89e-02, grad_scale: 4.0 +2023-04-01 03:47:45,468 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 03:48:33,397 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 03:48:38,931 INFO [train.py:903] (3/4) Epoch 4, batch 5600, loss[loss=0.3724, simple_loss=0.3979, pruned_loss=0.1735, over 13093.00 frames. ], tot_loss[loss=0.297, simple_loss=0.3528, pruned_loss=0.1206, over 3812410.19 frames. ], batch size: 135, lr: 1.89e-02, grad_scale: 8.0 +2023-04-01 03:49:22,555 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26120.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:49:34,176 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.637e+02 7.669e+02 9.359e+02 1.114e+03 3.409e+03, threshold=1.872e+03, percent-clipped=3.0 +2023-04-01 03:49:40,070 INFO [train.py:903] (3/4) Epoch 4, batch 5650, loss[loss=0.2432, simple_loss=0.3055, pruned_loss=0.09043, over 19499.00 frames. ], tot_loss[loss=0.2963, simple_loss=0.352, pruned_loss=0.1203, over 3823185.63 frames. ], batch size: 49, lr: 1.89e-02, grad_scale: 8.0 +2023-04-01 03:50:12,585 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26161.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:50:25,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 03:50:40,052 INFO [train.py:903] (3/4) Epoch 4, batch 5700, loss[loss=0.3051, simple_loss=0.3586, pruned_loss=0.1257, over 19671.00 frames. ], tot_loss[loss=0.2961, simple_loss=0.352, pruned_loss=0.1201, over 3824993.32 frames. ], batch size: 55, lr: 1.89e-02, grad_scale: 8.0 +2023-04-01 03:50:42,138 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26186.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:50:42,238 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26186.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:51:23,043 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=26220.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:51:35,334 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.235e+02 7.760e+02 9.507e+02 1.157e+03 2.773e+03, threshold=1.901e+03, percent-clipped=5.0 +2023-04-01 03:51:39,943 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 03:51:40,952 INFO [train.py:903] (3/4) Epoch 4, batch 5750, loss[loss=0.2954, simple_loss=0.3692, pruned_loss=0.1108, over 19089.00 frames. ], tot_loss[loss=0.297, simple_loss=0.3526, pruned_loss=0.1207, over 3802027.16 frames. ], batch size: 69, lr: 1.89e-02, grad_scale: 8.0 +2023-04-01 03:51:48,676 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 03:51:52,962 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 03:52:21,192 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26268.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:52:40,293 INFO [train.py:903] (3/4) Epoch 4, batch 5800, loss[loss=0.3028, simple_loss=0.3646, pruned_loss=0.1205, over 19680.00 frames. ], tot_loss[loss=0.2983, simple_loss=0.3535, pruned_loss=0.1216, over 3792716.26 frames. ], batch size: 59, lr: 1.88e-02, grad_scale: 8.0 +2023-04-01 03:53:36,643 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.499e+02 6.787e+02 8.542e+02 1.114e+03 2.576e+03, threshold=1.708e+03, percent-clipped=4.0 +2023-04-01 03:53:41,252 INFO [train.py:903] (3/4) Epoch 4, batch 5850, loss[loss=0.2672, simple_loss=0.3416, pruned_loss=0.09639, over 19734.00 frames. ], tot_loss[loss=0.2989, simple_loss=0.3541, pruned_loss=0.1219, over 3796234.05 frames. ], batch size: 63, lr: 1.88e-02, grad_scale: 4.0 +2023-04-01 03:53:41,613 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26335.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 03:54:40,815 INFO [train.py:903] (3/4) Epoch 4, batch 5900, loss[loss=0.3462, simple_loss=0.3847, pruned_loss=0.1538, over 13361.00 frames. ], tot_loss[loss=0.2978, simple_loss=0.353, pruned_loss=0.1213, over 3802696.44 frames. ], batch size: 136, lr: 1.88e-02, grad_scale: 4.0 +2023-04-01 03:54:41,863 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 03:55:03,848 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 03:55:37,699 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.672e+02 6.673e+02 8.574e+02 1.115e+03 3.080e+03, threshold=1.715e+03, percent-clipped=4.0 +2023-04-01 03:55:39,780 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26432.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:55:43,001 INFO [train.py:903] (3/4) Epoch 4, batch 5950, loss[loss=0.3308, simple_loss=0.3791, pruned_loss=0.1412, over 19355.00 frames. ], tot_loss[loss=0.2991, simple_loss=0.3542, pruned_loss=0.122, over 3793599.04 frames. ], batch size: 66, lr: 1.88e-02, grad_scale: 4.0 +2023-04-01 03:55:44,118 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2485, 3.6428, 3.9735, 4.0988, 1.4732, 3.7310, 3.3268, 3.1740], + device='cuda:3'), covar=tensor([0.1833, 0.1313, 0.0987, 0.0865, 0.5306, 0.0736, 0.0850, 0.2058], + device='cuda:3'), in_proj_covar=tensor([0.0470, 0.0419, 0.0547, 0.0442, 0.0539, 0.0316, 0.0360, 0.0522], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 03:56:15,420 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4313, 2.3929, 1.8728, 2.0287, 1.7854, 2.1484, 1.1569, 2.0423], + device='cuda:3'), covar=tensor([0.0187, 0.0255, 0.0205, 0.0243, 0.0358, 0.0268, 0.0408, 0.0313], + device='cuda:3'), in_proj_covar=tensor([0.0271, 0.0272, 0.0264, 0.0294, 0.0351, 0.0277, 0.0268, 0.0282], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 03:56:17,511 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=26464.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:56:43,852 INFO [train.py:903] (3/4) Epoch 4, batch 6000, loss[loss=0.2413, simple_loss=0.3003, pruned_loss=0.09113, over 19107.00 frames. ], tot_loss[loss=0.2986, simple_loss=0.3538, pruned_loss=0.1216, over 3794097.01 frames. ], batch size: 42, lr: 1.88e-02, grad_scale: 8.0 +2023-04-01 03:56:43,852 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 03:56:57,348 INFO [train.py:937] (3/4) Epoch 4, validation: loss=0.2103, simple_loss=0.3081, pruned_loss=0.05622, over 944034.00 frames. +2023-04-01 03:56:57,349 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 03:57:52,005 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=26530.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:57:52,999 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.403e+02 7.085e+02 8.722e+02 1.145e+03 2.334e+03, threshold=1.744e+03, percent-clipped=4.0 +2023-04-01 03:57:57,465 INFO [train.py:903] (3/4) Epoch 4, batch 6050, loss[loss=0.3192, simple_loss=0.3702, pruned_loss=0.1341, over 19555.00 frames. ], tot_loss[loss=0.2984, simple_loss=0.3536, pruned_loss=0.1216, over 3800610.13 frames. ], batch size: 61, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 03:58:18,096 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1905, 2.0875, 1.8477, 3.1557, 2.0939, 3.5849, 3.0345, 1.9234], + device='cuda:3'), covar=tensor([0.1895, 0.1439, 0.0770, 0.1008, 0.1779, 0.0405, 0.1295, 0.1284], + device='cuda:3'), in_proj_covar=tensor([0.0588, 0.0571, 0.0531, 0.0730, 0.0630, 0.0489, 0.0637, 0.0542], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 03:58:49,976 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26579.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:58:57,312 INFO [train.py:903] (3/4) Epoch 4, batch 6100, loss[loss=0.2842, simple_loss=0.35, pruned_loss=0.1092, over 19759.00 frames. ], tot_loss[loss=0.297, simple_loss=0.3527, pruned_loss=0.1207, over 3819683.98 frames. ], batch size: 54, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 03:59:04,787 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26591.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 03:59:29,876 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=26612.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 03:59:34,622 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26616.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 03:59:51,598 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.600e+02 6.768e+02 7.963e+02 1.032e+03 2.370e+03, threshold=1.593e+03, percent-clipped=5.0 +2023-04-01 03:59:56,257 INFO [train.py:903] (3/4) Epoch 4, batch 6150, loss[loss=0.3342, simple_loss=0.3894, pruned_loss=0.1395, over 19704.00 frames. ], tot_loss[loss=0.2981, simple_loss=0.3535, pruned_loss=0.1214, over 3811172.03 frames. ], batch size: 59, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 04:00:08,599 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26645.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:00:23,789 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 04:00:54,794 INFO [train.py:903] (3/4) Epoch 4, batch 6200, loss[loss=0.3103, simple_loss=0.3744, pruned_loss=0.123, over 19663.00 frames. ], tot_loss[loss=0.2996, simple_loss=0.3541, pruned_loss=0.1225, over 3806949.32 frames. ], batch size: 58, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 04:01:45,218 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26727.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:01:51,166 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.725e+02 7.340e+02 8.907e+02 1.200e+03 2.630e+03, threshold=1.781e+03, percent-clipped=5.0 +2023-04-01 04:01:55,636 INFO [train.py:903] (3/4) Epoch 4, batch 6250, loss[loss=0.3175, simple_loss=0.3742, pruned_loss=0.1304, over 18713.00 frames. ], tot_loss[loss=0.3, simple_loss=0.3547, pruned_loss=0.1226, over 3815130.42 frames. ], batch size: 74, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 04:02:10,574 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8162, 1.8881, 1.7552, 2.6565, 1.8987, 2.7218, 2.3952, 1.6597], + device='cuda:3'), covar=tensor([0.1661, 0.1300, 0.0816, 0.0778, 0.1430, 0.0460, 0.1448, 0.1394], + device='cuda:3'), in_proj_covar=tensor([0.0581, 0.0565, 0.0529, 0.0725, 0.0625, 0.0485, 0.0633, 0.0538], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:02:24,773 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 04:02:42,400 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8136, 4.0574, 4.4354, 4.4490, 1.7299, 4.0398, 3.7412, 3.9818], + device='cuda:3'), covar=tensor([0.0747, 0.0678, 0.0464, 0.0327, 0.3358, 0.0341, 0.0393, 0.0837], + device='cuda:3'), in_proj_covar=tensor([0.0463, 0.0410, 0.0554, 0.0441, 0.0536, 0.0316, 0.0366, 0.0519], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 04:02:44,567 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=26776.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:02:49,364 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 04:02:55,313 INFO [train.py:903] (3/4) Epoch 4, batch 6300, loss[loss=0.3307, simple_loss=0.3675, pruned_loss=0.147, over 19853.00 frames. ], tot_loss[loss=0.3002, simple_loss=0.3551, pruned_loss=0.1226, over 3808652.19 frames. ], batch size: 52, lr: 1.87e-02, grad_scale: 8.0 +2023-04-01 04:03:49,870 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.898e+02 7.473e+02 9.358e+02 1.215e+03 3.413e+03, threshold=1.872e+03, percent-clipped=4.0 +2023-04-01 04:03:54,565 INFO [train.py:903] (3/4) Epoch 4, batch 6350, loss[loss=0.293, simple_loss=0.3633, pruned_loss=0.1113, over 19663.00 frames. ], tot_loss[loss=0.2993, simple_loss=0.3544, pruned_loss=0.1222, over 3807599.31 frames. ], batch size: 53, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:03:55,024 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26835.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:04:26,067 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26860.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:04:37,816 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-04-01 04:04:45,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0791, 3.1761, 3.4998, 3.4991, 1.9255, 3.1407, 3.0459, 3.1768], + device='cuda:3'), covar=tensor([0.0904, 0.1610, 0.0570, 0.0452, 0.2977, 0.0631, 0.0452, 0.0936], + device='cuda:3'), in_proj_covar=tensor([0.0468, 0.0416, 0.0559, 0.0442, 0.0540, 0.0319, 0.0367, 0.0524], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 04:04:55,432 INFO [train.py:903] (3/4) Epoch 4, batch 6400, loss[loss=0.2838, simple_loss=0.3508, pruned_loss=0.1084, over 19627.00 frames. ], tot_loss[loss=0.298, simple_loss=0.3535, pruned_loss=0.1213, over 3808994.97 frames. ], batch size: 57, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:05:05,497 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26891.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:05:16,543 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26901.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:05:41,172 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26922.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:05:45,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26926.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:05:52,842 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.488e+02 6.991e+02 8.678e+02 1.023e+03 2.915e+03, threshold=1.736e+03, percent-clipped=2.0 +2023-04-01 04:05:58,084 INFO [train.py:903] (3/4) Epoch 4, batch 6450, loss[loss=0.3003, simple_loss=0.36, pruned_loss=0.1203, over 19667.00 frames. ], tot_loss[loss=0.2991, simple_loss=0.3548, pruned_loss=0.1217, over 3809762.85 frames. ], batch size: 55, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:06:37,698 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26968.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:06:40,563 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 04:06:56,681 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26983.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:06:58,607 INFO [train.py:903] (3/4) Epoch 4, batch 6500, loss[loss=0.2665, simple_loss=0.3282, pruned_loss=0.1024, over 19469.00 frames. ], tot_loss[loss=0.3, simple_loss=0.3556, pruned_loss=0.1222, over 3813783.85 frames. ], batch size: 49, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:07:03,008 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 04:07:25,287 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27008.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:07:53,581 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.310e+02 7.867e+02 9.982e+02 1.245e+03 2.621e+03, threshold=1.996e+03, percent-clipped=6.0 +2023-04-01 04:07:55,893 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5868, 1.3577, 1.3695, 1.6316, 1.6101, 1.4121, 1.3365, 1.5578], + device='cuda:3'), covar=tensor([0.0629, 0.1071, 0.0995, 0.0624, 0.0758, 0.0417, 0.0841, 0.0532], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0369, 0.0279, 0.0240, 0.0305, 0.0254, 0.0272, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 04:07:57,612 INFO [train.py:903] (3/4) Epoch 4, batch 6550, loss[loss=0.2972, simple_loss=0.3501, pruned_loss=0.1221, over 19669.00 frames. ], tot_loss[loss=0.3025, simple_loss=0.3572, pruned_loss=0.1239, over 3817696.90 frames. ], batch size: 53, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:08:57,153 INFO [train.py:903] (3/4) Epoch 4, batch 6600, loss[loss=0.2618, simple_loss=0.336, pruned_loss=0.09376, over 19674.00 frames. ], tot_loss[loss=0.301, simple_loss=0.3562, pruned_loss=0.1229, over 3825778.74 frames. ], batch size: 55, lr: 1.86e-02, grad_scale: 8.0 +2023-04-01 04:09:53,360 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.938e+02 7.635e+02 9.605e+02 1.185e+03 2.942e+03, threshold=1.921e+03, percent-clipped=6.0 +2023-04-01 04:09:58,813 INFO [train.py:903] (3/4) Epoch 4, batch 6650, loss[loss=0.2978, simple_loss=0.3607, pruned_loss=0.1175, over 19663.00 frames. ], tot_loss[loss=0.2999, simple_loss=0.3553, pruned_loss=0.1223, over 3828827.81 frames. ], batch size: 55, lr: 1.85e-02, grad_scale: 8.0 +2023-04-01 04:10:03,557 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9787, 2.0004, 1.8424, 2.8186, 1.9768, 2.6627, 2.5277, 1.8158], + device='cuda:3'), covar=tensor([0.1619, 0.1285, 0.0711, 0.0751, 0.1425, 0.0508, 0.1254, 0.1178], + device='cuda:3'), in_proj_covar=tensor([0.0589, 0.0575, 0.0531, 0.0731, 0.0628, 0.0486, 0.0644, 0.0549], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:10:13,164 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27147.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:10:35,302 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=27166.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:10:39,786 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.94 vs. limit=5.0 +2023-04-01 04:10:43,566 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27172.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:10:58,823 INFO [train.py:903] (3/4) Epoch 4, batch 6700, loss[loss=0.3074, simple_loss=0.362, pruned_loss=0.1264, over 19663.00 frames. ], tot_loss[loss=0.2999, simple_loss=0.355, pruned_loss=0.1224, over 3834563.63 frames. ], batch size: 58, lr: 1.85e-02, grad_scale: 8.0 +2023-04-01 04:11:52,642 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.097e+02 7.210e+02 9.176e+02 1.266e+03 4.477e+03, threshold=1.835e+03, percent-clipped=7.0 +2023-04-01 04:11:57,028 INFO [train.py:903] (3/4) Epoch 4, batch 6750, loss[loss=0.3078, simple_loss=0.3684, pruned_loss=0.1236, over 19660.00 frames. ], tot_loss[loss=0.2993, simple_loss=0.3546, pruned_loss=0.1219, over 3829850.46 frames. ], batch size: 55, lr: 1.85e-02, grad_scale: 8.0 +2023-04-01 04:12:31,970 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=27266.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:12:50,003 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7687, 3.1368, 3.2574, 3.2081, 1.0667, 2.9702, 2.7095, 2.9091], + device='cuda:3'), covar=tensor([0.1049, 0.0644, 0.0665, 0.0582, 0.3568, 0.0451, 0.0566, 0.1271], + device='cuda:3'), in_proj_covar=tensor([0.0472, 0.0416, 0.0562, 0.0445, 0.0543, 0.0321, 0.0367, 0.0525], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 04:12:52,218 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6976, 1.3088, 1.1684, 1.6083, 1.4312, 1.3623, 1.2582, 1.5430], + device='cuda:3'), covar=tensor([0.0759, 0.1272, 0.1288, 0.0809, 0.0981, 0.0591, 0.0997, 0.0648], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0365, 0.0277, 0.0240, 0.0301, 0.0253, 0.0267, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 04:12:52,952 INFO [train.py:903] (3/4) Epoch 4, batch 6800, loss[loss=0.3252, simple_loss=0.3744, pruned_loss=0.138, over 19687.00 frames. ], tot_loss[loss=0.2995, simple_loss=0.3543, pruned_loss=0.1223, over 3826985.02 frames. ], batch size: 59, lr: 1.85e-02, grad_scale: 8.0 +2023-04-01 04:13:36,970 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 04:13:37,407 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 04:13:40,505 INFO [train.py:903] (3/4) Epoch 5, batch 0, loss[loss=0.3108, simple_loss=0.3551, pruned_loss=0.1333, over 19505.00 frames. ], tot_loss[loss=0.3108, simple_loss=0.3551, pruned_loss=0.1333, over 19505.00 frames. ], batch size: 49, lr: 1.72e-02, grad_scale: 8.0 +2023-04-01 04:13:40,505 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 04:13:52,272 INFO [train.py:937] (3/4) Epoch 5, validation: loss=0.2121, simple_loss=0.3102, pruned_loss=0.05704, over 944034.00 frames. +2023-04-01 04:13:52,273 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 04:13:52,410 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=27312.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:14:04,621 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 04:14:16,053 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 5.218e+02 7.861e+02 9.859e+02 1.236e+03 2.711e+03, threshold=1.972e+03, percent-clipped=3.0 +2023-04-01 04:14:52,468 INFO [train.py:903] (3/4) Epoch 5, batch 50, loss[loss=0.2716, simple_loss=0.3346, pruned_loss=0.1043, over 19539.00 frames. ], tot_loss[loss=0.3023, simple_loss=0.3574, pruned_loss=0.1236, over 864619.01 frames. ], batch size: 54, lr: 1.72e-02, grad_scale: 8.0 +2023-04-01 04:15:15,069 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27381.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:15:26,079 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 04:15:53,859 INFO [train.py:903] (3/4) Epoch 5, batch 100, loss[loss=0.2454, simple_loss=0.3128, pruned_loss=0.08898, over 19807.00 frames. ], tot_loss[loss=0.2969, simple_loss=0.3547, pruned_loss=0.1196, over 1524617.15 frames. ], batch size: 49, lr: 1.72e-02, grad_scale: 8.0 +2023-04-01 04:16:05,289 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 04:16:11,155 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27427.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:16:15,198 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.977e+02 6.953e+02 8.679e+02 1.081e+03 2.199e+03, threshold=1.736e+03, percent-clipped=1.0 +2023-04-01 04:16:53,739 INFO [train.py:903] (3/4) Epoch 5, batch 150, loss[loss=0.3243, simple_loss=0.3759, pruned_loss=0.1363, over 19316.00 frames. ], tot_loss[loss=0.2968, simple_loss=0.354, pruned_loss=0.1198, over 2040066.00 frames. ], batch size: 66, lr: 1.72e-02, grad_scale: 8.0 +2023-04-01 04:17:52,401 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=27510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:17:54,451 INFO [train.py:903] (3/4) Epoch 5, batch 200, loss[loss=0.3072, simple_loss=0.3604, pruned_loss=0.127, over 19728.00 frames. ], tot_loss[loss=0.2933, simple_loss=0.3509, pruned_loss=0.1179, over 2425014.69 frames. ], batch size: 51, lr: 1.72e-02, grad_scale: 8.0 +2023-04-01 04:17:54,461 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 04:18:19,379 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.602e+02 6.870e+02 8.382e+02 1.064e+03 2.606e+03, threshold=1.676e+03, percent-clipped=3.0 +2023-04-01 04:18:56,504 INFO [train.py:903] (3/4) Epoch 5, batch 250, loss[loss=0.3033, simple_loss=0.3626, pruned_loss=0.122, over 19532.00 frames. ], tot_loss[loss=0.2901, simple_loss=0.3483, pruned_loss=0.1159, over 2744980.59 frames. ], batch size: 56, lr: 1.71e-02, grad_scale: 4.0 +2023-04-01 04:19:58,942 INFO [train.py:903] (3/4) Epoch 5, batch 300, loss[loss=0.2813, simple_loss=0.3399, pruned_loss=0.1113, over 19664.00 frames. ], tot_loss[loss=0.2913, simple_loss=0.3492, pruned_loss=0.1167, over 2989220.27 frames. ], batch size: 55, lr: 1.71e-02, grad_scale: 4.0 +2023-04-01 04:20:15,422 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27625.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:20:22,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.497e+02 6.581e+02 8.607e+02 1.103e+03 1.922e+03, threshold=1.721e+03, percent-clipped=6.0 +2023-04-01 04:20:28,921 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27637.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:21:01,017 INFO [train.py:903] (3/4) Epoch 5, batch 350, loss[loss=0.354, simple_loss=0.4054, pruned_loss=0.1513, over 18796.00 frames. ], tot_loss[loss=0.2927, simple_loss=0.3497, pruned_loss=0.1179, over 3175625.83 frames. ], batch size: 74, lr: 1.71e-02, grad_scale: 4.0 +2023-04-01 04:21:01,426 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27662.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:21:07,171 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 04:21:26,428 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27683.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:21:58,795 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27708.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:22:02,939 INFO [train.py:903] (3/4) Epoch 5, batch 400, loss[loss=0.2491, simple_loss=0.3293, pruned_loss=0.08441, over 19566.00 frames. ], tot_loss[loss=0.2915, simple_loss=0.3488, pruned_loss=0.1171, over 3327882.47 frames. ], batch size: 52, lr: 1.71e-02, grad_scale: 8.0 +2023-04-01 04:22:26,130 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1282, 2.1926, 1.9089, 3.2084, 2.1578, 3.5238, 2.9403, 2.0781], + device='cuda:3'), covar=tensor([0.2028, 0.1577, 0.0811, 0.1024, 0.1951, 0.0474, 0.1430, 0.1247], + device='cuda:3'), in_proj_covar=tensor([0.0595, 0.0583, 0.0534, 0.0742, 0.0638, 0.0499, 0.0647, 0.0548], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:22:27,973 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.948e+02 7.305e+02 9.017e+02 1.065e+03 1.815e+03, threshold=1.803e+03, percent-clipped=3.0 +2023-04-01 04:22:28,320 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=27732.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:22:43,460 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2612, 1.2807, 1.6207, 0.9032, 2.4398, 2.7763, 2.6693, 2.9179], + device='cuda:3'), covar=tensor([0.1308, 0.2674, 0.2609, 0.2121, 0.0469, 0.0220, 0.0251, 0.0163], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0281, 0.0319, 0.0259, 0.0197, 0.0115, 0.0206, 0.0135], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:23:04,214 INFO [train.py:903] (3/4) Epoch 5, batch 450, loss[loss=0.3029, simple_loss=0.3634, pruned_loss=0.1212, over 19337.00 frames. ], tot_loss[loss=0.2905, simple_loss=0.3484, pruned_loss=0.1163, over 3444199.75 frames. ], batch size: 70, lr: 1.71e-02, grad_scale: 8.0 +2023-04-01 04:23:45,999 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 04:23:47,128 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 04:24:07,106 INFO [train.py:903] (3/4) Epoch 5, batch 500, loss[loss=0.2807, simple_loss=0.337, pruned_loss=0.1123, over 18649.00 frames. ], tot_loss[loss=0.2901, simple_loss=0.3484, pruned_loss=0.1159, over 3530726.48 frames. ], batch size: 41, lr: 1.71e-02, grad_scale: 8.0 +2023-04-01 04:24:31,790 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.519e+02 6.150e+02 8.318e+02 1.057e+03 1.987e+03, threshold=1.664e+03, percent-clipped=1.0 +2023-04-01 04:25:11,324 INFO [train.py:903] (3/4) Epoch 5, batch 550, loss[loss=0.3549, simple_loss=0.4085, pruned_loss=0.1507, over 19686.00 frames. ], tot_loss[loss=0.2905, simple_loss=0.349, pruned_loss=0.116, over 3602974.23 frames. ], batch size: 59, lr: 1.71e-02, grad_scale: 8.0 +2023-04-01 04:25:35,647 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27881.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:25:56,960 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7179, 4.1863, 2.5526, 3.6392, 0.8576, 3.8130, 3.8568, 3.9345], + device='cuda:3'), covar=tensor([0.0559, 0.1055, 0.1852, 0.0775, 0.4158, 0.0793, 0.0726, 0.0796], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0305, 0.0354, 0.0276, 0.0345, 0.0294, 0.0270, 0.0301], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:26:07,957 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27906.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:26:13,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1338, 1.0967, 1.3419, 0.5387, 2.3671, 2.3581, 2.1365, 2.4693], + device='cuda:3'), covar=tensor([0.1311, 0.3001, 0.2924, 0.2162, 0.0365, 0.0180, 0.0355, 0.0201], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0280, 0.0316, 0.0255, 0.0195, 0.0114, 0.0204, 0.0134], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:26:14,567 INFO [train.py:903] (3/4) Epoch 5, batch 600, loss[loss=0.3294, simple_loss=0.3808, pruned_loss=0.139, over 19304.00 frames. ], tot_loss[loss=0.2901, simple_loss=0.3487, pruned_loss=0.1157, over 3657565.73 frames. ], batch size: 66, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:26:38,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.822e+02 6.595e+02 8.388e+02 1.023e+03 2.578e+03, threshold=1.678e+03, percent-clipped=3.0 +2023-04-01 04:27:02,854 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 04:27:12,903 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-01 04:27:17,953 INFO [train.py:903] (3/4) Epoch 5, batch 650, loss[loss=0.2655, simple_loss=0.3266, pruned_loss=0.1021, over 19730.00 frames. ], tot_loss[loss=0.2909, simple_loss=0.3487, pruned_loss=0.1165, over 3698594.85 frames. ], batch size: 51, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:27:22,971 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4040, 2.4349, 1.5508, 1.4405, 2.2157, 1.0573, 1.2297, 1.5720], + device='cuda:3'), covar=tensor([0.0755, 0.0412, 0.0823, 0.0537, 0.0343, 0.1032, 0.0625, 0.0439], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0254, 0.0305, 0.0234, 0.0215, 0.0309, 0.0271, 0.0244], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 04:27:59,453 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=5.01 vs. limit=5.0 +2023-04-01 04:28:20,084 INFO [train.py:903] (3/4) Epoch 5, batch 700, loss[loss=0.3683, simple_loss=0.3958, pruned_loss=0.1704, over 13463.00 frames. ], tot_loss[loss=0.2909, simple_loss=0.3488, pruned_loss=0.1165, over 3714978.09 frames. ], batch size: 136, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:28:47,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.733e+02 7.486e+02 9.333e+02 1.140e+03 2.488e+03, threshold=1.867e+03, percent-clipped=5.0 +2023-04-01 04:29:08,508 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4171, 2.2307, 1.8350, 1.7099, 1.4843, 1.7718, 0.3940, 1.4231], + device='cuda:3'), covar=tensor([0.0271, 0.0230, 0.0182, 0.0302, 0.0505, 0.0278, 0.0483, 0.0387], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0278, 0.0273, 0.0301, 0.0360, 0.0283, 0.0274, 0.0282], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:29:25,957 INFO [train.py:903] (3/4) Epoch 5, batch 750, loss[loss=0.3251, simple_loss=0.3627, pruned_loss=0.1437, over 19587.00 frames. ], tot_loss[loss=0.2921, simple_loss=0.3497, pruned_loss=0.1172, over 3743397.49 frames. ], batch size: 52, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:29:43,608 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28076.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:30:27,891 INFO [train.py:903] (3/4) Epoch 5, batch 800, loss[loss=0.3127, simple_loss=0.3683, pruned_loss=0.1286, over 19599.00 frames. ], tot_loss[loss=0.2923, simple_loss=0.3495, pruned_loss=0.1175, over 3747519.95 frames. ], batch size: 61, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:30:48,164 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 04:30:52,981 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.190e+02 6.427e+02 8.567e+02 1.032e+03 2.729e+03, threshold=1.713e+03, percent-clipped=3.0 +2023-04-01 04:31:32,204 INFO [train.py:903] (3/4) Epoch 5, batch 850, loss[loss=0.3072, simple_loss=0.3578, pruned_loss=0.1283, over 19533.00 frames. ], tot_loss[loss=0.2897, simple_loss=0.3474, pruned_loss=0.1161, over 3752805.58 frames. ], batch size: 54, lr: 1.70e-02, grad_scale: 8.0 +2023-04-01 04:32:03,333 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-04-01 04:32:08,670 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28191.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:32:14,506 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7481, 1.4445, 1.8654, 1.5164, 3.0623, 4.4961, 4.7504, 5.0691], + device='cuda:3'), covar=tensor([0.1215, 0.2533, 0.2383, 0.1664, 0.0384, 0.0137, 0.0124, 0.0056], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0279, 0.0314, 0.0254, 0.0196, 0.0115, 0.0204, 0.0133], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:3') +2023-04-01 04:32:29,187 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 04:32:33,626 INFO [train.py:903] (3/4) Epoch 5, batch 900, loss[loss=0.2157, simple_loss=0.2845, pruned_loss=0.07343, over 19732.00 frames. ], tot_loss[loss=0.2913, simple_loss=0.3486, pruned_loss=0.117, over 3754461.57 frames. ], batch size: 45, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:32:59,333 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.597e+02 6.938e+02 8.196e+02 1.125e+03 2.658e+03, threshold=1.639e+03, percent-clipped=4.0 +2023-04-01 04:33:17,402 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28247.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:33:22,480 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.47 vs. limit=5.0 +2023-04-01 04:33:36,690 INFO [train.py:903] (3/4) Epoch 5, batch 950, loss[loss=0.2512, simple_loss=0.3246, pruned_loss=0.08896, over 19718.00 frames. ], tot_loss[loss=0.289, simple_loss=0.3468, pruned_loss=0.1156, over 3776185.02 frames. ], batch size: 63, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:33:42,368 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 04:33:49,441 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28273.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:34:28,485 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28305.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:34:36,191 INFO [train.py:903] (3/4) Epoch 5, batch 1000, loss[loss=0.2281, simple_loss=0.2953, pruned_loss=0.08043, over 19738.00 frames. ], tot_loss[loss=0.2877, simple_loss=0.3459, pruned_loss=0.1148, over 3785400.45 frames. ], batch size: 45, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:34:59,300 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.517e+02 7.143e+02 8.888e+02 1.138e+03 2.880e+03, threshold=1.778e+03, percent-clipped=9.0 +2023-04-01 04:35:30,270 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 04:35:36,617 INFO [train.py:903] (3/4) Epoch 5, batch 1050, loss[loss=0.3527, simple_loss=0.3887, pruned_loss=0.1583, over 19628.00 frames. ], tot_loss[loss=0.2901, simple_loss=0.3478, pruned_loss=0.1162, over 3782317.63 frames. ], batch size: 57, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:36:09,495 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 04:36:36,348 INFO [train.py:903] (3/4) Epoch 5, batch 1100, loss[loss=0.3007, simple_loss=0.3659, pruned_loss=0.1177, over 19775.00 frames. ], tot_loss[loss=0.2897, simple_loss=0.3475, pruned_loss=0.1159, over 3789679.99 frames. ], batch size: 56, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:37:01,573 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.535e+02 6.917e+02 8.996e+02 1.190e+03 3.192e+03, threshold=1.799e+03, percent-clipped=6.0 +2023-04-01 04:37:19,648 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28447.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:37:37,665 INFO [train.py:903] (3/4) Epoch 5, batch 1150, loss[loss=0.269, simple_loss=0.3224, pruned_loss=0.1078, over 17745.00 frames. ], tot_loss[loss=0.2887, simple_loss=0.3468, pruned_loss=0.1153, over 3807223.33 frames. ], batch size: 39, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:37:50,296 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28472.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:38:11,055 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1976, 1.0833, 1.0769, 1.4466, 1.1863, 1.2642, 1.3590, 1.2087], + device='cuda:3'), covar=tensor([0.0878, 0.1107, 0.1148, 0.0681, 0.0851, 0.0854, 0.0846, 0.0797], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0254, 0.0245, 0.0279, 0.0280, 0.0237, 0.0242, 0.0229], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 04:38:29,212 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28504.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:38:36,139 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:38:38,084 INFO [train.py:903] (3/4) Epoch 5, batch 1200, loss[loss=0.2809, simple_loss=0.3482, pruned_loss=0.1068, over 19606.00 frames. ], tot_loss[loss=0.2887, simple_loss=0.3471, pruned_loss=0.1152, over 3811512.11 frames. ], batch size: 57, lr: 1.69e-02, grad_scale: 8.0 +2023-04-01 04:39:01,764 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.620e+02 6.666e+02 7.999e+02 1.012e+03 1.920e+03, threshold=1.600e+03, percent-clipped=0.0 +2023-04-01 04:39:12,212 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 04:39:37,128 INFO [train.py:903] (3/4) Epoch 5, batch 1250, loss[loss=0.3317, simple_loss=0.3832, pruned_loss=0.1401, over 19590.00 frames. ], tot_loss[loss=0.2899, simple_loss=0.3477, pruned_loss=0.116, over 3823932.43 frames. ], batch size: 61, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:40:13,035 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28591.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:40:37,857 INFO [train.py:903] (3/4) Epoch 5, batch 1300, loss[loss=0.2364, simple_loss=0.297, pruned_loss=0.08788, over 19775.00 frames. ], tot_loss[loss=0.289, simple_loss=0.3469, pruned_loss=0.1156, over 3816351.16 frames. ], batch size: 47, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:40:43,796 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28617.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:41:03,193 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.211e+02 7.072e+02 8.596e+02 1.219e+03 1.879e+03, threshold=1.719e+03, percent-clipped=8.0 +2023-04-01 04:41:11,362 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4885, 3.2055, 2.0340, 2.5612, 2.2992, 2.5013, 0.8495, 2.5256], + device='cuda:3'), covar=tensor([0.0270, 0.0218, 0.0289, 0.0323, 0.0384, 0.0338, 0.0546, 0.0371], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0277, 0.0275, 0.0296, 0.0360, 0.0277, 0.0273, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:41:23,371 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28649.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:41:39,846 INFO [train.py:903] (3/4) Epoch 5, batch 1350, loss[loss=0.3165, simple_loss=0.3585, pruned_loss=0.1372, over 19593.00 frames. ], tot_loss[loss=0.2879, simple_loss=0.3463, pruned_loss=0.1147, over 3820322.66 frames. ], batch size: 52, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:42:34,081 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28706.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:42:40,552 INFO [train.py:903] (3/4) Epoch 5, batch 1400, loss[loss=0.3196, simple_loss=0.3752, pruned_loss=0.132, over 19616.00 frames. ], tot_loss[loss=0.2872, simple_loss=0.3452, pruned_loss=0.1145, over 3822078.22 frames. ], batch size: 57, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:43:01,034 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0486, 4.9953, 5.9904, 5.8525, 1.8635, 5.4745, 4.9142, 5.3725], + device='cuda:3'), covar=tensor([0.0794, 0.0547, 0.0378, 0.0306, 0.3759, 0.0257, 0.0398, 0.0808], + device='cuda:3'), in_proj_covar=tensor([0.0495, 0.0445, 0.0587, 0.0466, 0.0564, 0.0337, 0.0379, 0.0543], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 04:43:04,142 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.747e+02 6.792e+02 9.169e+02 1.129e+03 1.829e+03, threshold=1.834e+03, percent-clipped=1.0 +2023-04-01 04:43:04,528 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28732.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:43:40,832 INFO [train.py:903] (3/4) Epoch 5, batch 1450, loss[loss=0.269, simple_loss=0.3414, pruned_loss=0.09829, over 19688.00 frames. ], tot_loss[loss=0.2873, simple_loss=0.3458, pruned_loss=0.1144, over 3826072.08 frames. ], batch size: 59, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:43:43,195 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 04:43:43,506 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28764.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:44:41,295 INFO [train.py:903] (3/4) Epoch 5, batch 1500, loss[loss=0.289, simple_loss=0.3445, pruned_loss=0.1168, over 19600.00 frames. ], tot_loss[loss=0.2876, simple_loss=0.3464, pruned_loss=0.1144, over 3813306.00 frames. ], batch size: 50, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:45:06,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.730e+02 6.912e+02 8.562e+02 1.022e+03 2.509e+03, threshold=1.712e+03, percent-clipped=1.0 +2023-04-01 04:45:24,303 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28848.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:45:31,741 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=28854.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:45:32,290 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-01 04:45:42,354 INFO [train.py:903] (3/4) Epoch 5, batch 1550, loss[loss=0.2861, simple_loss=0.3414, pruned_loss=0.1154, over 19467.00 frames. ], tot_loss[loss=0.2883, simple_loss=0.3464, pruned_loss=0.1151, over 3794745.23 frames. ], batch size: 49, lr: 1.68e-02, grad_scale: 8.0 +2023-04-01 04:46:02,294 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28879.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:46:41,467 INFO [train.py:903] (3/4) Epoch 5, batch 1600, loss[loss=0.3774, simple_loss=0.4008, pruned_loss=0.177, over 13647.00 frames. ], tot_loss[loss=0.2888, simple_loss=0.347, pruned_loss=0.1153, over 3786308.33 frames. ], batch size: 136, lr: 1.67e-02, grad_scale: 8.0 +2023-04-01 04:47:01,752 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9744, 2.0257, 1.9170, 2.8618, 1.8063, 2.6600, 2.6035, 1.8919], + device='cuda:3'), covar=tensor([0.1742, 0.1352, 0.0726, 0.0833, 0.1727, 0.0564, 0.1274, 0.1221], + device='cuda:3'), in_proj_covar=tensor([0.0605, 0.0591, 0.0544, 0.0751, 0.0651, 0.0507, 0.0656, 0.0561], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:47:04,538 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.940e+02 6.984e+02 8.420e+02 1.102e+03 2.946e+03, threshold=1.684e+03, percent-clipped=4.0 +2023-04-01 04:47:04,566 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 04:47:41,582 INFO [train.py:903] (3/4) Epoch 5, batch 1650, loss[loss=0.3141, simple_loss=0.3764, pruned_loss=0.1259, over 18729.00 frames. ], tot_loss[loss=0.2881, simple_loss=0.3464, pruned_loss=0.1149, over 3792307.03 frames. ], batch size: 74, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:47:42,054 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28962.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:47:43,184 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28963.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:47:49,877 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:48:12,492 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28987.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:48:14,445 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28988.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 04:48:42,352 INFO [train.py:903] (3/4) Epoch 5, batch 1700, loss[loss=0.3026, simple_loss=0.3602, pruned_loss=0.1225, over 19525.00 frames. ], tot_loss[loss=0.288, simple_loss=0.3462, pruned_loss=0.1149, over 3790845.95 frames. ], batch size: 64, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:48:43,875 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29013.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 04:48:52,758 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29020.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:48:52,890 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29020.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:49:08,764 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.635e+02 6.699e+02 8.227e+02 1.083e+03 2.721e+03, threshold=1.645e+03, percent-clipped=4.0 +2023-04-01 04:49:16,993 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3283, 2.1823, 2.0678, 3.3234, 2.1587, 3.5590, 3.1710, 2.0237], + device='cuda:3'), covar=tensor([0.1836, 0.1487, 0.0750, 0.0917, 0.1860, 0.0461, 0.1266, 0.1303], + device='cuda:3'), in_proj_covar=tensor([0.0599, 0.0589, 0.0536, 0.0743, 0.0645, 0.0502, 0.0657, 0.0557], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:49:21,046 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 04:49:22,488 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29045.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:49:36,773 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.34 vs. limit=5.0 +2023-04-01 04:49:40,703 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8757, 1.8762, 1.9000, 2.5226, 1.6013, 2.3577, 2.3877, 1.7705], + device='cuda:3'), covar=tensor([0.1607, 0.1272, 0.0717, 0.0685, 0.1438, 0.0535, 0.1355, 0.1274], + device='cuda:3'), in_proj_covar=tensor([0.0594, 0.0584, 0.0531, 0.0737, 0.0640, 0.0498, 0.0651, 0.0554], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:49:42,475 INFO [train.py:903] (3/4) Epoch 5, batch 1750, loss[loss=0.2999, simple_loss=0.3645, pruned_loss=0.1177, over 19650.00 frames. ], tot_loss[loss=0.2876, simple_loss=0.346, pruned_loss=0.1146, over 3791870.26 frames. ], batch size: 58, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:50:26,343 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-04-01 04:50:43,927 INFO [train.py:903] (3/4) Epoch 5, batch 1800, loss[loss=0.2746, simple_loss=0.3275, pruned_loss=0.1108, over 19748.00 frames. ], tot_loss[loss=0.2872, simple_loss=0.3457, pruned_loss=0.1143, over 3798664.38 frames. ], batch size: 48, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:50:59,221 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0532, 1.0743, 1.5482, 0.5666, 2.4575, 2.3998, 2.1037, 2.4949], + device='cuda:3'), covar=tensor([0.1276, 0.2862, 0.2621, 0.1968, 0.0308, 0.0183, 0.0349, 0.0193], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0278, 0.0314, 0.0253, 0.0197, 0.0115, 0.0205, 0.0136], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:51:07,718 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.740e+02 6.050e+02 7.849e+02 1.011e+03 2.328e+03, threshold=1.570e+03, percent-clipped=7.0 +2023-04-01 04:51:39,753 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 04:51:44,194 INFO [train.py:903] (3/4) Epoch 5, batch 1850, loss[loss=0.3217, simple_loss=0.3759, pruned_loss=0.1337, over 19777.00 frames. ], tot_loss[loss=0.2855, simple_loss=0.3444, pruned_loss=0.1132, over 3805956.92 frames. ], batch size: 54, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:51:52,844 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-04-01 04:52:17,655 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 04:52:43,729 INFO [train.py:903] (3/4) Epoch 5, batch 1900, loss[loss=0.2773, simple_loss=0.3346, pruned_loss=0.11, over 19660.00 frames. ], tot_loss[loss=0.2863, simple_loss=0.3454, pruned_loss=0.1136, over 3802024.91 frames. ], batch size: 60, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:52:53,176 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29219.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:52:59,218 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=29223.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:53:01,719 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29225.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:53:02,418 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 04:53:07,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 04:53:11,316 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.676e+02 6.988e+02 9.073e+02 1.156e+03 1.890e+03, threshold=1.815e+03, percent-clipped=5.0 +2023-04-01 04:53:23,912 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29244.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:53:28,917 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 04:53:30,453 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:53:35,744 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29254.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:53:44,960 INFO [train.py:903] (3/4) Epoch 5, batch 1950, loss[loss=0.2365, simple_loss=0.3077, pruned_loss=0.08264, over 19663.00 frames. ], tot_loss[loss=0.2881, simple_loss=0.3467, pruned_loss=0.1147, over 3800754.31 frames. ], batch size: 53, lr: 1.67e-02, grad_scale: 4.0 +2023-04-01 04:54:46,876 INFO [train.py:903] (3/4) Epoch 5, batch 2000, loss[loss=0.2836, simple_loss=0.3455, pruned_loss=0.1108, over 19669.00 frames. ], tot_loss[loss=0.2871, simple_loss=0.3461, pruned_loss=0.114, over 3806824.32 frames. ], batch size: 60, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:54:56,674 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.00 vs. limit=5.0 +2023-04-01 04:55:10,280 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.142e+02 7.102e+02 9.141e+02 1.135e+03 3.050e+03, threshold=1.828e+03, percent-clipped=2.0 +2023-04-01 04:55:16,951 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29338.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:55:41,133 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2122, 1.1421, 1.4834, 0.9909, 2.6959, 3.3165, 3.2327, 3.5610], + device='cuda:3'), covar=tensor([0.1336, 0.2926, 0.2867, 0.2059, 0.0439, 0.0153, 0.0200, 0.0125], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0279, 0.0314, 0.0256, 0.0196, 0.0114, 0.0204, 0.0135], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 04:55:43,163 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 04:55:46,567 INFO [train.py:903] (3/4) Epoch 5, batch 2050, loss[loss=0.2989, simple_loss=0.3637, pruned_loss=0.117, over 19531.00 frames. ], tot_loss[loss=0.2868, simple_loss=0.3456, pruned_loss=0.114, over 3816514.80 frames. ], batch size: 54, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:55:49,013 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=29364.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:55:59,871 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 04:56:00,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 04:56:12,693 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1916, 1.2991, 1.6689, 1.2910, 2.4267, 2.1789, 2.5361, 0.9497], + device='cuda:3'), covar=tensor([0.1689, 0.2675, 0.1438, 0.1417, 0.1018, 0.1252, 0.1103, 0.2491], + device='cuda:3'), in_proj_covar=tensor([0.0438, 0.0497, 0.0469, 0.0403, 0.0533, 0.0431, 0.0616, 0.0433], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:56:16,134 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-01 04:56:23,070 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 04:56:30,418 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5802, 1.6522, 1.5551, 2.2355, 1.4323, 1.8225, 1.9733, 1.5851], + device='cuda:3'), covar=tensor([0.1705, 0.1275, 0.0808, 0.0657, 0.1416, 0.0641, 0.1477, 0.1333], + device='cuda:3'), in_proj_covar=tensor([0.0607, 0.0597, 0.0543, 0.0750, 0.0652, 0.0509, 0.0656, 0.0559], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:56:37,212 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3520, 2.3957, 2.1954, 3.4669, 2.2335, 3.5237, 3.4653, 2.5830], + device='cuda:3'), covar=tensor([0.1947, 0.1418, 0.0743, 0.0967, 0.1988, 0.0543, 0.1191, 0.1096], + device='cuda:3'), in_proj_covar=tensor([0.0606, 0.0596, 0.0541, 0.0750, 0.0652, 0.0509, 0.0656, 0.0559], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 04:56:45,829 INFO [train.py:903] (3/4) Epoch 5, batch 2100, loss[loss=0.2885, simple_loss=0.3458, pruned_loss=0.1156, over 19673.00 frames. ], tot_loss[loss=0.2856, simple_loss=0.3443, pruned_loss=0.1134, over 3832188.73 frames. ], batch size: 53, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:57:12,304 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.101e+02 7.145e+02 9.347e+02 1.270e+03 4.921e+03, threshold=1.869e+03, percent-clipped=10.0 +2023-04-01 04:57:13,529 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 04:57:15,696 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-01 04:57:34,270 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 04:57:46,676 INFO [train.py:903] (3/4) Epoch 5, batch 2150, loss[loss=0.2805, simple_loss=0.3245, pruned_loss=0.1182, over 19297.00 frames. ], tot_loss[loss=0.2873, simple_loss=0.3453, pruned_loss=0.1146, over 3820237.61 frames. ], batch size: 44, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:58:08,905 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29479.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 04:58:48,388 INFO [train.py:903] (3/4) Epoch 5, batch 2200, loss[loss=0.2654, simple_loss=0.3212, pruned_loss=0.1047, over 19812.00 frames. ], tot_loss[loss=0.2872, simple_loss=0.3454, pruned_loss=0.1145, over 3811790.98 frames. ], batch size: 49, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:59:11,935 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.753e+02 6.760e+02 8.285e+02 1.117e+03 1.782e+03, threshold=1.657e+03, percent-clipped=0.0 +2023-04-01 04:59:47,494 INFO [train.py:903] (3/4) Epoch 5, batch 2250, loss[loss=0.2727, simple_loss=0.3356, pruned_loss=0.1049, over 19568.00 frames. ], tot_loss[loss=0.2874, simple_loss=0.3454, pruned_loss=0.1147, over 3814424.48 frames. ], batch size: 52, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 04:59:55,558 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29569.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:00:28,223 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29594.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:00:32,406 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=29598.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:00:48,025 INFO [train.py:903] (3/4) Epoch 5, batch 2300, loss[loss=0.3065, simple_loss=0.3603, pruned_loss=0.1263, over 19298.00 frames. ], tot_loss[loss=0.2873, simple_loss=0.3457, pruned_loss=0.1144, over 3815983.01 frames. ], batch size: 66, lr: 1.66e-02, grad_scale: 8.0 +2023-04-01 05:00:56,357 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29619.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:01:03,672 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 05:01:15,152 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.694e+02 6.380e+02 7.654e+02 9.572e+02 1.859e+03, threshold=1.531e+03, percent-clipped=2.0 +2023-04-01 05:01:15,583 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1902, 2.1768, 1.5657, 1.4439, 1.9673, 1.0154, 1.2389, 1.6303], + device='cuda:3'), covar=tensor([0.0626, 0.0427, 0.0834, 0.0428, 0.0367, 0.1006, 0.0531, 0.0364], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0263, 0.0310, 0.0235, 0.0213, 0.0307, 0.0282, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:01:48,836 INFO [train.py:903] (3/4) Epoch 5, batch 2350, loss[loss=0.3331, simple_loss=0.3793, pruned_loss=0.1434, over 19610.00 frames. ], tot_loss[loss=0.2862, simple_loss=0.3451, pruned_loss=0.1137, over 3818319.58 frames. ], batch size: 57, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:02:11,376 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29680.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:02:28,866 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 05:02:37,413 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-01 05:02:45,270 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 05:02:49,428 INFO [train.py:903] (3/4) Epoch 5, batch 2400, loss[loss=0.3731, simple_loss=0.4017, pruned_loss=0.1723, over 12816.00 frames. ], tot_loss[loss=0.2865, simple_loss=0.3454, pruned_loss=0.1138, over 3820660.03 frames. ], batch size: 136, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:02:50,600 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29713.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:03:12,595 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.376e+02 6.843e+02 7.889e+02 1.103e+03 3.246e+03, threshold=1.578e+03, percent-clipped=5.0 +2023-04-01 05:03:15,307 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29735.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:03:46,799 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29760.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:03:48,490 INFO [train.py:903] (3/4) Epoch 5, batch 2450, loss[loss=0.3472, simple_loss=0.3876, pruned_loss=0.1534, over 19307.00 frames. ], tot_loss[loss=0.2839, simple_loss=0.3429, pruned_loss=0.1124, over 3834351.25 frames. ], batch size: 66, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:04:48,601 INFO [train.py:903] (3/4) Epoch 5, batch 2500, loss[loss=0.2871, simple_loss=0.3485, pruned_loss=0.1129, over 18847.00 frames. ], tot_loss[loss=0.285, simple_loss=0.3439, pruned_loss=0.113, over 3807996.85 frames. ], batch size: 74, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:04:57,792 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29820.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:05:14,588 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.471e+02 6.790e+02 8.500e+02 1.086e+03 2.138e+03, threshold=1.700e+03, percent-clipped=3.0 +2023-04-01 05:05:30,623 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2696, 0.8461, 1.0220, 2.0467, 1.5173, 1.2337, 1.9670, 1.3391], + device='cuda:3'), covar=tensor([0.1247, 0.2028, 0.1607, 0.1028, 0.1220, 0.1631, 0.1162, 0.1184], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0253, 0.0244, 0.0281, 0.0274, 0.0231, 0.0238, 0.0228], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0004, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:05:48,266 INFO [train.py:903] (3/4) Epoch 5, batch 2550, loss[loss=0.2822, simple_loss=0.3392, pruned_loss=0.1126, over 19397.00 frames. ], tot_loss[loss=0.285, simple_loss=0.3438, pruned_loss=0.1131, over 3814872.40 frames. ], batch size: 48, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:06:05,776 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 05:06:16,590 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.42 vs. limit=5.0 +2023-04-01 05:06:40,384 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 05:06:48,581 INFO [train.py:903] (3/4) Epoch 5, batch 2600, loss[loss=0.3197, simple_loss=0.3689, pruned_loss=0.1353, over 12836.00 frames. ], tot_loss[loss=0.2854, simple_loss=0.3442, pruned_loss=0.1133, over 3805812.01 frames. ], batch size: 137, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:06:50,696 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=29913.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:07:05,059 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8933, 1.8594, 1.9371, 2.9515, 1.8832, 2.7479, 2.6095, 1.8543], + device='cuda:3'), covar=tensor([0.1919, 0.1613, 0.0797, 0.0774, 0.1775, 0.0609, 0.1428, 0.1427], + device='cuda:3'), in_proj_covar=tensor([0.0607, 0.0597, 0.0541, 0.0755, 0.0646, 0.0516, 0.0654, 0.0563], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 05:07:13,515 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.137e+02 6.882e+02 9.175e+02 1.239e+03 1.828e+03, threshold=1.835e+03, percent-clipped=5.0 +2023-04-01 05:07:50,212 INFO [train.py:903] (3/4) Epoch 5, batch 2650, loss[loss=0.3582, simple_loss=0.3979, pruned_loss=0.1593, over 19639.00 frames. ], tot_loss[loss=0.2858, simple_loss=0.3449, pruned_loss=0.1134, over 3820471.10 frames. ], batch size: 60, lr: 1.65e-02, grad_scale: 8.0 +2023-04-01 05:07:58,283 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:08:08,343 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 05:08:29,828 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29994.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:08:48,839 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3670, 1.9357, 1.4276, 1.5491, 1.8016, 1.1794, 1.2556, 1.6355], + device='cuda:3'), covar=tensor([0.0616, 0.0396, 0.0614, 0.0376, 0.0285, 0.0756, 0.0490, 0.0291], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0266, 0.0313, 0.0237, 0.0221, 0.0310, 0.0288, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:08:50,698 INFO [train.py:903] (3/4) Epoch 5, batch 2700, loss[loss=0.2955, simple_loss=0.364, pruned_loss=0.1135, over 19686.00 frames. ], tot_loss[loss=0.2841, simple_loss=0.3439, pruned_loss=0.1122, over 3824758.23 frames. ], batch size: 59, lr: 1.64e-02, grad_scale: 8.0 +2023-04-01 05:09:04,701 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=30024.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:09:07,346 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-01 05:09:10,274 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30028.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:09:17,305 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.255e+02 6.674e+02 8.375e+02 9.964e+02 1.932e+03, threshold=1.675e+03, percent-clipped=1.0 +2023-04-01 05:09:42,500 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8600, 1.8417, 1.8334, 2.8244, 1.9825, 2.8289, 2.5703, 1.8270], + device='cuda:3'), covar=tensor([0.1970, 0.1585, 0.0862, 0.0877, 0.1711, 0.0564, 0.1483, 0.1438], + device='cuda:3'), in_proj_covar=tensor([0.0607, 0.0597, 0.0544, 0.0757, 0.0653, 0.0519, 0.0663, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 05:09:49,950 INFO [train.py:903] (3/4) Epoch 5, batch 2750, loss[loss=0.3349, simple_loss=0.388, pruned_loss=0.1409, over 18827.00 frames. ], tot_loss[loss=0.2853, simple_loss=0.3446, pruned_loss=0.113, over 3809956.62 frames. ], batch size: 74, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:10:50,795 INFO [train.py:903] (3/4) Epoch 5, batch 2800, loss[loss=0.2431, simple_loss=0.2975, pruned_loss=0.09435, over 19750.00 frames. ], tot_loss[loss=0.2847, simple_loss=0.3443, pruned_loss=0.1125, over 3820644.24 frames. ], batch size: 47, lr: 1.64e-02, grad_scale: 8.0 +2023-04-01 05:11:17,048 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.279e+02 7.041e+02 9.072e+02 1.191e+03 2.188e+03, threshold=1.814e+03, percent-clipped=6.0 +2023-04-01 05:11:22,788 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30139.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:11:27,366 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30143.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:11:51,984 INFO [train.py:903] (3/4) Epoch 5, batch 2850, loss[loss=0.3138, simple_loss=0.3753, pruned_loss=0.1262, over 19601.00 frames. ], tot_loss[loss=0.2869, simple_loss=0.3464, pruned_loss=0.1137, over 3822627.49 frames. ], batch size: 61, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:11:54,315 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=30164.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:12:15,909 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9502, 1.5387, 1.4811, 2.0133, 1.5768, 2.1178, 2.2507, 2.0980], + device='cuda:3'), covar=tensor([0.0698, 0.0983, 0.1036, 0.0991, 0.1023, 0.0757, 0.0819, 0.0599], + device='cuda:3'), in_proj_covar=tensor([0.0237, 0.0259, 0.0249, 0.0288, 0.0282, 0.0239, 0.0242, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:12:46,623 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2216, 2.1859, 1.5752, 1.3323, 1.9500, 1.0645, 1.1257, 1.8120], + device='cuda:3'), covar=tensor([0.0685, 0.0442, 0.0824, 0.0552, 0.0367, 0.1064, 0.0635, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0265, 0.0310, 0.0235, 0.0219, 0.0310, 0.0286, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:12:51,658 INFO [train.py:903] (3/4) Epoch 5, batch 2900, loss[loss=0.2837, simple_loss=0.3472, pruned_loss=0.1101, over 19766.00 frames. ], tot_loss[loss=0.2846, simple_loss=0.3443, pruned_loss=0.1124, over 3836588.96 frames. ], batch size: 54, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:12:51,680 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 05:13:09,293 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30227.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:13:20,147 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.023e+02 7.237e+02 9.091e+02 1.153e+03 2.755e+03, threshold=1.818e+03, percent-clipped=7.0 +2023-04-01 05:13:28,928 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7239, 1.4269, 1.3958, 1.9893, 1.4491, 1.9980, 2.0885, 1.9939], + device='cuda:3'), covar=tensor([0.0808, 0.1060, 0.1106, 0.0920, 0.1058, 0.0778, 0.0842, 0.0650], + device='cuda:3'), in_proj_covar=tensor([0.0238, 0.0259, 0.0248, 0.0286, 0.0280, 0.0237, 0.0242, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:13:51,431 INFO [train.py:903] (3/4) Epoch 5, batch 2950, loss[loss=0.2587, simple_loss=0.3153, pruned_loss=0.101, over 19754.00 frames. ], tot_loss[loss=0.2843, simple_loss=0.3445, pruned_loss=0.112, over 3837619.02 frames. ], batch size: 47, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:14:12,815 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30279.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:14:18,231 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30284.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:14:29,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7617, 1.4528, 1.3169, 2.0326, 1.3990, 2.1257, 2.0884, 2.0234], + device='cuda:3'), covar=tensor([0.0847, 0.1136, 0.1165, 0.0997, 0.1176, 0.0688, 0.0880, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0236, 0.0258, 0.0247, 0.0282, 0.0279, 0.0235, 0.0240, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:14:46,942 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30309.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:14:50,602 INFO [train.py:903] (3/4) Epoch 5, batch 3000, loss[loss=0.2998, simple_loss=0.3618, pruned_loss=0.1189, over 19674.00 frames. ], tot_loss[loss=0.2848, simple_loss=0.3451, pruned_loss=0.1123, over 3842551.62 frames. ], batch size: 59, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:14:50,602 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 05:15:03,126 INFO [train.py:937] (3/4) Epoch 5, validation: loss=0.2047, simple_loss=0.3034, pruned_loss=0.05296, over 944034.00 frames. +2023-04-01 05:15:03,127 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 05:15:05,713 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 05:15:27,705 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.77 vs. limit=5.0 +2023-04-01 05:15:33,556 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.690e+02 7.155e+02 8.736e+02 1.085e+03 2.346e+03, threshold=1.747e+03, percent-clipped=4.0 +2023-04-01 05:16:05,815 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.43 vs. limit=5.0 +2023-04-01 05:16:06,358 INFO [train.py:903] (3/4) Epoch 5, batch 3050, loss[loss=0.237, simple_loss=0.3012, pruned_loss=0.0864, over 19708.00 frames. ], tot_loss[loss=0.2843, simple_loss=0.3444, pruned_loss=0.1121, over 3834619.85 frames. ], batch size: 45, lr: 1.64e-02, grad_scale: 4.0 +2023-04-01 05:16:07,006 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-04-01 05:16:26,394 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30378.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:16:45,476 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30395.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:17:07,569 INFO [train.py:903] (3/4) Epoch 5, batch 3100, loss[loss=0.287, simple_loss=0.3374, pruned_loss=0.1183, over 19490.00 frames. ], tot_loss[loss=0.2838, simple_loss=0.344, pruned_loss=0.1118, over 3828324.68 frames. ], batch size: 49, lr: 1.63e-02, grad_scale: 4.0 +2023-04-01 05:17:17,194 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30420.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:17:26,216 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0170, 1.3737, 1.6023, 2.0670, 1.7487, 1.7666, 1.8622, 1.9890], + device='cuda:3'), covar=tensor([0.0779, 0.1601, 0.1217, 0.0754, 0.1067, 0.0430, 0.0845, 0.0537], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0368, 0.0285, 0.0241, 0.0302, 0.0247, 0.0269, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:17:33,697 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.705e+02 6.847e+02 8.274e+02 1.001e+03 3.134e+03, threshold=1.655e+03, percent-clipped=2.0 +2023-04-01 05:17:43,912 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2827, 1.1277, 1.2421, 1.4510, 2.1248, 1.0927, 1.6617, 2.1549], + device='cuda:3'), covar=tensor([0.0393, 0.1717, 0.1695, 0.1010, 0.0482, 0.1499, 0.1393, 0.0394], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0308, 0.0308, 0.0286, 0.0304, 0.0314, 0.0287, 0.0302], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:18:06,702 INFO [train.py:903] (3/4) Epoch 5, batch 3150, loss[loss=0.281, simple_loss=0.3466, pruned_loss=0.1077, over 19532.00 frames. ], tot_loss[loss=0.2842, simple_loss=0.3441, pruned_loss=0.1121, over 3830410.49 frames. ], batch size: 56, lr: 1.63e-02, grad_scale: 4.0 +2023-04-01 05:18:34,303 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 05:18:37,217 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=30487.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:19:06,273 INFO [train.py:903] (3/4) Epoch 5, batch 3200, loss[loss=0.263, simple_loss=0.3183, pruned_loss=0.1038, over 19759.00 frames. ], tot_loss[loss=0.2857, simple_loss=0.3454, pruned_loss=0.113, over 3833822.70 frames. ], batch size: 46, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:19:35,545 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.065e+02 7.005e+02 8.588e+02 1.128e+03 3.335e+03, threshold=1.718e+03, percent-clipped=13.0 +2023-04-01 05:19:35,975 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30535.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:20:04,489 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30560.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:20:06,323 INFO [train.py:903] (3/4) Epoch 5, batch 3250, loss[loss=0.2662, simple_loss=0.3326, pruned_loss=0.09985, over 19485.00 frames. ], tot_loss[loss=0.285, simple_loss=0.345, pruned_loss=0.1125, over 3840230.11 frames. ], batch size: 49, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:20:20,424 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=30571.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:20:55,355 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30602.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:21:08,678 INFO [train.py:903] (3/4) Epoch 5, batch 3300, loss[loss=0.2732, simple_loss=0.3302, pruned_loss=0.1081, over 19764.00 frames. ], tot_loss[loss=0.2853, simple_loss=0.3456, pruned_loss=0.1125, over 3849487.54 frames. ], batch size: 47, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:21:16,490 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 05:21:29,548 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.26 vs. limit=5.0 +2023-04-01 05:21:35,516 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.601e+02 6.699e+02 7.755e+02 9.198e+02 2.155e+03, threshold=1.551e+03, percent-clipped=1.0 +2023-04-01 05:22:09,095 INFO [train.py:903] (3/4) Epoch 5, batch 3350, loss[loss=0.3032, simple_loss=0.3657, pruned_loss=0.1204, over 19538.00 frames. ], tot_loss[loss=0.2854, simple_loss=0.3455, pruned_loss=0.1126, over 3832157.00 frames. ], batch size: 56, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:22:21,653 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30673.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:22:39,147 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30686.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:22:49,175 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8782, 3.5743, 2.3804, 3.2729, 1.0347, 3.2375, 3.1050, 3.2831], + device='cuda:3'), covar=tensor([0.0876, 0.1367, 0.2012, 0.0755, 0.3797, 0.0989, 0.0928, 0.0998], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0315, 0.0366, 0.0288, 0.0355, 0.0303, 0.0282, 0.0316], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:23:09,774 INFO [train.py:903] (3/4) Epoch 5, batch 3400, loss[loss=0.2867, simple_loss=0.3262, pruned_loss=0.1236, over 19736.00 frames. ], tot_loss[loss=0.2843, simple_loss=0.3443, pruned_loss=0.1122, over 3828626.16 frames. ], batch size: 46, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:23:22,327 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=30722.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:23:39,932 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.254e+02 6.893e+02 8.724e+02 1.073e+03 2.213e+03, threshold=1.745e+03, percent-clipped=7.0 +2023-04-01 05:23:57,768 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2303, 1.0575, 1.0190, 1.4093, 1.2353, 1.2717, 1.3983, 1.1784], + device='cuda:3'), covar=tensor([0.1030, 0.1268, 0.1426, 0.0760, 0.0920, 0.0957, 0.0936, 0.0912], + device='cuda:3'), in_proj_covar=tensor([0.0237, 0.0256, 0.0251, 0.0282, 0.0276, 0.0232, 0.0237, 0.0228], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:24:11,824 INFO [train.py:903] (3/4) Epoch 5, batch 3450, loss[loss=0.2314, simple_loss=0.2999, pruned_loss=0.08145, over 19399.00 frames. ], tot_loss[loss=0.2831, simple_loss=0.3433, pruned_loss=0.1115, over 3825607.37 frames. ], batch size: 48, lr: 1.63e-02, grad_scale: 8.0 +2023-04-01 05:24:15,168 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 05:24:27,457 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0006, 1.4306, 1.5472, 1.9170, 1.8308, 1.8181, 1.6924, 1.9379], + device='cuda:3'), covar=tensor([0.0718, 0.1392, 0.1238, 0.0785, 0.1004, 0.0416, 0.0813, 0.0512], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0358, 0.0278, 0.0235, 0.0301, 0.0242, 0.0265, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:24:52,628 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2881, 2.8487, 2.0841, 2.2231, 1.8590, 2.4135, 0.7177, 2.1533], + device='cuda:3'), covar=tensor([0.0257, 0.0248, 0.0236, 0.0387, 0.0496, 0.0354, 0.0550, 0.0440], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0280, 0.0279, 0.0302, 0.0370, 0.0293, 0.0276, 0.0287], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:25:13,755 INFO [train.py:903] (3/4) Epoch 5, batch 3500, loss[loss=0.2737, simple_loss=0.3404, pruned_loss=0.1036, over 18139.00 frames. ], tot_loss[loss=0.2833, simple_loss=0.3435, pruned_loss=0.1116, over 3824815.15 frames. ], batch size: 83, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:25:39,221 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.495e+02 7.015e+02 8.195e+02 1.097e+03 2.546e+03, threshold=1.639e+03, percent-clipped=5.0 +2023-04-01 05:25:41,891 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30837.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:26:09,323 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30858.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:26:13,374 INFO [train.py:903] (3/4) Epoch 5, batch 3550, loss[loss=0.2861, simple_loss=0.3554, pruned_loss=0.1084, over 19776.00 frames. ], tot_loss[loss=0.2858, simple_loss=0.3456, pruned_loss=0.113, over 3821435.54 frames. ], batch size: 56, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:26:15,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5190, 2.5178, 1.7385, 1.5534, 2.2084, 1.3229, 1.2566, 1.6298], + device='cuda:3'), covar=tensor([0.0647, 0.0308, 0.0615, 0.0432, 0.0320, 0.0688, 0.0538, 0.0351], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0271, 0.0316, 0.0244, 0.0223, 0.0313, 0.0288, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:26:38,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30883.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:27:13,894 INFO [train.py:903] (3/4) Epoch 5, batch 3600, loss[loss=0.3329, simple_loss=0.3812, pruned_loss=0.1423, over 19810.00 frames. ], tot_loss[loss=0.2836, simple_loss=0.3438, pruned_loss=0.1117, over 3839143.14 frames. ], batch size: 56, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:27:24,165 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30921.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:27:43,234 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.778e+02 6.769e+02 8.289e+02 1.066e+03 2.218e+03, threshold=1.658e+03, percent-clipped=4.0 +2023-04-01 05:27:51,226 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30942.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:27:52,206 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30943.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:28:02,213 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30952.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:28:13,484 INFO [train.py:903] (3/4) Epoch 5, batch 3650, loss[loss=0.2856, simple_loss=0.3543, pruned_loss=0.1084, over 19669.00 frames. ], tot_loss[loss=0.2833, simple_loss=0.3433, pruned_loss=0.1116, over 3832510.62 frames. ], batch size: 58, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:28:20,070 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30967.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:28:45,370 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1930, 1.3214, 1.1018, 1.0125, 0.9706, 1.1347, 0.1028, 0.3928], + device='cuda:3'), covar=tensor([0.0254, 0.0246, 0.0159, 0.0181, 0.0500, 0.0195, 0.0466, 0.0423], + device='cuda:3'), in_proj_covar=tensor([0.0286, 0.0284, 0.0287, 0.0309, 0.0376, 0.0298, 0.0280, 0.0293], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:29:02,305 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1655, 2.1250, 1.6237, 1.4329, 1.9435, 1.0463, 1.0714, 1.6325], + device='cuda:3'), covar=tensor([0.0715, 0.0498, 0.0875, 0.0500, 0.0406, 0.1039, 0.0585, 0.0388], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0274, 0.0317, 0.0243, 0.0225, 0.0314, 0.0286, 0.0256], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:29:12,118 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.66 vs. limit=5.0 +2023-04-01 05:29:14,322 INFO [train.py:903] (3/4) Epoch 5, batch 3700, loss[loss=0.2593, simple_loss=0.3143, pruned_loss=0.1022, over 19780.00 frames. ], tot_loss[loss=0.2841, simple_loss=0.3443, pruned_loss=0.112, over 3839210.87 frames. ], batch size: 48, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:29:21,014 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31017.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:29:34,019 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31029.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:29:40,640 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.405e+02 7.099e+02 8.962e+02 1.140e+03 3.223e+03, threshold=1.792e+03, percent-clipped=9.0 +2023-04-01 05:30:15,306 INFO [train.py:903] (3/4) Epoch 5, batch 3750, loss[loss=0.331, simple_loss=0.3746, pruned_loss=0.1437, over 13360.00 frames. ], tot_loss[loss=0.2861, simple_loss=0.3455, pruned_loss=0.1133, over 3836681.75 frames. ], batch size: 135, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:30:53,648 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31093.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:31:15,740 INFO [train.py:903] (3/4) Epoch 5, batch 3800, loss[loss=0.2837, simple_loss=0.3507, pruned_loss=0.1084, over 19666.00 frames. ], tot_loss[loss=0.286, simple_loss=0.3455, pruned_loss=0.1133, over 3840849.93 frames. ], batch size: 55, lr: 1.62e-02, grad_scale: 8.0 +2023-04-01 05:31:22,823 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31118.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:31:25,989 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4504, 1.8084, 1.9955, 2.6188, 2.4098, 2.1082, 1.9158, 2.5886], + device='cuda:3'), covar=tensor([0.0690, 0.1529, 0.1166, 0.0665, 0.0941, 0.0418, 0.0910, 0.0450], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0359, 0.0279, 0.0231, 0.0299, 0.0240, 0.0265, 0.0225], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:31:40,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31132.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:31:44,606 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.073e+02 5.856e+02 8.467e+02 1.115e+03 2.554e+03, threshold=1.693e+03, percent-clipped=5.0 +2023-04-01 05:31:49,119 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 05:32:15,363 INFO [train.py:903] (3/4) Epoch 5, batch 3850, loss[loss=0.2976, simple_loss=0.3631, pruned_loss=0.116, over 19669.00 frames. ], tot_loss[loss=0.2863, simple_loss=0.3455, pruned_loss=0.1136, over 3825325.23 frames. ], batch size: 60, lr: 1.61e-02, grad_scale: 8.0 +2023-04-01 05:32:37,689 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.73 vs. limit=2.0 +2023-04-01 05:33:19,169 INFO [train.py:903] (3/4) Epoch 5, batch 3900, loss[loss=0.3215, simple_loss=0.3713, pruned_loss=0.1358, over 19597.00 frames. ], tot_loss[loss=0.2868, simple_loss=0.3459, pruned_loss=0.1138, over 3819849.02 frames. ], batch size: 61, lr: 1.61e-02, grad_scale: 8.0 +2023-04-01 05:33:45,289 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.653e+02 6.709e+02 8.069e+02 1.055e+03 2.198e+03, threshold=1.614e+03, percent-clipped=3.0 +2023-04-01 05:34:10,691 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1798, 3.7315, 2.2346, 2.3767, 3.3491, 1.8309, 1.1893, 1.9246], + device='cuda:3'), covar=tensor([0.0910, 0.0335, 0.0751, 0.0529, 0.0406, 0.0832, 0.0800, 0.0589], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0270, 0.0313, 0.0233, 0.0219, 0.0306, 0.0279, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:34:18,968 INFO [train.py:903] (3/4) Epoch 5, batch 3950, loss[loss=0.2788, simple_loss=0.3386, pruned_loss=0.1095, over 19651.00 frames. ], tot_loss[loss=0.286, simple_loss=0.3452, pruned_loss=0.1134, over 3821128.68 frames. ], batch size: 53, lr: 1.61e-02, grad_scale: 4.0 +2023-04-01 05:34:22,383 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31265.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:34:24,551 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 05:34:47,812 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31287.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:35:00,481 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31296.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:35:18,261 INFO [train.py:903] (3/4) Epoch 5, batch 4000, loss[loss=0.3037, simple_loss=0.3681, pruned_loss=0.1197, over 18736.00 frames. ], tot_loss[loss=0.2866, simple_loss=0.3455, pruned_loss=0.1138, over 3814309.66 frames. ], batch size: 74, lr: 1.61e-02, grad_scale: 8.0 +2023-04-01 05:35:48,954 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.459e+02 7.061e+02 8.767e+02 1.081e+03 2.366e+03, threshold=1.753e+03, percent-clipped=7.0 +2023-04-01 05:36:06,101 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 05:36:18,327 INFO [train.py:903] (3/4) Epoch 5, batch 4050, loss[loss=0.2838, simple_loss=0.3526, pruned_loss=0.1075, over 19614.00 frames. ], tot_loss[loss=0.2858, simple_loss=0.3449, pruned_loss=0.1134, over 3821271.22 frames. ], batch size: 57, lr: 1.61e-02, grad_scale: 4.0 +2023-04-01 05:36:20,397 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-01 05:36:33,372 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31373.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 05:36:42,206 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31380.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:36:51,301 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31388.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:37:07,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31402.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:37:18,623 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31411.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:37:19,400 INFO [train.py:903] (3/4) Epoch 5, batch 4100, loss[loss=0.2454, simple_loss=0.3147, pruned_loss=0.08804, over 19734.00 frames. ], tot_loss[loss=0.2854, simple_loss=0.3446, pruned_loss=0.1132, over 3824857.07 frames. ], batch size: 51, lr: 1.61e-02, grad_scale: 4.0 +2023-04-01 05:37:20,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31413.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:37:37,698 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31427.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:37:48,916 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.057e+02 5.964e+02 7.397e+02 9.605e+02 3.908e+03, threshold=1.479e+03, percent-clipped=5.0 +2023-04-01 05:37:49,658 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 05:37:53,720 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 05:38:20,925 INFO [train.py:903] (3/4) Epoch 5, batch 4150, loss[loss=0.2728, simple_loss=0.327, pruned_loss=0.1092, over 19711.00 frames. ], tot_loss[loss=0.2846, simple_loss=0.3442, pruned_loss=0.1125, over 3822454.81 frames. ], batch size: 46, lr: 1.61e-02, grad_scale: 4.0 +2023-04-01 05:38:51,006 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31488.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:39:16,906 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31509.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 05:39:20,064 INFO [train.py:903] (3/4) Epoch 5, batch 4200, loss[loss=0.2879, simple_loss=0.3511, pruned_loss=0.1124, over 18000.00 frames. ], tot_loss[loss=0.2852, simple_loss=0.3446, pruned_loss=0.1129, over 3816512.50 frames. ], batch size: 83, lr: 1.61e-02, grad_scale: 4.0 +2023-04-01 05:39:23,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 05:39:50,934 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.017e+02 6.856e+02 8.101e+02 1.045e+03 3.023e+03, threshold=1.620e+03, percent-clipped=6.0 +2023-04-01 05:40:19,245 INFO [train.py:903] (3/4) Epoch 5, batch 4250, loss[loss=0.2057, simple_loss=0.2765, pruned_loss=0.0674, over 19746.00 frames. ], tot_loss[loss=0.2833, simple_loss=0.3431, pruned_loss=0.1117, over 3823749.46 frames. ], batch size: 46, lr: 1.60e-02, grad_scale: 4.0 +2023-04-01 05:40:21,995 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3730, 1.0805, 1.0603, 1.5685, 1.1308, 1.5046, 1.6948, 1.4962], + device='cuda:3'), covar=tensor([0.0882, 0.1169, 0.1291, 0.0941, 0.1069, 0.0845, 0.0952, 0.0722], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0253, 0.0245, 0.0280, 0.0274, 0.0234, 0.0237, 0.0223], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:40:35,456 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 05:40:46,345 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 05:41:20,812 INFO [train.py:903] (3/4) Epoch 5, batch 4300, loss[loss=0.2385, simple_loss=0.3007, pruned_loss=0.08819, over 16083.00 frames. ], tot_loss[loss=0.2833, simple_loss=0.3431, pruned_loss=0.1118, over 3815727.31 frames. ], batch size: 35, lr: 1.60e-02, grad_scale: 4.0 +2023-04-01 05:41:42,005 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7830, 1.0800, 1.1586, 2.0003, 1.7065, 1.6374, 1.9078, 1.7780], + device='cuda:3'), covar=tensor([0.0735, 0.1257, 0.1163, 0.0903, 0.0957, 0.0873, 0.0895, 0.0726], + device='cuda:3'), in_proj_covar=tensor([0.0229, 0.0251, 0.0245, 0.0277, 0.0271, 0.0232, 0.0233, 0.0221], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 05:41:51,068 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31636.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:41:51,770 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.826e+02 7.144e+02 8.425e+02 1.091e+03 2.021e+03, threshold=1.685e+03, percent-clipped=5.0 +2023-04-01 05:42:13,884 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 05:42:18,615 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31658.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:42:21,709 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31661.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:42:22,335 INFO [train.py:903] (3/4) Epoch 5, batch 4350, loss[loss=0.333, simple_loss=0.3758, pruned_loss=0.1452, over 19643.00 frames. ], tot_loss[loss=0.2849, simple_loss=0.3442, pruned_loss=0.1128, over 3812333.94 frames. ], batch size: 58, lr: 1.60e-02, grad_scale: 4.0 +2023-04-01 05:42:28,428 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31667.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:42:45,966 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31683.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:42:57,836 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31692.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:43:22,084 INFO [train.py:903] (3/4) Epoch 5, batch 4400, loss[loss=0.3596, simple_loss=0.3922, pruned_loss=0.1635, over 13689.00 frames. ], tot_loss[loss=0.2864, simple_loss=0.3455, pruned_loss=0.1137, over 3798357.01 frames. ], batch size: 136, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:43:33,985 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31722.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:43:44,926 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 05:43:53,319 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.183e+02 7.191e+02 8.879e+02 1.082e+03 1.961e+03, threshold=1.776e+03, percent-clipped=1.0 +2023-04-01 05:43:55,412 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 05:44:00,347 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8074, 1.9607, 2.3309, 2.2351, 3.0003, 3.5272, 3.6316, 3.7556], + device='cuda:3'), covar=tensor([0.1174, 0.2116, 0.2056, 0.1370, 0.0685, 0.0210, 0.0151, 0.0119], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0279, 0.0315, 0.0251, 0.0198, 0.0119, 0.0203, 0.0138], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:44:02,740 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31744.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:44:20,764 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2832, 1.3002, 1.5413, 1.2425, 2.7211, 3.6043, 3.5963, 3.9144], + device='cuda:3'), covar=tensor([0.1364, 0.2870, 0.2838, 0.1821, 0.0451, 0.0109, 0.0179, 0.0103], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0282, 0.0318, 0.0253, 0.0200, 0.0120, 0.0204, 0.0139], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:44:22,687 INFO [train.py:903] (3/4) Epoch 5, batch 4450, loss[loss=0.222, simple_loss=0.2991, pruned_loss=0.07248, over 19845.00 frames. ], tot_loss[loss=0.285, simple_loss=0.3442, pruned_loss=0.1128, over 3807041.93 frames. ], batch size: 52, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:44:30,939 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31769.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:44:32,881 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31771.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:45:16,703 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31806.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:45:23,179 INFO [train.py:903] (3/4) Epoch 5, batch 4500, loss[loss=0.2583, simple_loss=0.332, pruned_loss=0.09228, over 19540.00 frames. ], tot_loss[loss=0.285, simple_loss=0.3445, pruned_loss=0.1128, over 3815578.00 frames. ], batch size: 54, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:45:53,689 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.229e+02 6.666e+02 8.776e+02 1.149e+03 2.550e+03, threshold=1.755e+03, percent-clipped=7.0 +2023-04-01 05:45:54,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9063, 2.5808, 1.7052, 1.8010, 1.7668, 1.8164, 0.5694, 2.0309], + device='cuda:3'), covar=tensor([0.0375, 0.0372, 0.0451, 0.0620, 0.0703, 0.0613, 0.0733, 0.0583], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0280, 0.0279, 0.0301, 0.0374, 0.0293, 0.0273, 0.0289], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:45:58,554 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:46:12,840 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=31853.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 05:46:24,486 INFO [train.py:903] (3/4) Epoch 5, batch 4550, loss[loss=0.2287, simple_loss=0.2932, pruned_loss=0.08203, over 19394.00 frames. ], tot_loss[loss=0.2839, simple_loss=0.3438, pruned_loss=0.112, over 3818045.65 frames. ], batch size: 48, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:46:30,251 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 05:46:51,959 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 05:46:52,276 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31886.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:46:54,762 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1002, 1.9976, 1.9209, 3.2164, 2.1082, 3.0780, 3.0184, 1.8070], + device='cuda:3'), covar=tensor([0.2332, 0.1997, 0.0933, 0.1166, 0.2303, 0.0744, 0.1596, 0.1683], + device='cuda:3'), in_proj_covar=tensor([0.0626, 0.0616, 0.0555, 0.0771, 0.0661, 0.0531, 0.0682, 0.0585], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 05:47:06,472 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31896.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:47:25,656 INFO [train.py:903] (3/4) Epoch 5, batch 4600, loss[loss=0.3564, simple_loss=0.394, pruned_loss=0.1594, over 13477.00 frames. ], tot_loss[loss=0.2832, simple_loss=0.3432, pruned_loss=0.1116, over 3807636.31 frames. ], batch size: 135, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:47:50,399 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31933.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:47:55,744 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.582e+02 6.775e+02 7.918e+02 9.900e+02 3.222e+03, threshold=1.584e+03, percent-clipped=3.0 +2023-04-01 05:48:25,675 INFO [train.py:903] (3/4) Epoch 5, batch 4650, loss[loss=0.4524, simple_loss=0.4554, pruned_loss=0.2247, over 13395.00 frames. ], tot_loss[loss=0.2838, simple_loss=0.3436, pruned_loss=0.112, over 3806152.97 frames. ], batch size: 136, lr: 1.60e-02, grad_scale: 8.0 +2023-04-01 05:48:32,757 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31968.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:48:32,827 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31968.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:48:41,063 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 05:48:43,626 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5111, 1.5196, 1.6368, 2.3504, 1.9173, 2.1832, 2.6609, 2.1566], + device='cuda:3'), covar=tensor([0.0717, 0.1303, 0.1206, 0.1122, 0.1068, 0.0874, 0.0920, 0.0777], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0257, 0.0248, 0.0282, 0.0273, 0.0235, 0.0236, 0.0226], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 05:48:52,753 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 05:49:00,905 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0386, 2.0890, 2.1013, 2.9941, 2.4407, 2.8065, 2.6612, 2.0637], + device='cuda:3'), covar=tensor([0.1828, 0.1320, 0.0716, 0.0747, 0.1326, 0.0468, 0.1151, 0.1169], + device='cuda:3'), in_proj_covar=tensor([0.0626, 0.0613, 0.0552, 0.0768, 0.0656, 0.0529, 0.0679, 0.0581], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 05:49:25,341 INFO [train.py:903] (3/4) Epoch 5, batch 4700, loss[loss=0.3595, simple_loss=0.3989, pruned_loss=0.16, over 13045.00 frames. ], tot_loss[loss=0.2834, simple_loss=0.343, pruned_loss=0.1119, over 3804069.84 frames. ], batch size: 136, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:49:37,012 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-04-01 05:49:48,442 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 05:49:56,507 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.386e+02 6.540e+02 8.637e+02 1.061e+03 2.519e+03, threshold=1.727e+03, percent-clipped=5.0 +2023-04-01 05:50:27,325 INFO [train.py:903] (3/4) Epoch 5, batch 4750, loss[loss=0.2739, simple_loss=0.3207, pruned_loss=0.1135, over 19320.00 frames. ], tot_loss[loss=0.2817, simple_loss=0.3413, pruned_loss=0.111, over 3822272.35 frames. ], batch size: 44, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:50:32,837 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32066.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:50:34,227 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8655, 1.3152, 0.9981, 0.9865, 1.2162, 0.8529, 0.7240, 1.2640], + device='cuda:3'), covar=tensor([0.0522, 0.0565, 0.0923, 0.0432, 0.0373, 0.1029, 0.0537, 0.0372], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0264, 0.0312, 0.0234, 0.0215, 0.0309, 0.0281, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 05:51:28,013 INFO [train.py:903] (3/4) Epoch 5, batch 4800, loss[loss=0.2692, simple_loss=0.3348, pruned_loss=0.1018, over 19614.00 frames. ], tot_loss[loss=0.2831, simple_loss=0.3431, pruned_loss=0.1115, over 3834890.28 frames. ], batch size: 61, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:51:57,642 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.741e+02 7.051e+02 8.763e+02 1.042e+03 3.094e+03, threshold=1.753e+03, percent-clipped=4.0 +2023-04-01 05:52:01,617 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1541, 1.3733, 1.6837, 1.2695, 2.8626, 3.5162, 3.3997, 3.7312], + device='cuda:3'), covar=tensor([0.1594, 0.2798, 0.2885, 0.1958, 0.0442, 0.0248, 0.0200, 0.0119], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0280, 0.0317, 0.0251, 0.0197, 0.0117, 0.0204, 0.0138], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:52:04,804 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32142.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:52:14,250 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32150.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:52:27,937 INFO [train.py:903] (3/4) Epoch 5, batch 4850, loss[loss=0.2566, simple_loss=0.3364, pruned_loss=0.08841, over 19645.00 frames. ], tot_loss[loss=0.2826, simple_loss=0.3427, pruned_loss=0.1112, over 3832797.83 frames. ], batch size: 58, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:52:34,183 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32167.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:52:51,467 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32181.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:52:54,109 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 05:52:57,426 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32185.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:53:12,908 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 05:53:17,713 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 05:53:18,687 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 05:53:21,340 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8067, 4.1762, 4.4732, 4.4216, 1.7776, 4.0467, 3.6709, 4.0561], + device='cuda:3'), covar=tensor([0.0845, 0.0722, 0.0472, 0.0399, 0.3897, 0.0436, 0.0498, 0.0966], + device='cuda:3'), in_proj_covar=tensor([0.0509, 0.0454, 0.0591, 0.0499, 0.0578, 0.0362, 0.0392, 0.0558], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 05:53:26,880 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 05:53:27,951 INFO [train.py:903] (3/4) Epoch 5, batch 4900, loss[loss=0.357, simple_loss=0.3963, pruned_loss=0.1588, over 13766.00 frames. ], tot_loss[loss=0.2829, simple_loss=0.343, pruned_loss=0.1114, over 3816532.32 frames. ], batch size: 136, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:53:45,222 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32224.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:53:48,942 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 05:53:59,194 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.307e+02 6.585e+02 7.912e+02 1.039e+03 2.328e+03, threshold=1.582e+03, percent-clipped=1.0 +2023-04-01 05:54:02,583 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32240.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:54:13,283 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32249.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 05:54:16,862 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-01 05:54:29,206 INFO [train.py:903] (3/4) Epoch 5, batch 4950, loss[loss=0.2309, simple_loss=0.3016, pruned_loss=0.08012, over 19479.00 frames. ], tot_loss[loss=0.2837, simple_loss=0.3437, pruned_loss=0.1119, over 3795721.41 frames. ], batch size: 49, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:54:33,932 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32265.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:54:48,159 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32277.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:54:49,209 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 05:55:12,820 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 05:55:15,577 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32300.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:55:30,487 INFO [train.py:903] (3/4) Epoch 5, batch 5000, loss[loss=0.3172, simple_loss=0.3664, pruned_loss=0.1339, over 19682.00 frames. ], tot_loss[loss=0.2847, simple_loss=0.3441, pruned_loss=0.1126, over 3795590.68 frames. ], batch size: 59, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:55:30,633 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=32312.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:55:40,423 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 05:55:50,530 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 05:55:58,320 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.834e+02 7.030e+02 8.789e+02 1.165e+03 2.289e+03, threshold=1.758e+03, percent-clipped=7.0 +2023-04-01 05:56:21,778 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32355.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:56:29,553 INFO [train.py:903] (3/4) Epoch 5, batch 5050, loss[loss=0.2747, simple_loss=0.3452, pruned_loss=0.1021, over 18208.00 frames. ], tot_loss[loss=0.2857, simple_loss=0.3451, pruned_loss=0.1131, over 3791013.39 frames. ], batch size: 83, lr: 1.59e-02, grad_scale: 8.0 +2023-04-01 05:57:05,113 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 05:57:07,770 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32392.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:57:30,263 INFO [train.py:903] (3/4) Epoch 5, batch 5100, loss[loss=0.3014, simple_loss=0.3636, pruned_loss=0.1196, over 19663.00 frames. ], tot_loss[loss=0.2848, simple_loss=0.3444, pruned_loss=0.1127, over 3807483.05 frames. ], batch size: 60, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 05:57:40,583 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 05:57:44,922 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 05:57:51,244 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 05:57:51,520 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32427.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:58:01,497 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4787, 1.1581, 1.5895, 1.2151, 2.7026, 3.6269, 3.5213, 3.8646], + device='cuda:3'), covar=tensor([0.1336, 0.3049, 0.2890, 0.1905, 0.0443, 0.0131, 0.0191, 0.0111], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0279, 0.0316, 0.0250, 0.0195, 0.0115, 0.0204, 0.0137], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 05:58:02,218 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.811e+02 6.879e+02 8.267e+02 1.044e+03 2.791e+03, threshold=1.653e+03, percent-clipped=3.0 +2023-04-01 05:58:02,655 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32437.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:58:20,131 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9211, 1.6460, 1.6372, 2.2430, 1.7143, 2.2198, 2.1562, 2.0886], + device='cuda:3'), covar=tensor([0.0769, 0.0944, 0.0994, 0.0878, 0.0945, 0.0721, 0.0884, 0.0617], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0248, 0.0244, 0.0276, 0.0267, 0.0232, 0.0230, 0.0223], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 05:58:32,011 INFO [train.py:903] (3/4) Epoch 5, batch 5150, loss[loss=0.3529, simple_loss=0.3955, pruned_loss=0.1551, over 19530.00 frames. ], tot_loss[loss=0.2832, simple_loss=0.3431, pruned_loss=0.1117, over 3808632.90 frames. ], batch size: 54, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 05:58:32,383 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32462.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:58:45,277 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 05:59:19,201 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 05:59:33,452 INFO [train.py:903] (3/4) Epoch 5, batch 5200, loss[loss=0.2672, simple_loss=0.3378, pruned_loss=0.09831, over 19777.00 frames. ], tot_loss[loss=0.2818, simple_loss=0.342, pruned_loss=0.1108, over 3803868.81 frames. ], batch size: 56, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 05:59:43,942 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32521.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 05:59:45,955 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 06:00:02,762 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.318e+02 6.175e+02 7.903e+02 1.065e+03 1.799e+03, threshold=1.581e+03, percent-clipped=1.0 +2023-04-01 06:00:15,194 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32546.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:00:26,342 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32556.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:00:29,355 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 06:00:32,702 INFO [train.py:903] (3/4) Epoch 5, batch 5250, loss[loss=0.3066, simple_loss=0.3674, pruned_loss=0.1229, over 18104.00 frames. ], tot_loss[loss=0.2822, simple_loss=0.3426, pruned_loss=0.1109, over 3815265.13 frames. ], batch size: 83, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 06:00:55,249 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32581.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:01:32,207 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32611.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:01:32,915 INFO [train.py:903] (3/4) Epoch 5, batch 5300, loss[loss=0.3231, simple_loss=0.3697, pruned_loss=0.1383, over 19679.00 frames. ], tot_loss[loss=0.2806, simple_loss=0.3416, pruned_loss=0.1098, over 3827899.10 frames. ], batch size: 53, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 06:01:41,079 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1765, 1.2071, 1.8402, 1.4263, 2.5847, 2.1618, 2.8314, 1.0994], + device='cuda:3'), covar=tensor([0.1948, 0.3349, 0.1667, 0.1584, 0.1268, 0.1609, 0.1336, 0.2968], + device='cuda:3'), in_proj_covar=tensor([0.0446, 0.0507, 0.0480, 0.0407, 0.0555, 0.0446, 0.0625, 0.0442], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:01:51,391 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 06:02:04,555 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32636.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:02:05,308 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.788e+02 6.583e+02 8.041e+02 1.027e+03 2.106e+03, threshold=1.608e+03, percent-clipped=4.0 +2023-04-01 06:02:17,829 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32648.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:02:34,326 INFO [train.py:903] (3/4) Epoch 5, batch 5350, loss[loss=0.2798, simple_loss=0.3307, pruned_loss=0.1144, over 19714.00 frames. ], tot_loss[loss=0.282, simple_loss=0.3427, pruned_loss=0.1107, over 3807682.11 frames. ], batch size: 51, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 06:02:36,861 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2566, 1.2689, 1.8119, 1.4978, 2.5149, 2.2784, 2.6489, 0.8981], + device='cuda:3'), covar=tensor([0.1624, 0.2794, 0.1431, 0.1327, 0.1084, 0.1251, 0.1283, 0.2655], + device='cuda:3'), in_proj_covar=tensor([0.0441, 0.0502, 0.0475, 0.0403, 0.0549, 0.0439, 0.0621, 0.0437], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:02:44,362 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5029, 1.1046, 1.3157, 1.3424, 2.1128, 0.9949, 1.8936, 2.1887], + device='cuda:3'), covar=tensor([0.0558, 0.2420, 0.2247, 0.1348, 0.0799, 0.1758, 0.0880, 0.0564], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0314, 0.0316, 0.0286, 0.0306, 0.0316, 0.0289, 0.0306], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:02:48,913 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32673.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:03:00,322 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32683.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:03:06,684 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 06:03:31,063 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32708.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:03:35,126 INFO [train.py:903] (3/4) Epoch 5, batch 5400, loss[loss=0.2889, simple_loss=0.3477, pruned_loss=0.1151, over 19582.00 frames. ], tot_loss[loss=0.2818, simple_loss=0.3423, pruned_loss=0.1107, over 3814236.86 frames. ], batch size: 52, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 06:03:48,191 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.79 vs. limit=5.0 +2023-04-01 06:04:03,210 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.192e+02 6.673e+02 8.431e+02 1.064e+03 2.658e+03, threshold=1.686e+03, percent-clipped=8.0 +2023-04-01 06:04:21,185 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9951, 2.0716, 1.9476, 3.0134, 1.9369, 2.8927, 2.6766, 1.9070], + device='cuda:3'), covar=tensor([0.2070, 0.1619, 0.0839, 0.0929, 0.1997, 0.0636, 0.1482, 0.1403], + device='cuda:3'), in_proj_covar=tensor([0.0628, 0.0621, 0.0554, 0.0777, 0.0659, 0.0541, 0.0678, 0.0586], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:04:34,474 INFO [train.py:903] (3/4) Epoch 5, batch 5450, loss[loss=0.2614, simple_loss=0.33, pruned_loss=0.09641, over 19799.00 frames. ], tot_loss[loss=0.2819, simple_loss=0.3427, pruned_loss=0.1106, over 3811903.27 frames. ], batch size: 56, lr: 1.58e-02, grad_scale: 8.0 +2023-04-01 06:04:47,441 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.58 vs. limit=5.0 +2023-04-01 06:05:26,452 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.58 vs. limit=5.0 +2023-04-01 06:05:34,680 INFO [train.py:903] (3/4) Epoch 5, batch 5500, loss[loss=0.353, simple_loss=0.4024, pruned_loss=0.1518, over 19734.00 frames. ], tot_loss[loss=0.2826, simple_loss=0.3433, pruned_loss=0.111, over 3806154.06 frames. ], batch size: 63, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:05:56,705 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 06:06:05,439 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.564e+02 6.283e+02 7.782e+02 1.000e+03 2.107e+03, threshold=1.556e+03, percent-clipped=4.0 +2023-04-01 06:06:32,678 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7824, 1.7873, 1.7762, 2.8104, 1.7828, 2.6861, 2.4431, 1.7364], + device='cuda:3'), covar=tensor([0.2023, 0.1701, 0.0898, 0.0898, 0.1942, 0.0623, 0.1610, 0.1574], + device='cuda:3'), in_proj_covar=tensor([0.0629, 0.0621, 0.0555, 0.0781, 0.0667, 0.0544, 0.0681, 0.0586], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:06:34,472 INFO [train.py:903] (3/4) Epoch 5, batch 5550, loss[loss=0.3573, simple_loss=0.3925, pruned_loss=0.1611, over 19314.00 frames. ], tot_loss[loss=0.2837, simple_loss=0.3441, pruned_loss=0.1117, over 3806834.54 frames. ], batch size: 66, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:06:40,852 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 06:07:08,641 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2368, 1.3926, 1.9168, 1.4385, 2.5519, 2.1653, 2.7274, 0.9945], + device='cuda:3'), covar=tensor([0.1758, 0.2931, 0.1446, 0.1433, 0.1129, 0.1406, 0.1131, 0.2815], + device='cuda:3'), in_proj_covar=tensor([0.0446, 0.0505, 0.0480, 0.0407, 0.0553, 0.0445, 0.0628, 0.0445], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:07:27,160 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4766, 3.9197, 4.0649, 4.0644, 1.4234, 3.7240, 3.3273, 3.6689], + device='cuda:3'), covar=tensor([0.0975, 0.0613, 0.0579, 0.0470, 0.4152, 0.0431, 0.0574, 0.1105], + device='cuda:3'), in_proj_covar=tensor([0.0511, 0.0449, 0.0600, 0.0496, 0.0575, 0.0365, 0.0388, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 06:07:29,977 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 06:07:36,761 INFO [train.py:903] (3/4) Epoch 5, batch 5600, loss[loss=0.3085, simple_loss=0.361, pruned_loss=0.1279, over 19511.00 frames. ], tot_loss[loss=0.2827, simple_loss=0.3435, pruned_loss=0.111, over 3815734.10 frames. ], batch size: 64, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:08:06,606 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.479e+02 7.188e+02 9.159e+02 1.163e+03 2.158e+03, threshold=1.832e+03, percent-clipped=9.0 +2023-04-01 06:08:31,824 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5915, 1.1857, 1.4699, 1.8993, 1.6205, 2.0270, 2.2576, 2.0529], + device='cuda:3'), covar=tensor([0.0883, 0.1093, 0.1048, 0.0945, 0.0892, 0.0701, 0.0769, 0.0571], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0249, 0.0245, 0.0274, 0.0267, 0.0232, 0.0232, 0.0223], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 06:08:38,297 INFO [train.py:903] (3/4) Epoch 5, batch 5650, loss[loss=0.2826, simple_loss=0.3475, pruned_loss=0.1089, over 19319.00 frames. ], tot_loss[loss=0.2813, simple_loss=0.3423, pruned_loss=0.1102, over 3824000.56 frames. ], batch size: 66, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:09:04,640 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32984.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:09:24,864 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 06:09:38,280 INFO [train.py:903] (3/4) Epoch 5, batch 5700, loss[loss=0.2619, simple_loss=0.3345, pruned_loss=0.09466, over 19743.00 frames. ], tot_loss[loss=0.2826, simple_loss=0.343, pruned_loss=0.1111, over 3832158.72 frames. ], batch size: 63, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:10:09,259 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.491e+02 7.117e+02 8.783e+02 1.086e+03 2.576e+03, threshold=1.757e+03, percent-clipped=4.0 +2023-04-01 06:10:21,039 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2323, 1.1181, 1.3390, 1.3050, 1.8635, 1.8564, 1.8830, 0.4375], + device='cuda:3'), covar=tensor([0.1706, 0.2894, 0.1586, 0.1423, 0.1004, 0.1445, 0.1018, 0.2845], + device='cuda:3'), in_proj_covar=tensor([0.0440, 0.0495, 0.0475, 0.0401, 0.0544, 0.0440, 0.0614, 0.0436], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:10:38,545 INFO [train.py:903] (3/4) Epoch 5, batch 5750, loss[loss=0.2389, simple_loss=0.3095, pruned_loss=0.08414, over 19633.00 frames. ], tot_loss[loss=0.2816, simple_loss=0.342, pruned_loss=0.1106, over 3847257.03 frames. ], batch size: 50, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:10:39,675 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 06:10:47,555 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 06:10:52,568 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 06:11:09,100 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33087.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:11:18,384 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4894, 3.8457, 4.0517, 3.9770, 1.6779, 3.7382, 3.3993, 3.6938], + device='cuda:3'), covar=tensor([0.0840, 0.0572, 0.0506, 0.0498, 0.3438, 0.0380, 0.0445, 0.1013], + device='cuda:3'), in_proj_covar=tensor([0.0514, 0.0448, 0.0602, 0.0500, 0.0577, 0.0366, 0.0388, 0.0566], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 06:11:30,086 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 06:11:40,193 INFO [train.py:903] (3/4) Epoch 5, batch 5800, loss[loss=0.2709, simple_loss=0.3219, pruned_loss=0.1099, over 19743.00 frames. ], tot_loss[loss=0.2815, simple_loss=0.3422, pruned_loss=0.1104, over 3840497.09 frames. ], batch size: 51, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:12:08,910 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.601e+02 7.070e+02 8.520e+02 1.129e+03 2.712e+03, threshold=1.704e+03, percent-clipped=8.0 +2023-04-01 06:12:40,681 INFO [train.py:903] (3/4) Epoch 5, batch 5850, loss[loss=0.2981, simple_loss=0.3627, pruned_loss=0.1168, over 19478.00 frames. ], tot_loss[loss=0.2841, simple_loss=0.344, pruned_loss=0.1122, over 3833954.37 frames. ], batch size: 64, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:12:53,496 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33173.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:13:13,036 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1894, 1.3142, 1.1254, 1.0120, 1.0114, 1.0751, 0.0155, 0.3560], + device='cuda:3'), covar=tensor([0.0315, 0.0303, 0.0175, 0.0222, 0.0587, 0.0245, 0.0489, 0.0463], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0283, 0.0282, 0.0298, 0.0370, 0.0292, 0.0274, 0.0291], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 06:13:40,940 INFO [train.py:903] (3/4) Epoch 5, batch 5900, loss[loss=0.2247, simple_loss=0.2899, pruned_loss=0.07978, over 19410.00 frames. ], tot_loss[loss=0.2844, simple_loss=0.3444, pruned_loss=0.1122, over 3824981.76 frames. ], batch size: 48, lr: 1.57e-02, grad_scale: 8.0 +2023-04-01 06:13:43,338 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 06:14:04,514 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 06:14:11,990 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.345e+02 7.066e+02 8.762e+02 1.130e+03 2.300e+03, threshold=1.752e+03, percent-clipped=6.0 +2023-04-01 06:14:18,954 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3650, 2.0966, 1.7110, 1.3382, 2.0104, 1.0202, 1.1644, 1.6819], + device='cuda:3'), covar=tensor([0.0577, 0.0453, 0.0672, 0.0536, 0.0307, 0.0981, 0.0585, 0.0296], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0263, 0.0307, 0.0236, 0.0216, 0.0305, 0.0283, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:14:41,682 INFO [train.py:903] (3/4) Epoch 5, batch 5950, loss[loss=0.2623, simple_loss=0.3365, pruned_loss=0.09404, over 19763.00 frames. ], tot_loss[loss=0.2823, simple_loss=0.3428, pruned_loss=0.1109, over 3828682.47 frames. ], batch size: 63, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:14:44,252 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 06:15:43,948 INFO [train.py:903] (3/4) Epoch 5, batch 6000, loss[loss=0.2644, simple_loss=0.3285, pruned_loss=0.1002, over 19691.00 frames. ], tot_loss[loss=0.2811, simple_loss=0.3415, pruned_loss=0.1104, over 3830889.25 frames. ], batch size: 53, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:15:43,948 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 06:15:56,873 INFO [train.py:937] (3/4) Epoch 5, validation: loss=0.203, simple_loss=0.3017, pruned_loss=0.05213, over 944034.00 frames. +2023-04-01 06:15:56,874 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 06:16:05,083 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2711, 1.4154, 2.0617, 1.5852, 3.2091, 2.5296, 3.5702, 1.6227], + device='cuda:3'), covar=tensor([0.1981, 0.3321, 0.1801, 0.1421, 0.1220, 0.1538, 0.1387, 0.2716], + device='cuda:3'), in_proj_covar=tensor([0.0448, 0.0507, 0.0484, 0.0410, 0.0560, 0.0447, 0.0623, 0.0442], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:16:18,146 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=33328.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:16:28,853 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.050e+02 6.893e+02 8.503e+02 1.056e+03 1.945e+03, threshold=1.701e+03, percent-clipped=4.0 +2023-04-01 06:16:37,619 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-04-01 06:16:59,411 INFO [train.py:903] (3/4) Epoch 5, batch 6050, loss[loss=0.261, simple_loss=0.3301, pruned_loss=0.09597, over 19662.00 frames. ], tot_loss[loss=0.2812, simple_loss=0.3413, pruned_loss=0.1105, over 3822726.80 frames. ], batch size: 58, lr: 1.56e-02, grad_scale: 16.0 +2023-04-01 06:18:00,558 INFO [train.py:903] (3/4) Epoch 5, batch 6100, loss[loss=0.3357, simple_loss=0.3825, pruned_loss=0.1444, over 19735.00 frames. ], tot_loss[loss=0.2813, simple_loss=0.3411, pruned_loss=0.1108, over 3818443.34 frames. ], batch size: 63, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:18:23,051 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=33431.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:18:32,605 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.702e+02 6.262e+02 7.464e+02 9.853e+02 2.581e+03, threshold=1.493e+03, percent-clipped=2.0 +2023-04-01 06:18:38,640 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-04-01 06:18:39,234 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33443.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:19:00,962 INFO [train.py:903] (3/4) Epoch 5, batch 6150, loss[loss=0.2646, simple_loss=0.3272, pruned_loss=0.101, over 19598.00 frames. ], tot_loss[loss=0.2822, simple_loss=0.3422, pruned_loss=0.1111, over 3825662.66 frames. ], batch size: 52, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:19:29,652 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 06:19:35,840 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33490.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:20:01,185 INFO [train.py:903] (3/4) Epoch 5, batch 6200, loss[loss=0.2861, simple_loss=0.3469, pruned_loss=0.1127, over 19680.00 frames. ], tot_loss[loss=0.2815, simple_loss=0.3413, pruned_loss=0.1108, over 3814995.11 frames. ], batch size: 53, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:20:08,934 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=33517.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:20:34,185 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.213e+02 6.728e+02 8.677e+02 1.165e+03 2.777e+03, threshold=1.735e+03, percent-clipped=13.0 +2023-04-01 06:20:43,514 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33546.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:21:03,397 INFO [train.py:903] (3/4) Epoch 5, batch 6250, loss[loss=0.228, simple_loss=0.2961, pruned_loss=0.07992, over 19771.00 frames. ], tot_loss[loss=0.2791, simple_loss=0.3394, pruned_loss=0.1094, over 3810674.31 frames. ], batch size: 48, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:21:31,254 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 06:22:03,934 INFO [train.py:903] (3/4) Epoch 5, batch 6300, loss[loss=0.2353, simple_loss=0.305, pruned_loss=0.08279, over 19494.00 frames. ], tot_loss[loss=0.2801, simple_loss=0.3402, pruned_loss=0.11, over 3829945.83 frames. ], batch size: 49, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:22:28,415 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33632.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:22:35,562 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.303e+02 6.435e+02 8.030e+02 9.840e+02 2.632e+03, threshold=1.606e+03, percent-clipped=3.0 +2023-04-01 06:22:47,560 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.62 vs. limit=5.0 +2023-04-01 06:23:04,055 INFO [train.py:903] (3/4) Epoch 5, batch 6350, loss[loss=0.263, simple_loss=0.3168, pruned_loss=0.1046, over 19029.00 frames. ], tot_loss[loss=0.281, simple_loss=0.3412, pruned_loss=0.1104, over 3831558.32 frames. ], batch size: 42, lr: 1.56e-02, grad_scale: 8.0 +2023-04-01 06:23:45,912 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33695.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:23:50,609 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33699.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:24:05,038 INFO [train.py:903] (3/4) Epoch 5, batch 6400, loss[loss=0.3198, simple_loss=0.3936, pruned_loss=0.123, over 19691.00 frames. ], tot_loss[loss=0.2819, simple_loss=0.3423, pruned_loss=0.1108, over 3822862.96 frames. ], batch size: 59, lr: 1.55e-02, grad_scale: 8.0 +2023-04-01 06:24:20,335 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33724.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:24:37,367 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.704e+02 6.874e+02 8.420e+02 1.031e+03 3.616e+03, threshold=1.684e+03, percent-clipped=3.0 +2023-04-01 06:25:00,331 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3290, 2.3180, 1.8483, 1.2337, 2.3248, 0.9845, 1.0253, 1.6387], + device='cuda:3'), covar=tensor([0.0759, 0.0512, 0.0640, 0.0678, 0.0370, 0.0992, 0.0689, 0.0414], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0269, 0.0312, 0.0237, 0.0225, 0.0304, 0.0284, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:25:05,847 INFO [train.py:903] (3/4) Epoch 5, batch 6450, loss[loss=0.3407, simple_loss=0.3804, pruned_loss=0.1505, over 19269.00 frames. ], tot_loss[loss=0.2802, simple_loss=0.3409, pruned_loss=0.1098, over 3837562.21 frames. ], batch size: 66, lr: 1.55e-02, grad_scale: 8.0 +2023-04-01 06:25:48,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 06:25:54,908 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33802.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:25:59,532 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.80 vs. limit=5.0 +2023-04-01 06:26:06,361 INFO [train.py:903] (3/4) Epoch 5, batch 6500, loss[loss=0.2379, simple_loss=0.3191, pruned_loss=0.07837, over 19623.00 frames. ], tot_loss[loss=0.2817, simple_loss=0.3422, pruned_loss=0.1106, over 3826899.43 frames. ], batch size: 57, lr: 1.55e-02, grad_scale: 8.0 +2023-04-01 06:26:12,177 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 06:26:24,764 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33827.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:26:32,488 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=33834.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:26:36,708 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.072e+02 6.680e+02 8.171e+02 1.132e+03 2.519e+03, threshold=1.634e+03, percent-clipped=6.0 +2023-04-01 06:27:07,200 INFO [train.py:903] (3/4) Epoch 5, batch 6550, loss[loss=0.2682, simple_loss=0.3423, pruned_loss=0.09703, over 19658.00 frames. ], tot_loss[loss=0.281, simple_loss=0.3417, pruned_loss=0.1102, over 3817206.34 frames. ], batch size: 59, lr: 1.55e-02, grad_scale: 8.0 +2023-04-01 06:27:38,517 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33888.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:28:06,990 INFO [train.py:903] (3/4) Epoch 5, batch 6600, loss[loss=0.2672, simple_loss=0.3355, pruned_loss=0.09948, over 19780.00 frames. ], tot_loss[loss=0.2809, simple_loss=0.3418, pruned_loss=0.11, over 3803857.44 frames. ], batch size: 56, lr: 1.55e-02, grad_scale: 4.0 +2023-04-01 06:28:08,467 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33913.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:28:40,230 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.407e+02 7.502e+02 9.137e+02 1.060e+03 2.817e+03, threshold=1.827e+03, percent-clipped=6.0 +2023-04-01 06:28:52,899 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33949.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:29:01,860 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33957.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:29:09,038 INFO [train.py:903] (3/4) Epoch 5, batch 6650, loss[loss=0.2906, simple_loss=0.3667, pruned_loss=0.1073, over 19739.00 frames. ], tot_loss[loss=0.2787, simple_loss=0.3399, pruned_loss=0.1088, over 3817839.66 frames. ], batch size: 63, lr: 1.55e-02, grad_scale: 4.0 +2023-04-01 06:30:10,061 INFO [train.py:903] (3/4) Epoch 5, batch 6700, loss[loss=0.2281, simple_loss=0.2927, pruned_loss=0.08178, over 19797.00 frames. ], tot_loss[loss=0.2814, simple_loss=0.3417, pruned_loss=0.1106, over 3814760.08 frames. ], batch size: 48, lr: 1.55e-02, grad_scale: 4.0 +2023-04-01 06:30:25,148 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 06:30:40,326 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.004e+02 7.257e+02 9.131e+02 1.100e+03 2.314e+03, threshold=1.826e+03, percent-clipped=7.0 +2023-04-01 06:30:40,478 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34039.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:30:42,851 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.49 vs. limit=5.0 +2023-04-01 06:31:06,262 INFO [train.py:903] (3/4) Epoch 5, batch 6750, loss[loss=0.3109, simple_loss=0.3668, pruned_loss=0.1275, over 19533.00 frames. ], tot_loss[loss=0.2812, simple_loss=0.3414, pruned_loss=0.1105, over 3824222.64 frames. ], batch size: 56, lr: 1.55e-02, grad_scale: 4.0 +2023-04-01 06:31:06,532 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34062.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:31:22,477 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4369, 2.3808, 1.6823, 1.5563, 2.3081, 1.0878, 1.1680, 1.6649], + device='cuda:3'), covar=tensor([0.0709, 0.0405, 0.0735, 0.0531, 0.0303, 0.0941, 0.0628, 0.0430], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0269, 0.0312, 0.0240, 0.0223, 0.0304, 0.0287, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:32:02,677 INFO [train.py:903] (3/4) Epoch 5, batch 6800, loss[loss=0.3152, simple_loss=0.3632, pruned_loss=0.1335, over 19614.00 frames. ], tot_loss[loss=0.2807, simple_loss=0.3409, pruned_loss=0.1103, over 3817476.03 frames. ], batch size: 50, lr: 1.55e-02, grad_scale: 8.0 +2023-04-01 06:32:30,647 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.271e+02 6.317e+02 7.808e+02 9.230e+02 1.582e+03, threshold=1.562e+03, percent-clipped=0.0 +2023-04-01 06:32:47,562 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 06:32:48,597 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 06:32:50,734 INFO [train.py:903] (3/4) Epoch 6, batch 0, loss[loss=0.3268, simple_loss=0.3786, pruned_loss=0.1375, over 19659.00 frames. ], tot_loss[loss=0.3268, simple_loss=0.3786, pruned_loss=0.1375, over 19659.00 frames. ], batch size: 58, lr: 1.44e-02, grad_scale: 8.0 +2023-04-01 06:32:50,734 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 06:33:02,090 INFO [train.py:937] (3/4) Epoch 6, validation: loss=0.2022, simple_loss=0.3015, pruned_loss=0.05149, over 944034.00 frames. +2023-04-01 06:33:02,091 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 06:33:15,298 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 06:33:20,354 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34154.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:33:30,313 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34163.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:34:03,531 INFO [train.py:903] (3/4) Epoch 6, batch 50, loss[loss=0.268, simple_loss=0.3273, pruned_loss=0.1044, over 19458.00 frames. ], tot_loss[loss=0.2777, simple_loss=0.3401, pruned_loss=0.1077, over 875423.28 frames. ], batch size: 49, lr: 1.44e-02, grad_scale: 8.0 +2023-04-01 06:34:15,334 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 06:34:22,176 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34205.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:34:29,751 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 06:34:40,712 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 06:34:54,295 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34230.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:35:05,215 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.721e+02 5.756e+02 7.149e+02 1.025e+03 3.166e+03, threshold=1.430e+03, percent-clipped=7.0 +2023-04-01 06:35:06,288 INFO [train.py:903] (3/4) Epoch 6, batch 100, loss[loss=0.2804, simple_loss=0.3517, pruned_loss=0.1045, over 19666.00 frames. ], tot_loss[loss=0.2726, simple_loss=0.3358, pruned_loss=0.1047, over 1542797.87 frames. ], batch size: 58, lr: 1.44e-02, grad_scale: 8.0 +2023-04-01 06:35:18,583 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 06:35:20,113 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4823, 1.1202, 1.1822, 1.3395, 2.1405, 0.9924, 1.6863, 2.2018], + device='cuda:3'), covar=tensor([0.0561, 0.2340, 0.2403, 0.1305, 0.0727, 0.1916, 0.0989, 0.0504], + device='cuda:3'), in_proj_covar=tensor([0.0286, 0.0311, 0.0313, 0.0286, 0.0307, 0.0316, 0.0284, 0.0301], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:36:06,449 INFO [train.py:903] (3/4) Epoch 6, batch 150, loss[loss=0.2439, simple_loss=0.3195, pruned_loss=0.0842, over 19656.00 frames. ], tot_loss[loss=0.2751, simple_loss=0.3377, pruned_loss=0.1062, over 2059303.60 frames. ], batch size: 55, lr: 1.44e-02, grad_scale: 4.0 +2023-04-01 06:36:19,472 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34301.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:37:08,895 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.164e+02 6.478e+02 8.283e+02 9.993e+02 1.951e+03, threshold=1.657e+03, percent-clipped=7.0 +2023-04-01 06:37:08,914 INFO [train.py:903] (3/4) Epoch 6, batch 200, loss[loss=0.3262, simple_loss=0.3925, pruned_loss=0.1299, over 19357.00 frames. ], tot_loss[loss=0.2753, simple_loss=0.3378, pruned_loss=0.1064, over 2444450.95 frames. ], batch size: 70, lr: 1.44e-02, grad_scale: 4.0 +2023-04-01 06:37:08,921 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 06:38:12,100 INFO [train.py:903] (3/4) Epoch 6, batch 250, loss[loss=0.2733, simple_loss=0.3451, pruned_loss=0.1008, over 19302.00 frames. ], tot_loss[loss=0.2762, simple_loss=0.3383, pruned_loss=0.107, over 2760346.10 frames. ], batch size: 66, lr: 1.44e-02, grad_scale: 4.0 +2023-04-01 06:38:33,065 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34406.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:38:37,984 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34410.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:38:44,900 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34416.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:39:08,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34435.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:39:14,113 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.775e+02 6.949e+02 8.663e+02 1.115e+03 2.860e+03, threshold=1.733e+03, percent-clipped=3.0 +2023-04-01 06:39:14,131 INFO [train.py:903] (3/4) Epoch 6, batch 300, loss[loss=0.2286, simple_loss=0.2956, pruned_loss=0.08084, over 19362.00 frames. ], tot_loss[loss=0.2754, simple_loss=0.3379, pruned_loss=0.1065, over 3004723.70 frames. ], batch size: 47, lr: 1.44e-02, grad_scale: 4.0 +2023-04-01 06:40:17,202 INFO [train.py:903] (3/4) Epoch 6, batch 350, loss[loss=0.2715, simple_loss=0.345, pruned_loss=0.09906, over 19527.00 frames. ], tot_loss[loss=0.2745, simple_loss=0.3375, pruned_loss=0.1058, over 3191399.74 frames. ], batch size: 56, lr: 1.43e-02, grad_scale: 4.0 +2023-04-01 06:40:22,957 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 06:40:37,843 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34507.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:40:45,481 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-01 06:40:55,534 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34521.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:41:02,905 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-01 06:41:18,579 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.691e+02 6.755e+02 8.258e+02 9.979e+02 1.871e+03, threshold=1.652e+03, percent-clipped=1.0 +2023-04-01 06:41:18,598 INFO [train.py:903] (3/4) Epoch 6, batch 400, loss[loss=0.2746, simple_loss=0.3405, pruned_loss=0.1043, over 19610.00 frames. ], tot_loss[loss=0.2739, simple_loss=0.3366, pruned_loss=0.1056, over 3324348.71 frames. ], batch size: 61, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:41:51,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9912, 1.4906, 1.5877, 2.1289, 1.9167, 1.7996, 1.7657, 1.8696], + device='cuda:3'), covar=tensor([0.0773, 0.1471, 0.1275, 0.0737, 0.1041, 0.0441, 0.0881, 0.0586], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0362, 0.0287, 0.0235, 0.0301, 0.0245, 0.0272, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:42:20,648 INFO [train.py:903] (3/4) Epoch 6, batch 450, loss[loss=0.2485, simple_loss=0.3194, pruned_loss=0.08874, over 19716.00 frames. ], tot_loss[loss=0.2766, simple_loss=0.3388, pruned_loss=0.1072, over 3438131.61 frames. ], batch size: 51, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:42:49,006 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34611.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:42:54,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 06:42:55,454 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 06:43:01,473 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34622.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:43:23,826 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.432e+02 6.970e+02 8.269e+02 1.071e+03 2.551e+03, threshold=1.654e+03, percent-clipped=6.0 +2023-04-01 06:43:23,844 INFO [train.py:903] (3/4) Epoch 6, batch 500, loss[loss=0.3231, simple_loss=0.38, pruned_loss=0.1331, over 19793.00 frames. ], tot_loss[loss=0.2762, simple_loss=0.3383, pruned_loss=0.107, over 3532393.60 frames. ], batch size: 56, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:43:25,159 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34640.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:43:29,871 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34644.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:43:29,995 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0401, 3.6415, 1.9826, 2.1354, 3.2075, 1.6008, 1.2716, 1.9841], + device='cuda:3'), covar=tensor([0.0815, 0.0305, 0.0770, 0.0531, 0.0299, 0.0852, 0.0756, 0.0482], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0271, 0.0314, 0.0238, 0.0227, 0.0305, 0.0286, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:43:35,808 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9147, 4.3647, 4.6102, 4.4859, 1.6288, 4.1147, 3.7032, 4.1918], + device='cuda:3'), covar=tensor([0.0916, 0.0497, 0.0390, 0.0431, 0.4086, 0.0356, 0.0464, 0.0872], + device='cuda:3'), in_proj_covar=tensor([0.0532, 0.0464, 0.0610, 0.0510, 0.0598, 0.0381, 0.0390, 0.0576], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 06:44:03,447 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34672.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:44:27,125 INFO [train.py:903] (3/4) Epoch 6, batch 550, loss[loss=0.2121, simple_loss=0.2793, pruned_loss=0.07244, over 19765.00 frames. ], tot_loss[loss=0.275, simple_loss=0.3374, pruned_loss=0.1063, over 3589986.27 frames. ], batch size: 46, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:44:37,142 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34697.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:44:42,301 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-01 06:45:17,137 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34728.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 06:45:31,818 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.620e+02 6.191e+02 8.093e+02 9.820e+02 1.880e+03, threshold=1.619e+03, percent-clipped=2.0 +2023-04-01 06:45:31,837 INFO [train.py:903] (3/4) Epoch 6, batch 600, loss[loss=0.2865, simple_loss=0.343, pruned_loss=0.115, over 19776.00 frames. ], tot_loss[loss=0.2746, simple_loss=0.3373, pruned_loss=0.106, over 3646298.08 frames. ], batch size: 54, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:46:13,306 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 06:46:20,296 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34777.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:46:35,493 INFO [train.py:903] (3/4) Epoch 6, batch 650, loss[loss=0.228, simple_loss=0.2933, pruned_loss=0.08135, over 19613.00 frames. ], tot_loss[loss=0.2754, simple_loss=0.3376, pruned_loss=0.1066, over 3698366.41 frames. ], batch size: 50, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:46:51,809 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34802.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:47:38,641 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.675e+02 6.280e+02 8.587e+02 1.153e+03 3.497e+03, threshold=1.717e+03, percent-clipped=9.0 +2023-04-01 06:47:38,661 INFO [train.py:903] (3/4) Epoch 6, batch 700, loss[loss=0.3804, simple_loss=0.4038, pruned_loss=0.1786, over 13354.00 frames. ], tot_loss[loss=0.2753, simple_loss=0.3376, pruned_loss=0.1066, over 3716810.34 frames. ], batch size: 136, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:47:51,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.7746, 5.1343, 2.5779, 4.5140, 1.1512, 4.8040, 4.9415, 5.1986], + device='cuda:3'), covar=tensor([0.0453, 0.0900, 0.2149, 0.0597, 0.4111, 0.0617, 0.0628, 0.0797], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0315, 0.0376, 0.0290, 0.0360, 0.0310, 0.0289, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 06:47:54,283 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3956, 2.9990, 2.0706, 2.0659, 2.2826, 2.3928, 0.7668, 2.1805], + device='cuda:3'), covar=tensor([0.0249, 0.0239, 0.0327, 0.0437, 0.0402, 0.0443, 0.0614, 0.0476], + device='cuda:3'), in_proj_covar=tensor([0.0284, 0.0288, 0.0292, 0.0306, 0.0374, 0.0295, 0.0286, 0.0296], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 06:48:27,772 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34878.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:48:43,534 INFO [train.py:903] (3/4) Epoch 6, batch 750, loss[loss=0.3174, simple_loss=0.371, pruned_loss=0.1319, over 19794.00 frames. ], tot_loss[loss=0.2775, simple_loss=0.3399, pruned_loss=0.1076, over 3750829.71 frames. ], batch size: 63, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:48:47,836 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9677, 1.9979, 1.9447, 2.8705, 1.9668, 2.6859, 2.5804, 1.8458], + device='cuda:3'), covar=tensor([0.2150, 0.1651, 0.0861, 0.0995, 0.1971, 0.0697, 0.1622, 0.1556], + device='cuda:3'), in_proj_covar=tensor([0.0652, 0.0638, 0.0570, 0.0800, 0.0681, 0.0557, 0.0701, 0.0605], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 06:49:00,092 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34903.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:49:45,088 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.784e+02 6.254e+02 7.861e+02 1.094e+03 2.828e+03, threshold=1.572e+03, percent-clipped=5.0 +2023-04-01 06:49:45,108 INFO [train.py:903] (3/4) Epoch 6, batch 800, loss[loss=0.3084, simple_loss=0.3674, pruned_loss=0.1247, over 19340.00 frames. ], tot_loss[loss=0.2765, simple_loss=0.3385, pruned_loss=0.1072, over 3767472.34 frames. ], batch size: 66, lr: 1.43e-02, grad_scale: 8.0 +2023-04-01 06:50:02,467 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 06:50:03,738 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34955.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:50:37,442 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2016, 1.3615, 1.3818, 1.7686, 2.7546, 1.0784, 2.0899, 3.0333], + device='cuda:3'), covar=tensor([0.0437, 0.2441, 0.2286, 0.1238, 0.0641, 0.2148, 0.1064, 0.0392], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0317, 0.0317, 0.0287, 0.0313, 0.0315, 0.0289, 0.0307], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:50:41,648 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34984.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:50:46,241 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=34988.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:50:48,468 INFO [train.py:903] (3/4) Epoch 6, batch 850, loss[loss=0.3071, simple_loss=0.3694, pruned_loss=0.1224, over 19649.00 frames. ], tot_loss[loss=0.2761, simple_loss=0.3382, pruned_loss=0.107, over 3775821.32 frames. ], batch size: 58, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:50:57,232 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-01 06:51:37,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2197, 2.8094, 2.1261, 2.1579, 2.1115, 2.4022, 0.5696, 1.9972], + device='cuda:3'), covar=tensor([0.0292, 0.0289, 0.0317, 0.0425, 0.0467, 0.0402, 0.0618, 0.0532], + device='cuda:3'), in_proj_covar=tensor([0.0288, 0.0289, 0.0289, 0.0308, 0.0376, 0.0295, 0.0285, 0.0298], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 06:51:42,524 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 06:51:49,537 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.027e+02 6.223e+02 7.935e+02 9.772e+02 2.166e+03, threshold=1.587e+03, percent-clipped=2.0 +2023-04-01 06:51:49,557 INFO [train.py:903] (3/4) Epoch 6, batch 900, loss[loss=0.2873, simple_loss=0.3641, pruned_loss=0.1052, over 19310.00 frames. ], tot_loss[loss=0.2756, simple_loss=0.3378, pruned_loss=0.1067, over 3800593.25 frames. ], batch size: 66, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:52:28,250 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35070.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:52:30,300 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=35072.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 06:52:51,734 INFO [train.py:903] (3/4) Epoch 6, batch 950, loss[loss=0.3078, simple_loss=0.363, pruned_loss=0.1263, over 19686.00 frames. ], tot_loss[loss=0.276, simple_loss=0.3382, pruned_loss=0.1069, over 3812557.40 frames. ], batch size: 59, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:52:57,573 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 06:53:04,407 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35099.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:53:09,177 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35103.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:53:18,122 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35110.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:53:37,153 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0178, 1.4386, 1.4741, 1.5107, 2.6819, 0.9981, 1.8563, 2.7871], + device='cuda:3'), covar=tensor([0.0340, 0.1742, 0.1748, 0.1115, 0.0505, 0.1866, 0.0954, 0.0314], + device='cuda:3'), in_proj_covar=tensor([0.0289, 0.0311, 0.0311, 0.0283, 0.0307, 0.0314, 0.0286, 0.0304], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 06:53:55,220 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.390e+02 7.194e+02 8.589e+02 1.083e+03 2.096e+03, threshold=1.718e+03, percent-clipped=5.0 +2023-04-01 06:53:55,239 INFO [train.py:903] (3/4) Epoch 6, batch 1000, loss[loss=0.2737, simple_loss=0.3293, pruned_loss=0.109, over 19600.00 frames. ], tot_loss[loss=0.2767, simple_loss=0.3384, pruned_loss=0.1075, over 3799841.41 frames. ], batch size: 50, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:54:48,383 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 06:54:52,216 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35187.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 06:54:55,181 INFO [train.py:903] (3/4) Epoch 6, batch 1050, loss[loss=0.2506, simple_loss=0.3198, pruned_loss=0.0907, over 19670.00 frames. ], tot_loss[loss=0.2769, simple_loss=0.3385, pruned_loss=0.1077, over 3811402.15 frames. ], batch size: 53, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:55:30,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 06:55:57,171 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.102e+02 7.066e+02 8.619e+02 1.238e+03 3.302e+03, threshold=1.724e+03, percent-clipped=8.0 +2023-04-01 06:55:57,191 INFO [train.py:903] (3/4) Epoch 6, batch 1100, loss[loss=0.2732, simple_loss=0.3401, pruned_loss=0.1032, over 19409.00 frames. ], tot_loss[loss=0.2783, simple_loss=0.3395, pruned_loss=0.1085, over 3812466.36 frames. ], batch size: 70, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:56:59,591 INFO [train.py:903] (3/4) Epoch 6, batch 1150, loss[loss=0.2848, simple_loss=0.3529, pruned_loss=0.1084, over 19301.00 frames. ], tot_loss[loss=0.2774, simple_loss=0.3386, pruned_loss=0.1081, over 3805150.21 frames. ], batch size: 66, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:57:45,502 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35326.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:58:04,200 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.615e+02 6.124e+02 7.469e+02 9.050e+02 1.884e+03, threshold=1.494e+03, percent-clipped=1.0 +2023-04-01 06:58:04,219 INFO [train.py:903] (3/4) Epoch 6, batch 1200, loss[loss=0.2611, simple_loss=0.3253, pruned_loss=0.09845, over 19847.00 frames. ], tot_loss[loss=0.2779, simple_loss=0.3389, pruned_loss=0.1084, over 3804440.39 frames. ], batch size: 52, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:58:17,345 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35351.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:58:22,860 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9239, 3.5308, 2.3629, 3.2695, 1.0032, 3.2220, 3.2437, 3.2960], + device='cuda:3'), covar=tensor([0.0777, 0.1087, 0.1908, 0.0726, 0.3519, 0.0915, 0.0782, 0.0940], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0318, 0.0380, 0.0294, 0.0357, 0.0313, 0.0294, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 06:58:23,077 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35355.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:58:27,684 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35359.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:58:32,232 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35363.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 06:58:36,592 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 06:58:55,126 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35380.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:58:59,609 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35384.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:59:05,735 INFO [train.py:903] (3/4) Epoch 6, batch 1250, loss[loss=0.3282, simple_loss=0.3754, pruned_loss=0.1405, over 19765.00 frames. ], tot_loss[loss=0.2773, simple_loss=0.3383, pruned_loss=0.1081, over 3821542.63 frames. ], batch size: 63, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 06:59:09,344 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35392.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 06:59:26,319 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-01 07:00:08,110 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.399e+02 6.685e+02 8.523e+02 1.077e+03 2.432e+03, threshold=1.705e+03, percent-clipped=5.0 +2023-04-01 07:00:08,129 INFO [train.py:903] (3/4) Epoch 6, batch 1300, loss[loss=0.2525, simple_loss=0.331, pruned_loss=0.08701, over 19539.00 frames. ], tot_loss[loss=0.2766, simple_loss=0.3377, pruned_loss=0.1077, over 3815650.20 frames. ], batch size: 56, lr: 1.42e-02, grad_scale: 8.0 +2023-04-01 07:00:12,118 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35443.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:00:26,240 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=35454.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:00:43,939 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35468.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:01:11,034 INFO [train.py:903] (3/4) Epoch 6, batch 1350, loss[loss=0.2755, simple_loss=0.3335, pruned_loss=0.1087, over 19746.00 frames. ], tot_loss[loss=0.2762, simple_loss=0.3375, pruned_loss=0.1074, over 3816817.85 frames. ], batch size: 46, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:01:26,078 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5559, 1.0798, 1.4250, 1.5718, 2.8829, 1.1193, 2.1312, 3.0979], + device='cuda:3'), covar=tensor([0.0475, 0.3122, 0.2770, 0.1720, 0.0867, 0.2499, 0.1273, 0.0554], + device='cuda:3'), in_proj_covar=tensor([0.0291, 0.0313, 0.0314, 0.0284, 0.0309, 0.0312, 0.0284, 0.0305], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:01:30,441 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9457, 2.1627, 2.2800, 2.2580, 1.0555, 2.0962, 1.9058, 2.0648], + device='cuda:3'), covar=tensor([0.0917, 0.1254, 0.0562, 0.0542, 0.2684, 0.0625, 0.0494, 0.0901], + device='cuda:3'), in_proj_covar=tensor([0.0532, 0.0463, 0.0618, 0.0507, 0.0592, 0.0378, 0.0396, 0.0578], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 07:01:30,555 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9970, 0.7932, 0.7288, 0.9539, 0.8166, 0.8967, 0.7020, 0.8853], + device='cuda:3'), covar=tensor([0.0635, 0.0909, 0.0994, 0.0556, 0.0731, 0.0372, 0.0810, 0.0491], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0359, 0.0285, 0.0233, 0.0300, 0.0239, 0.0267, 0.0229], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:02:13,028 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.489e+02 6.783e+02 8.495e+02 1.071e+03 2.340e+03, threshold=1.699e+03, percent-clipped=3.0 +2023-04-01 07:02:13,047 INFO [train.py:903] (3/4) Epoch 6, batch 1400, loss[loss=0.2533, simple_loss=0.31, pruned_loss=0.09828, over 19763.00 frames. ], tot_loss[loss=0.2765, simple_loss=0.3376, pruned_loss=0.1077, over 3805915.30 frames. ], batch size: 47, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:02:22,780 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35548.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:02:47,194 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35569.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:03:13,362 INFO [train.py:903] (3/4) Epoch 6, batch 1450, loss[loss=0.2926, simple_loss=0.361, pruned_loss=0.1121, over 19777.00 frames. ], tot_loss[loss=0.2774, simple_loss=0.3385, pruned_loss=0.1082, over 3811814.91 frames. ], batch size: 56, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:03:13,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 07:04:15,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.374e+02 6.609e+02 8.520e+02 1.101e+03 2.891e+03, threshold=1.704e+03, percent-clipped=3.0 +2023-04-01 07:04:15,942 INFO [train.py:903] (3/4) Epoch 6, batch 1500, loss[loss=0.3247, simple_loss=0.3715, pruned_loss=0.139, over 19757.00 frames. ], tot_loss[loss=0.2769, simple_loss=0.3383, pruned_loss=0.1078, over 3817796.33 frames. ], batch size: 63, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:04:24,661 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0492, 2.7039, 1.7109, 1.9005, 1.7399, 2.2820, 0.5529, 1.8352], + device='cuda:3'), covar=tensor([0.0281, 0.0321, 0.0349, 0.0501, 0.0562, 0.0469, 0.0659, 0.0592], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0296, 0.0292, 0.0312, 0.0381, 0.0299, 0.0286, 0.0299], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:04:43,320 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3343, 1.4251, 1.4007, 1.4368, 2.9179, 1.0231, 1.9714, 3.0008], + device='cuda:3'), covar=tensor([0.0374, 0.2239, 0.2301, 0.1479, 0.0568, 0.2109, 0.1156, 0.0407], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0313, 0.0316, 0.0285, 0.0310, 0.0311, 0.0286, 0.0304], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:05:17,134 INFO [train.py:903] (3/4) Epoch 6, batch 1550, loss[loss=0.3385, simple_loss=0.3742, pruned_loss=0.1514, over 19764.00 frames. ], tot_loss[loss=0.2797, simple_loss=0.3403, pruned_loss=0.1096, over 3802571.45 frames. ], batch size: 47, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:05:18,540 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0313, 1.8585, 1.4456, 1.1873, 1.7408, 1.0185, 0.9853, 1.5561], + device='cuda:3'), covar=tensor([0.0584, 0.0481, 0.0728, 0.0518, 0.0303, 0.0897, 0.0545, 0.0306], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0273, 0.0314, 0.0239, 0.0227, 0.0308, 0.0284, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:05:39,897 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=35707.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:06:02,414 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5953, 1.2935, 1.3119, 2.0960, 1.6544, 2.0571, 2.1472, 1.8242], + device='cuda:3'), covar=tensor([0.0842, 0.1071, 0.1107, 0.0849, 0.0962, 0.0674, 0.0741, 0.0686], + device='cuda:3'), in_proj_covar=tensor([0.0229, 0.0248, 0.0240, 0.0271, 0.0264, 0.0232, 0.0224, 0.0219], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 07:06:09,043 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9139, 2.0619, 1.8269, 1.9118, 4.4840, 0.9925, 2.1473, 4.1028], + device='cuda:3'), covar=tensor([0.0270, 0.2148, 0.2283, 0.1436, 0.0469, 0.2387, 0.1298, 0.0360], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0312, 0.0315, 0.0286, 0.0309, 0.0311, 0.0285, 0.0304], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:06:16,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=35736.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:06:22,435 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.104e+02 6.450e+02 9.026e+02 1.093e+03 2.835e+03, threshold=1.805e+03, percent-clipped=5.0 +2023-04-01 07:06:22,459 INFO [train.py:903] (3/4) Epoch 6, batch 1600, loss[loss=0.2441, simple_loss=0.3114, pruned_loss=0.08845, over 19480.00 frames. ], tot_loss[loss=0.2783, simple_loss=0.3394, pruned_loss=0.1086, over 3820895.05 frames. ], batch size: 49, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:06:44,235 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 07:07:23,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35789.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:07:24,378 INFO [train.py:903] (3/4) Epoch 6, batch 1650, loss[loss=0.2718, simple_loss=0.3435, pruned_loss=0.1, over 19690.00 frames. ], tot_loss[loss=0.2775, simple_loss=0.3388, pruned_loss=0.1081, over 3815638.16 frames. ], batch size: 59, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:08:05,730 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35822.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:08:09,320 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35825.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:08:13,941 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35829.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:08:23,387 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7102, 1.7636, 1.5318, 1.3144, 1.2335, 1.5045, 0.1258, 0.6329], + device='cuda:3'), covar=tensor([0.0280, 0.0280, 0.0159, 0.0254, 0.0638, 0.0251, 0.0490, 0.0478], + device='cuda:3'), in_proj_covar=tensor([0.0290, 0.0297, 0.0291, 0.0311, 0.0382, 0.0300, 0.0286, 0.0301], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:08:27,429 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.226e+02 6.593e+02 7.720e+02 9.793e+02 2.227e+03, threshold=1.544e+03, percent-clipped=1.0 +2023-04-01 07:08:27,448 INFO [train.py:903] (3/4) Epoch 6, batch 1700, loss[loss=0.304, simple_loss=0.3532, pruned_loss=0.1273, over 19523.00 frames. ], tot_loss[loss=0.2762, simple_loss=0.3381, pruned_loss=0.1072, over 3836220.09 frames. ], batch size: 54, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:08:38,917 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35850.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:08:40,015 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35851.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:08:54,021 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 07:09:06,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 07:09:18,234 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0625, 2.8448, 1.9417, 2.0086, 1.7984, 2.3504, 0.7551, 1.9287], + device='cuda:3'), covar=tensor([0.0251, 0.0255, 0.0263, 0.0387, 0.0508, 0.0402, 0.0564, 0.0482], + device='cuda:3'), in_proj_covar=tensor([0.0288, 0.0295, 0.0290, 0.0309, 0.0379, 0.0299, 0.0282, 0.0298], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:09:29,413 INFO [train.py:903] (3/4) Epoch 6, batch 1750, loss[loss=0.2682, simple_loss=0.3295, pruned_loss=0.1034, over 19404.00 frames. ], tot_loss[loss=0.2746, simple_loss=0.337, pruned_loss=0.1061, over 3837239.55 frames. ], batch size: 48, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:09:31,982 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=35892.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:09:56,884 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.48 vs. limit=5.0 +2023-04-01 07:10:33,851 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.141e+02 6.695e+02 8.372e+02 1.116e+03 2.634e+03, threshold=1.674e+03, percent-clipped=7.0 +2023-04-01 07:10:33,872 INFO [train.py:903] (3/4) Epoch 6, batch 1800, loss[loss=0.2619, simple_loss=0.3359, pruned_loss=0.09395, over 19774.00 frames. ], tot_loss[loss=0.2748, simple_loss=0.3373, pruned_loss=0.1062, over 3830313.27 frames. ], batch size: 54, lr: 1.41e-02, grad_scale: 8.0 +2023-04-01 07:10:36,673 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9325, 1.2645, 0.9620, 0.9630, 1.1217, 0.9155, 0.8380, 1.2357], + device='cuda:3'), covar=tensor([0.0391, 0.0505, 0.0843, 0.0401, 0.0336, 0.0790, 0.0445, 0.0307], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0276, 0.0315, 0.0241, 0.0228, 0.0307, 0.0281, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:11:11,642 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8619, 1.9415, 1.8691, 2.8139, 1.8381, 2.6619, 2.6436, 1.8522], + device='cuda:3'), covar=tensor([0.2552, 0.1907, 0.0990, 0.1126, 0.2275, 0.0805, 0.1823, 0.1812], + device='cuda:3'), in_proj_covar=tensor([0.0663, 0.0647, 0.0579, 0.0805, 0.0685, 0.0572, 0.0701, 0.0610], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 07:11:31,916 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 07:11:36,739 INFO [train.py:903] (3/4) Epoch 6, batch 1850, loss[loss=0.2637, simple_loss=0.3114, pruned_loss=0.108, over 19744.00 frames. ], tot_loss[loss=0.2741, simple_loss=0.3368, pruned_loss=0.1057, over 3831835.02 frames. ], batch size: 45, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:11:59,170 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36007.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:12:11,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 07:12:40,877 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.024e+02 7.330e+02 9.053e+02 1.086e+03 1.723e+03, threshold=1.811e+03, percent-clipped=2.0 +2023-04-01 07:12:40,895 INFO [train.py:903] (3/4) Epoch 6, batch 1900, loss[loss=0.2465, simple_loss=0.3122, pruned_loss=0.09035, over 19340.00 frames. ], tot_loss[loss=0.2747, simple_loss=0.3374, pruned_loss=0.106, over 3829745.66 frames. ], batch size: 47, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:12:57,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 07:13:04,075 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 07:13:26,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5574, 2.2594, 1.5926, 1.6001, 2.0107, 1.3317, 1.3359, 1.7704], + device='cuda:3'), covar=tensor([0.0713, 0.0533, 0.0843, 0.0536, 0.0423, 0.0882, 0.0561, 0.0409], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0278, 0.0311, 0.0240, 0.0230, 0.0308, 0.0284, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:13:27,614 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 07:13:29,155 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36078.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:13:42,462 INFO [train.py:903] (3/4) Epoch 6, batch 1950, loss[loss=0.2411, simple_loss=0.2991, pruned_loss=0.09152, over 19376.00 frames. ], tot_loss[loss=0.2748, simple_loss=0.3373, pruned_loss=0.1062, over 3826636.99 frames. ], batch size: 47, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:14:00,226 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36103.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:14:04,978 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36107.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:14:35,250 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36132.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:14:36,151 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36133.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:14:36,381 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8571, 1.4476, 1.4808, 1.7784, 1.6140, 1.6454, 1.5225, 1.5975], + device='cuda:3'), covar=tensor([0.0772, 0.1353, 0.1213, 0.0801, 0.1068, 0.0412, 0.0904, 0.0620], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0364, 0.0287, 0.0238, 0.0307, 0.0245, 0.0272, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:14:45,005 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.894e+02 6.352e+02 7.868e+02 9.806e+02 1.510e+03, threshold=1.574e+03, percent-clipped=0.0 +2023-04-01 07:14:45,023 INFO [train.py:903] (3/4) Epoch 6, batch 2000, loss[loss=0.2086, simple_loss=0.2753, pruned_loss=0.07094, over 19309.00 frames. ], tot_loss[loss=0.2743, simple_loss=0.3368, pruned_loss=0.1058, over 3825034.53 frames. ], batch size: 44, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:15:24,944 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36173.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:15:30,618 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8630, 1.9210, 1.7572, 2.8038, 1.9005, 2.6130, 2.4262, 1.7406], + device='cuda:3'), covar=tensor([0.2435, 0.1899, 0.1070, 0.1064, 0.2191, 0.0784, 0.1968, 0.1902], + device='cuda:3'), in_proj_covar=tensor([0.0663, 0.0650, 0.0578, 0.0808, 0.0685, 0.0571, 0.0702, 0.0609], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 07:15:42,599 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 07:15:42,939 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5852, 1.4240, 1.3489, 2.0139, 1.7925, 2.0971, 2.1329, 1.7910], + device='cuda:3'), covar=tensor([0.0750, 0.0892, 0.1010, 0.0888, 0.0815, 0.0610, 0.0842, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0226, 0.0244, 0.0236, 0.0272, 0.0262, 0.0225, 0.0224, 0.0216], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 07:15:46,070 INFO [train.py:903] (3/4) Epoch 6, batch 2050, loss[loss=0.2726, simple_loss=0.3341, pruned_loss=0.1056, over 19764.00 frames. ], tot_loss[loss=0.2726, simple_loss=0.3357, pruned_loss=0.1047, over 3834879.26 frames. ], batch size: 56, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:15:57,095 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36199.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:16:00,523 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 07:16:03,078 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 07:16:22,976 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 07:16:47,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.109e+02 6.690e+02 8.798e+02 1.173e+03 2.442e+03, threshold=1.760e+03, percent-clipped=12.0 +2023-04-01 07:16:47,793 INFO [train.py:903] (3/4) Epoch 6, batch 2100, loss[loss=0.2472, simple_loss=0.3286, pruned_loss=0.08287, over 19620.00 frames. ], tot_loss[loss=0.2724, simple_loss=0.3357, pruned_loss=0.1046, over 3836246.76 frames. ], batch size: 57, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:16:53,764 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9865, 1.7741, 1.5672, 1.9502, 2.0046, 1.7500, 1.6551, 1.8656], + device='cuda:3'), covar=tensor([0.0719, 0.1292, 0.1213, 0.0740, 0.0896, 0.0453, 0.0925, 0.0588], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0363, 0.0287, 0.0237, 0.0304, 0.0244, 0.0270, 0.0231], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:16:57,879 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36248.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:17:17,867 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36263.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:17:18,600 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 07:17:39,982 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 07:17:49,931 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36288.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:17:49,971 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36288.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:17:51,847 INFO [train.py:903] (3/4) Epoch 6, batch 2150, loss[loss=0.301, simple_loss=0.3663, pruned_loss=0.1179, over 19778.00 frames. ], tot_loss[loss=0.2725, simple_loss=0.3357, pruned_loss=0.1047, over 3824807.31 frames. ], batch size: 56, lr: 1.40e-02, grad_scale: 16.0 +2023-04-01 07:18:53,907 INFO [train.py:903] (3/4) Epoch 6, batch 2200, loss[loss=0.2813, simple_loss=0.3394, pruned_loss=0.1117, over 19846.00 frames. ], tot_loss[loss=0.2728, simple_loss=0.3359, pruned_loss=0.1048, over 3825633.30 frames. ], batch size: 52, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:18:55,064 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.621e+02 6.300e+02 8.031e+02 1.073e+03 2.013e+03, threshold=1.606e+03, percent-clipped=1.0 +2023-04-01 07:19:57,018 INFO [train.py:903] (3/4) Epoch 6, batch 2250, loss[loss=0.3108, simple_loss=0.3525, pruned_loss=0.1345, over 13333.00 frames. ], tot_loss[loss=0.2736, simple_loss=0.3365, pruned_loss=0.1054, over 3807965.20 frames. ], batch size: 136, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:20:58,359 INFO [train.py:903] (3/4) Epoch 6, batch 2300, loss[loss=0.2867, simple_loss=0.3506, pruned_loss=0.1114, over 19321.00 frames. ], tot_loss[loss=0.2733, simple_loss=0.3363, pruned_loss=0.1051, over 3814313.29 frames. ], batch size: 66, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:20:59,554 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.730e+02 6.689e+02 7.482e+02 9.822e+02 1.768e+03, threshold=1.496e+03, percent-clipped=2.0 +2023-04-01 07:21:14,335 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 07:22:00,470 INFO [train.py:903] (3/4) Epoch 6, batch 2350, loss[loss=0.2539, simple_loss=0.3153, pruned_loss=0.09625, over 19715.00 frames. ], tot_loss[loss=0.2714, simple_loss=0.3346, pruned_loss=0.1041, over 3800348.80 frames. ], batch size: 45, lr: 1.40e-02, grad_scale: 8.0 +2023-04-01 07:22:20,239 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36504.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:22:24,766 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-01 07:22:43,441 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 07:22:49,437 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36529.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:23:00,583 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 07:23:02,882 INFO [train.py:903] (3/4) Epoch 6, batch 2400, loss[loss=0.2918, simple_loss=0.3556, pruned_loss=0.114, over 19777.00 frames. ], tot_loss[loss=0.2712, simple_loss=0.3346, pruned_loss=0.1039, over 3794108.26 frames. ], batch size: 56, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:23:04,017 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.742e+02 5.696e+02 7.249e+02 9.157e+02 1.479e+03, threshold=1.450e+03, percent-clipped=0.0 +2023-04-01 07:23:07,407 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36543.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:23:08,806 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36544.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:23:39,050 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36569.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:23:48,693 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.34 vs. limit=5.0 +2023-04-01 07:24:06,981 INFO [train.py:903] (3/4) Epoch 6, batch 2450, loss[loss=0.2506, simple_loss=0.3203, pruned_loss=0.0904, over 19680.00 frames. ], tot_loss[loss=0.271, simple_loss=0.3345, pruned_loss=0.1037, over 3807127.53 frames. ], batch size: 53, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:24:46,707 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36623.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:24:51,078 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36626.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:24:56,983 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36630.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:25:07,903 INFO [train.py:903] (3/4) Epoch 6, batch 2500, loss[loss=0.2589, simple_loss=0.3289, pruned_loss=0.09443, over 19704.00 frames. ], tot_loss[loss=0.2709, simple_loss=0.3342, pruned_loss=0.1038, over 3819525.67 frames. ], batch size: 59, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:25:09,076 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.995e+02 6.265e+02 8.006e+02 9.274e+02 1.564e+03, threshold=1.601e+03, percent-clipped=1.0 +2023-04-01 07:25:30,341 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36658.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:26:09,577 INFO [train.py:903] (3/4) Epoch 6, batch 2550, loss[loss=0.2924, simple_loss=0.3543, pruned_loss=0.1153, over 19534.00 frames. ], tot_loss[loss=0.2714, simple_loss=0.3345, pruned_loss=0.1041, over 3814166.14 frames. ], batch size: 54, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:26:44,768 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36718.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:26:47,042 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4907, 1.5497, 2.1196, 2.7310, 2.2352, 2.2933, 2.1497, 2.6538], + device='cuda:3'), covar=tensor([0.0777, 0.1930, 0.1190, 0.0696, 0.1092, 0.0393, 0.0843, 0.0523], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0365, 0.0287, 0.0234, 0.0303, 0.0240, 0.0269, 0.0229], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:27:05,086 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 07:27:10,912 INFO [train.py:903] (3/4) Epoch 6, batch 2600, loss[loss=0.2926, simple_loss=0.3534, pruned_loss=0.1159, over 18231.00 frames. ], tot_loss[loss=0.2725, simple_loss=0.3356, pruned_loss=0.1047, over 3821019.64 frames. ], batch size: 83, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:27:12,686 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.869e+02 6.492e+02 8.253e+02 1.085e+03 2.742e+03, threshold=1.651e+03, percent-clipped=10.0 +2023-04-01 07:28:14,514 INFO [train.py:903] (3/4) Epoch 6, batch 2650, loss[loss=0.2223, simple_loss=0.2939, pruned_loss=0.07538, over 19583.00 frames. ], tot_loss[loss=0.271, simple_loss=0.3342, pruned_loss=0.1038, over 3826816.62 frames. ], batch size: 52, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:28:19,433 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1573, 2.1132, 2.2824, 3.4756, 2.2162, 3.5994, 3.2778, 2.0241], + device='cuda:3'), covar=tensor([0.2708, 0.2078, 0.0989, 0.1243, 0.2487, 0.0728, 0.1794, 0.1883], + device='cuda:3'), in_proj_covar=tensor([0.0659, 0.0650, 0.0573, 0.0801, 0.0679, 0.0567, 0.0694, 0.0608], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 07:28:21,697 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9300, 1.3470, 0.9998, 0.9540, 1.1897, 0.8853, 0.6923, 1.2473], + device='cuda:3'), covar=tensor([0.0455, 0.0532, 0.0880, 0.0470, 0.0375, 0.1017, 0.0614, 0.0328], + device='cuda:3'), in_proj_covar=tensor([0.0283, 0.0277, 0.0309, 0.0239, 0.0226, 0.0311, 0.0281, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:28:37,581 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 07:28:55,412 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36823.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:29:16,462 INFO [train.py:903] (3/4) Epoch 6, batch 2700, loss[loss=0.2779, simple_loss=0.3448, pruned_loss=0.1054, over 19771.00 frames. ], tot_loss[loss=0.2694, simple_loss=0.3334, pruned_loss=0.1027, over 3839364.91 frames. ], batch size: 56, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:29:17,597 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.977e+02 6.342e+02 7.427e+02 9.812e+02 2.890e+03, threshold=1.485e+03, percent-clipped=2.0 +2023-04-01 07:29:30,673 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2640, 3.8233, 2.3526, 3.4545, 1.1502, 3.4992, 3.5585, 3.5836], + device='cuda:3'), covar=tensor([0.0746, 0.1213, 0.2067, 0.0739, 0.3564, 0.0934, 0.0741, 0.0995], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0314, 0.0376, 0.0289, 0.0354, 0.0313, 0.0293, 0.0326], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:30:18,803 INFO [train.py:903] (3/4) Epoch 6, batch 2750, loss[loss=0.233, simple_loss=0.2941, pruned_loss=0.08592, over 19747.00 frames. ], tot_loss[loss=0.2707, simple_loss=0.334, pruned_loss=0.1037, over 3808523.06 frames. ], batch size: 47, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:30:51,740 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0265, 3.3177, 1.9061, 2.0643, 3.0066, 1.5779, 1.2362, 1.8705], + device='cuda:3'), covar=tensor([0.0904, 0.0400, 0.0784, 0.0571, 0.0342, 0.0992, 0.0803, 0.0596], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0276, 0.0307, 0.0237, 0.0224, 0.0306, 0.0279, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:30:51,760 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36914.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:31:09,604 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36929.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:31:22,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36939.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:31:23,033 INFO [train.py:903] (3/4) Epoch 6, batch 2800, loss[loss=0.2494, simple_loss=0.3289, pruned_loss=0.08495, over 19284.00 frames. ], tot_loss[loss=0.2705, simple_loss=0.3338, pruned_loss=0.1036, over 3810568.60 frames. ], batch size: 66, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:31:24,217 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.898e+02 7.168e+02 8.701e+02 1.243e+03 3.330e+03, threshold=1.740e+03, percent-clipped=17.0 +2023-04-01 07:31:43,504 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5823, 1.5285, 1.3839, 1.5170, 3.0760, 1.0416, 1.9904, 3.2121], + device='cuda:3'), covar=tensor([0.0335, 0.2300, 0.2472, 0.1641, 0.0613, 0.2397, 0.1381, 0.0368], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0317, 0.0319, 0.0296, 0.0312, 0.0317, 0.0293, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:31:56,917 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36967.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:32:00,392 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36970.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:32:02,129 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.45 vs. limit=5.0 +2023-04-01 07:32:06,249 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=36974.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:32:08,545 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36976.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:32:26,411 INFO [train.py:903] (3/4) Epoch 6, batch 2850, loss[loss=0.2214, simple_loss=0.2968, pruned_loss=0.07298, over 19728.00 frames. ], tot_loss[loss=0.2708, simple_loss=0.3342, pruned_loss=0.1037, over 3813726.31 frames. ], batch size: 51, lr: 1.39e-02, grad_scale: 8.0 +2023-04-01 07:33:05,605 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37022.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:33:28,652 INFO [train.py:903] (3/4) Epoch 6, batch 2900, loss[loss=0.2606, simple_loss=0.3342, pruned_loss=0.09354, over 19542.00 frames. ], tot_loss[loss=0.2713, simple_loss=0.3341, pruned_loss=0.1043, over 3798418.55 frames. ], batch size: 56, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:33:28,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 07:33:29,876 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.185e+02 6.105e+02 7.947e+02 1.025e+03 2.308e+03, threshold=1.589e+03, percent-clipped=2.0 +2023-04-01 07:33:55,110 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37062.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:34:20,477 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37082.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:34:24,058 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37085.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:34:28,768 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37089.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:34:29,562 INFO [train.py:903] (3/4) Epoch 6, batch 2950, loss[loss=0.24, simple_loss=0.3169, pruned_loss=0.08153, over 19685.00 frames. ], tot_loss[loss=0.2707, simple_loss=0.3338, pruned_loss=0.1038, over 3807394.35 frames. ], batch size: 53, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:35:31,157 INFO [train.py:903] (3/4) Epoch 6, batch 3000, loss[loss=0.3201, simple_loss=0.3708, pruned_loss=0.1347, over 13507.00 frames. ], tot_loss[loss=0.273, simple_loss=0.3355, pruned_loss=0.1052, over 3794894.48 frames. ], batch size: 135, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:35:31,158 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 07:35:43,638 INFO [train.py:937] (3/4) Epoch 6, validation: loss=0.1968, simple_loss=0.2962, pruned_loss=0.04867, over 944034.00 frames. +2023-04-01 07:35:43,639 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 07:35:44,844 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.497e+02 6.001e+02 7.289e+02 9.626e+02 1.809e+03, threshold=1.458e+03, percent-clipped=5.0 +2023-04-01 07:35:48,618 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 07:36:18,750 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37167.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:36:29,238 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4545, 1.2130, 1.6261, 1.3008, 2.6736, 3.6234, 3.3565, 3.7847], + device='cuda:3'), covar=tensor([0.1359, 0.3103, 0.2891, 0.1828, 0.0450, 0.0131, 0.0211, 0.0126], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0283, 0.0316, 0.0249, 0.0201, 0.0125, 0.0202, 0.0153], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:36:30,324 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37177.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:36:45,871 INFO [train.py:903] (3/4) Epoch 6, batch 3050, loss[loss=0.2411, simple_loss=0.3069, pruned_loss=0.08767, over 19402.00 frames. ], tot_loss[loss=0.2727, simple_loss=0.3351, pruned_loss=0.1051, over 3793482.04 frames. ], batch size: 48, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:37:48,504 INFO [train.py:903] (3/4) Epoch 6, batch 3100, loss[loss=0.2341, simple_loss=0.2955, pruned_loss=0.08639, over 19772.00 frames. ], tot_loss[loss=0.2724, simple_loss=0.3349, pruned_loss=0.1049, over 3802126.63 frames. ], batch size: 48, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:37:49,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.764e+02 6.699e+02 8.375e+02 1.038e+03 2.239e+03, threshold=1.675e+03, percent-clipped=7.0 +2023-04-01 07:38:08,564 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9651, 5.3264, 2.7875, 4.7110, 1.2276, 5.1352, 5.2574, 5.2654], + device='cuda:3'), covar=tensor([0.0422, 0.0869, 0.1923, 0.0561, 0.3919, 0.0605, 0.0492, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0322, 0.0380, 0.0292, 0.0357, 0.0317, 0.0296, 0.0331], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 07:38:28,951 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37273.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:38:42,259 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37282.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 07:38:46,571 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37286.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:38:50,733 INFO [train.py:903] (3/4) Epoch 6, batch 3150, loss[loss=0.2674, simple_loss=0.3321, pruned_loss=0.1014, over 18761.00 frames. ], tot_loss[loss=0.2729, simple_loss=0.3355, pruned_loss=0.1052, over 3805342.50 frames. ], batch size: 74, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:38:52,204 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4903, 1.0678, 1.2243, 1.2609, 2.1377, 0.8447, 1.7894, 2.1972], + device='cuda:3'), covar=tensor([0.0570, 0.2560, 0.2548, 0.1370, 0.0775, 0.2004, 0.1001, 0.0564], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0316, 0.0315, 0.0290, 0.0310, 0.0312, 0.0288, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:39:13,511 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 07:39:27,508 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37320.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:39:49,090 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37338.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:39:51,049 INFO [train.py:903] (3/4) Epoch 6, batch 3200, loss[loss=0.2896, simple_loss=0.3549, pruned_loss=0.1122, over 19715.00 frames. ], tot_loss[loss=0.275, simple_loss=0.3374, pruned_loss=0.1063, over 3802392.32 frames. ], batch size: 63, lr: 1.38e-02, grad_scale: 8.0 +2023-04-01 07:39:52,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.326e+02 6.253e+02 8.171e+02 9.916e+02 1.975e+03, threshold=1.634e+03, percent-clipped=4.0 +2023-04-01 07:39:52,632 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37341.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:39:58,108 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37345.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:20,457 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37363.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:24,439 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37366.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:24,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37366.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:29,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37370.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:42,615 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37382.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:50,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37388.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:40:52,606 INFO [train.py:903] (3/4) Epoch 6, batch 3250, loss[loss=0.2315, simple_loss=0.3018, pruned_loss=0.08054, over 19761.00 frames. ], tot_loss[loss=0.2752, simple_loss=0.337, pruned_loss=0.1067, over 3794677.71 frames. ], batch size: 51, lr: 1.38e-02, grad_scale: 4.0 +2023-04-01 07:41:45,891 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37433.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:41:48,059 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37435.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:41:53,529 INFO [train.py:903] (3/4) Epoch 6, batch 3300, loss[loss=0.2602, simple_loss=0.3397, pruned_loss=0.09042, over 19697.00 frames. ], tot_loss[loss=0.2742, simple_loss=0.3366, pruned_loss=0.1059, over 3808969.42 frames. ], batch size: 59, lr: 1.38e-02, grad_scale: 4.0 +2023-04-01 07:41:57,411 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.613e+02 6.108e+02 8.102e+02 1.002e+03 3.053e+03, threshold=1.620e+03, percent-clipped=3.0 +2023-04-01 07:42:01,111 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 07:42:17,161 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37458.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:42:44,807 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37481.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:42:56,289 INFO [train.py:903] (3/4) Epoch 6, batch 3350, loss[loss=0.282, simple_loss=0.3417, pruned_loss=0.1112, over 19791.00 frames. ], tot_loss[loss=0.2754, simple_loss=0.3376, pruned_loss=0.1066, over 3820446.07 frames. ], batch size: 56, lr: 1.38e-02, grad_scale: 4.0 +2023-04-01 07:43:04,921 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8308, 4.4227, 2.6326, 3.8896, 1.1065, 3.9940, 4.0455, 4.2715], + device='cuda:3'), covar=tensor([0.0526, 0.0984, 0.1960, 0.0688, 0.4056, 0.0995, 0.0721, 0.0802], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0324, 0.0385, 0.0296, 0.0363, 0.0320, 0.0300, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 07:43:49,279 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0695, 2.0577, 1.7363, 1.7627, 1.6276, 1.8419, 0.9073, 1.5242], + device='cuda:3'), covar=tensor([0.0208, 0.0274, 0.0203, 0.0238, 0.0386, 0.0305, 0.0482, 0.0366], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0298, 0.0297, 0.0315, 0.0387, 0.0309, 0.0283, 0.0302], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:43:56,139 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37538.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:43:57,994 INFO [train.py:903] (3/4) Epoch 6, batch 3400, loss[loss=0.2883, simple_loss=0.3472, pruned_loss=0.1147, over 18194.00 frames. ], tot_loss[loss=0.2736, simple_loss=0.3363, pruned_loss=0.1055, over 3818482.13 frames. ], batch size: 83, lr: 1.38e-02, grad_scale: 4.0 +2023-04-01 07:44:00,243 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.822e+02 6.377e+02 8.364e+02 1.096e+03 2.128e+03, threshold=1.673e+03, percent-clipped=5.0 +2023-04-01 07:44:27,169 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37563.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 07:44:59,262 INFO [train.py:903] (3/4) Epoch 6, batch 3450, loss[loss=0.2857, simple_loss=0.3518, pruned_loss=0.1098, over 18258.00 frames. ], tot_loss[loss=0.2739, simple_loss=0.3365, pruned_loss=0.1057, over 3822651.02 frames. ], batch size: 83, lr: 1.37e-02, grad_scale: 4.0 +2023-04-01 07:45:07,191 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 07:45:49,573 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37630.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:46:02,066 INFO [train.py:903] (3/4) Epoch 6, batch 3500, loss[loss=0.2932, simple_loss=0.356, pruned_loss=0.1152, over 18780.00 frames. ], tot_loss[loss=0.2713, simple_loss=0.3342, pruned_loss=0.1042, over 3825781.81 frames. ], batch size: 74, lr: 1.37e-02, grad_scale: 4.0 +2023-04-01 07:46:04,578 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 6.340e+02 8.060e+02 1.060e+03 3.220e+03, threshold=1.612e+03, percent-clipped=3.0 +2023-04-01 07:46:07,151 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.3414, 4.7938, 2.9211, 4.2749, 1.5503, 4.2604, 4.5433, 4.7405], + device='cuda:3'), covar=tensor([0.0492, 0.0992, 0.1996, 0.0673, 0.3786, 0.0814, 0.0724, 0.0811], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0323, 0.0383, 0.0293, 0.0360, 0.0318, 0.0297, 0.0331], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 07:46:07,358 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37644.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:46:38,300 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37669.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:47:04,850 INFO [train.py:903] (3/4) Epoch 6, batch 3550, loss[loss=0.342, simple_loss=0.4054, pruned_loss=0.1393, over 19595.00 frames. ], tot_loss[loss=0.2724, simple_loss=0.3346, pruned_loss=0.1051, over 3829661.49 frames. ], batch size: 61, lr: 1.37e-02, grad_scale: 4.0 +2023-04-01 07:47:07,128 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37691.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:47:34,916 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37716.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:47:49,120 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=37726.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:47:57,202 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2168, 1.2231, 1.5582, 0.9220, 2.4120, 2.9227, 2.6236, 3.0141], + device='cuda:3'), covar=tensor([0.1482, 0.3051, 0.2962, 0.2207, 0.0501, 0.0197, 0.0287, 0.0187], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0279, 0.0314, 0.0246, 0.0200, 0.0125, 0.0202, 0.0153], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:48:01,740 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6091, 1.6819, 1.3081, 1.1935, 1.1429, 1.3802, 0.1850, 0.7079], + device='cuda:3'), covar=tensor([0.0265, 0.0251, 0.0186, 0.0245, 0.0515, 0.0261, 0.0474, 0.0425], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0298, 0.0296, 0.0314, 0.0384, 0.0309, 0.0282, 0.0301], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:48:03,009 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37737.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:48:05,908 INFO [train.py:903] (3/4) Epoch 6, batch 3600, loss[loss=0.3299, simple_loss=0.3864, pruned_loss=0.1367, over 18243.00 frames. ], tot_loss[loss=0.272, simple_loss=0.3342, pruned_loss=0.1049, over 3816277.20 frames. ], batch size: 84, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:48:08,222 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.206e+02 7.040e+02 8.572e+02 1.227e+03 4.209e+03, threshold=1.714e+03, percent-clipped=12.0 +2023-04-01 07:48:12,119 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37745.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:48:33,210 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37762.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:48:35,748 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 07:48:41,991 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5484, 1.5714, 1.6483, 2.1281, 1.2881, 1.6906, 1.9772, 1.6359], + device='cuda:3'), covar=tensor([0.2313, 0.1829, 0.1024, 0.0973, 0.2090, 0.0946, 0.2147, 0.1731], + device='cuda:3'), in_proj_covar=tensor([0.0669, 0.0664, 0.0578, 0.0820, 0.0693, 0.0583, 0.0705, 0.0617], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 07:49:08,249 INFO [train.py:903] (3/4) Epoch 6, batch 3650, loss[loss=0.2741, simple_loss=0.3408, pruned_loss=0.1037, over 18772.00 frames. ], tot_loss[loss=0.2731, simple_loss=0.335, pruned_loss=0.1056, over 3782035.55 frames. ], batch size: 74, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:49:49,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4040, 1.3217, 1.7292, 1.5256, 3.1193, 4.4811, 4.4981, 4.8559], + device='cuda:3'), covar=tensor([0.1459, 0.2997, 0.2931, 0.1796, 0.0408, 0.0105, 0.0136, 0.0074], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0281, 0.0314, 0.0247, 0.0200, 0.0125, 0.0202, 0.0154], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 07:50:10,246 INFO [train.py:903] (3/4) Epoch 6, batch 3700, loss[loss=0.2654, simple_loss=0.336, pruned_loss=0.09741, over 19530.00 frames. ], tot_loss[loss=0.2716, simple_loss=0.3338, pruned_loss=0.1047, over 3784408.24 frames. ], batch size: 54, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:50:11,766 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:50:12,518 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.880e+02 5.969e+02 7.272e+02 1.002e+03 1.787e+03, threshold=1.454e+03, percent-clipped=1.0 +2023-04-01 07:50:21,877 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4621, 1.0890, 1.3121, 1.1805, 2.1743, 0.8824, 1.6597, 2.1371], + device='cuda:3'), covar=tensor([0.0589, 0.2500, 0.2319, 0.1387, 0.0733, 0.1959, 0.0977, 0.0592], + device='cuda:3'), in_proj_covar=tensor([0.0299, 0.0317, 0.0316, 0.0291, 0.0314, 0.0315, 0.0291, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:50:48,321 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0641, 2.2059, 2.1702, 3.3035, 2.1149, 3.3230, 3.0165, 1.9738], + device='cuda:3'), covar=tensor([0.2627, 0.1933, 0.0914, 0.1273, 0.2596, 0.0717, 0.1813, 0.1754], + device='cuda:3'), in_proj_covar=tensor([0.0664, 0.0658, 0.0576, 0.0811, 0.0689, 0.0579, 0.0699, 0.0610], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 07:51:13,476 INFO [train.py:903] (3/4) Epoch 6, batch 3750, loss[loss=0.2194, simple_loss=0.2855, pruned_loss=0.07664, over 18997.00 frames. ], tot_loss[loss=0.2702, simple_loss=0.3328, pruned_loss=0.1038, over 3786279.85 frames. ], batch size: 42, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:52:16,179 INFO [train.py:903] (3/4) Epoch 6, batch 3800, loss[loss=0.2526, simple_loss=0.316, pruned_loss=0.09464, over 19381.00 frames. ], tot_loss[loss=0.2697, simple_loss=0.3324, pruned_loss=0.1035, over 3803842.95 frames. ], batch size: 47, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:52:18,421 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.239e+02 5.762e+02 7.572e+02 9.178e+02 2.007e+03, threshold=1.514e+03, percent-clipped=4.0 +2023-04-01 07:52:47,826 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 07:53:17,777 INFO [train.py:903] (3/4) Epoch 6, batch 3850, loss[loss=0.2254, simple_loss=0.3018, pruned_loss=0.07452, over 19600.00 frames. ], tot_loss[loss=0.2713, simple_loss=0.3342, pruned_loss=0.1042, over 3804185.18 frames. ], batch size: 57, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:53:25,815 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37997.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:53:32,710 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38001.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:54:03,092 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=38026.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:54:20,062 INFO [train.py:903] (3/4) Epoch 6, batch 3900, loss[loss=0.2864, simple_loss=0.3444, pruned_loss=0.1142, over 19634.00 frames. ], tot_loss[loss=0.2721, simple_loss=0.3347, pruned_loss=0.1047, over 3819881.53 frames. ], batch size: 60, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:54:22,364 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.032e+02 6.773e+02 7.927e+02 9.711e+02 2.220e+03, threshold=1.585e+03, percent-clipped=5.0 +2023-04-01 07:55:08,056 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8479, 4.4635, 2.5248, 3.9340, 1.0508, 4.1057, 4.1802, 4.3455], + device='cuda:3'), covar=tensor([0.0587, 0.1058, 0.1910, 0.0668, 0.3843, 0.0656, 0.0595, 0.0728], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0328, 0.0387, 0.0297, 0.0362, 0.0322, 0.0302, 0.0336], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 07:55:24,003 INFO [train.py:903] (3/4) Epoch 6, batch 3950, loss[loss=0.318, simple_loss=0.3711, pruned_loss=0.1324, over 19481.00 frames. ], tot_loss[loss=0.2695, simple_loss=0.3329, pruned_loss=0.103, over 3831315.63 frames. ], batch size: 64, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:55:27,716 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 07:55:33,029 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38097.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:56:03,036 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=38122.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 07:56:25,972 INFO [train.py:903] (3/4) Epoch 6, batch 4000, loss[loss=0.205, simple_loss=0.2725, pruned_loss=0.06875, over 19753.00 frames. ], tot_loss[loss=0.2695, simple_loss=0.3328, pruned_loss=0.1031, over 3824613.01 frames. ], batch size: 46, lr: 1.37e-02, grad_scale: 8.0 +2023-04-01 07:56:28,273 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.559e+02 6.005e+02 7.523e+02 9.814e+02 1.567e+03, threshold=1.505e+03, percent-clipped=0.0 +2023-04-01 07:57:10,236 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 07:57:18,207 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3000, 3.8809, 2.2985, 3.5300, 0.9060, 3.4828, 3.5801, 3.7255], + device='cuda:3'), covar=tensor([0.0675, 0.1204, 0.2154, 0.0782, 0.4056, 0.0837, 0.0813, 0.0883], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0327, 0.0385, 0.0298, 0.0360, 0.0321, 0.0303, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 07:57:24,172 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0890, 3.6980, 2.0157, 2.0099, 3.1773, 1.5794, 1.1171, 2.0216], + device='cuda:3'), covar=tensor([0.0861, 0.0277, 0.0746, 0.0570, 0.0283, 0.0869, 0.0841, 0.0531], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0279, 0.0314, 0.0240, 0.0221, 0.0314, 0.0283, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 07:57:25,997 INFO [train.py:903] (3/4) Epoch 6, batch 4050, loss[loss=0.2665, simple_loss=0.3122, pruned_loss=0.1104, over 18523.00 frames. ], tot_loss[loss=0.27, simple_loss=0.3335, pruned_loss=0.1033, over 3823315.72 frames. ], batch size: 41, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 07:57:36,493 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7939, 4.2365, 4.4330, 4.3859, 1.5401, 4.0014, 3.6815, 4.0411], + device='cuda:3'), covar=tensor([0.0954, 0.0514, 0.0496, 0.0449, 0.4572, 0.0412, 0.0504, 0.1019], + device='cuda:3'), in_proj_covar=tensor([0.0565, 0.0485, 0.0671, 0.0543, 0.0623, 0.0401, 0.0420, 0.0616], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 07:58:27,529 INFO [train.py:903] (3/4) Epoch 6, batch 4100, loss[loss=0.2689, simple_loss=0.3281, pruned_loss=0.1049, over 19381.00 frames. ], tot_loss[loss=0.2694, simple_loss=0.3329, pruned_loss=0.1029, over 3834486.75 frames. ], batch size: 48, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 07:58:30,590 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.078e+02 6.486e+02 7.817e+02 9.790e+02 2.532e+03, threshold=1.563e+03, percent-clipped=8.0 +2023-04-01 07:59:04,837 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 07:59:30,607 INFO [train.py:903] (3/4) Epoch 6, batch 4150, loss[loss=0.2656, simple_loss=0.3334, pruned_loss=0.09886, over 19759.00 frames. ], tot_loss[loss=0.2694, simple_loss=0.3329, pruned_loss=0.1029, over 3838956.96 frames. ], batch size: 54, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:00:34,880 INFO [train.py:903] (3/4) Epoch 6, batch 4200, loss[loss=0.2525, simple_loss=0.3148, pruned_loss=0.09511, over 19751.00 frames. ], tot_loss[loss=0.2704, simple_loss=0.3336, pruned_loss=0.1036, over 3835259.62 frames. ], batch size: 51, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:00:36,333 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=38341.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:00:37,309 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.522e+02 6.104e+02 7.462e+02 9.650e+02 2.123e+03, threshold=1.492e+03, percent-clipped=5.0 +2023-04-01 08:00:39,356 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 08:01:21,329 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=38378.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:01:30,688 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5058, 1.1420, 1.3421, 1.1504, 2.1706, 0.8877, 1.7512, 2.1956], + device='cuda:3'), covar=tensor([0.0545, 0.2410, 0.2251, 0.1424, 0.0705, 0.1975, 0.0944, 0.0562], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0320, 0.0323, 0.0297, 0.0320, 0.0317, 0.0293, 0.0313], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:01:34,728 INFO [train.py:903] (3/4) Epoch 6, batch 4250, loss[loss=0.3281, simple_loss=0.3712, pruned_loss=0.1425, over 19780.00 frames. ], tot_loss[loss=0.2704, simple_loss=0.3334, pruned_loss=0.1037, over 3838495.76 frames. ], batch size: 56, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:01:48,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 08:02:01,564 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 08:02:03,924 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6097, 4.1900, 2.6159, 3.6897, 1.0703, 3.7435, 3.8190, 3.8996], + device='cuda:3'), covar=tensor([0.0589, 0.1049, 0.1901, 0.0716, 0.3847, 0.0829, 0.0687, 0.0815], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0323, 0.0382, 0.0294, 0.0361, 0.0321, 0.0304, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 08:02:34,847 INFO [train.py:903] (3/4) Epoch 6, batch 4300, loss[loss=0.2878, simple_loss=0.3411, pruned_loss=0.1172, over 13293.00 frames. ], tot_loss[loss=0.2735, simple_loss=0.3362, pruned_loss=0.1054, over 3814773.01 frames. ], batch size: 136, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:02:37,124 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.280e+02 6.582e+02 8.580e+02 1.078e+03 2.349e+03, threshold=1.716e+03, percent-clipped=8.0 +2023-04-01 08:02:55,998 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=38456.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:02:58,139 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0701, 3.6602, 2.0490, 1.7890, 3.1446, 1.6401, 1.3216, 1.9979], + device='cuda:3'), covar=tensor([0.0673, 0.0244, 0.0571, 0.0502, 0.0347, 0.0683, 0.0644, 0.0435], + device='cuda:3'), in_proj_covar=tensor([0.0284, 0.0284, 0.0319, 0.0245, 0.0229, 0.0318, 0.0289, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:03:27,667 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 08:03:35,904 INFO [train.py:903] (3/4) Epoch 6, batch 4350, loss[loss=0.1979, simple_loss=0.2735, pruned_loss=0.06115, over 19782.00 frames. ], tot_loss[loss=0.2736, simple_loss=0.336, pruned_loss=0.1056, over 3822629.10 frames. ], batch size: 47, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:04:04,774 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.67 vs. limit=2.0 +2023-04-01 08:04:40,411 INFO [train.py:903] (3/4) Epoch 6, batch 4400, loss[loss=0.2784, simple_loss=0.3451, pruned_loss=0.1059, over 19617.00 frames. ], tot_loss[loss=0.2737, simple_loss=0.3361, pruned_loss=0.1056, over 3813247.32 frames. ], batch size: 50, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:04:42,534 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.411e+02 6.588e+02 8.122e+02 1.160e+03 2.348e+03, threshold=1.624e+03, percent-clipped=4.0 +2023-04-01 08:05:02,028 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7233, 1.7827, 1.7799, 2.5644, 1.6326, 2.1662, 2.2425, 1.7236], + device='cuda:3'), covar=tensor([0.2231, 0.1779, 0.0982, 0.0910, 0.2089, 0.0849, 0.1929, 0.1740], + device='cuda:3'), in_proj_covar=tensor([0.0671, 0.0672, 0.0584, 0.0814, 0.0698, 0.0585, 0.0714, 0.0618], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:05:05,908 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 08:05:13,686 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 08:05:40,417 INFO [train.py:903] (3/4) Epoch 6, batch 4450, loss[loss=0.2352, simple_loss=0.303, pruned_loss=0.0837, over 19743.00 frames. ], tot_loss[loss=0.2737, simple_loss=0.3361, pruned_loss=0.1056, over 3822662.57 frames. ], batch size: 51, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:06:15,731 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9056, 1.7778, 1.9481, 2.2245, 4.4381, 1.3753, 2.1945, 4.4220], + device='cuda:3'), covar=tensor([0.0276, 0.2226, 0.2245, 0.1300, 0.0488, 0.2140, 0.1284, 0.0273], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0316, 0.0318, 0.0292, 0.0317, 0.0312, 0.0290, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:06:42,148 INFO [train.py:903] (3/4) Epoch 6, batch 4500, loss[loss=0.3086, simple_loss=0.3723, pruned_loss=0.1225, over 19659.00 frames. ], tot_loss[loss=0.2715, simple_loss=0.3345, pruned_loss=0.1042, over 3829976.31 frames. ], batch size: 58, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:06:44,515 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.035e+02 6.139e+02 7.598e+02 9.442e+02 2.713e+03, threshold=1.520e+03, percent-clipped=3.0 +2023-04-01 08:07:42,587 INFO [train.py:903] (3/4) Epoch 6, batch 4550, loss[loss=0.3502, simple_loss=0.3894, pruned_loss=0.1555, over 13641.00 frames. ], tot_loss[loss=0.2713, simple_loss=0.3345, pruned_loss=0.1041, over 3802452.99 frames. ], batch size: 136, lr: 1.36e-02, grad_scale: 8.0 +2023-04-01 08:07:53,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 08:08:10,869 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38712.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:08:17,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 08:08:22,020 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=38722.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:08:41,576 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=38737.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:08:45,375 INFO [train.py:903] (3/4) Epoch 6, batch 4600, loss[loss=0.2433, simple_loss=0.3057, pruned_loss=0.09048, over 19724.00 frames. ], tot_loss[loss=0.2712, simple_loss=0.3344, pruned_loss=0.104, over 3799599.13 frames. ], batch size: 51, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:08:47,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.208e+02 6.588e+02 8.068e+02 1.040e+03 1.807e+03, threshold=1.614e+03, percent-clipped=3.0 +2023-04-01 08:09:02,884 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.53 vs. limit=5.0 +2023-04-01 08:09:45,875 INFO [train.py:903] (3/4) Epoch 6, batch 4650, loss[loss=0.3004, simple_loss=0.3474, pruned_loss=0.1267, over 19735.00 frames. ], tot_loss[loss=0.2708, simple_loss=0.3346, pruned_loss=0.1035, over 3805182.12 frames. ], batch size: 51, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:10:01,546 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 08:10:05,092 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2460, 2.9313, 2.1368, 2.7456, 1.1586, 2.7587, 2.6271, 2.7216], + device='cuda:3'), covar=tensor([0.1139, 0.1384, 0.2066, 0.0863, 0.3477, 0.1154, 0.1037, 0.1280], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0323, 0.0384, 0.0290, 0.0357, 0.0318, 0.0298, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 08:10:11,691 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 08:10:42,127 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9405, 1.5110, 1.6994, 2.1444, 1.7695, 1.8107, 1.7011, 1.8874], + device='cuda:3'), covar=tensor([0.0863, 0.1597, 0.1256, 0.0822, 0.1144, 0.0460, 0.1012, 0.0639], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0358, 0.0285, 0.0235, 0.0301, 0.0241, 0.0271, 0.0229], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:10:43,249 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1125, 1.2709, 1.5683, 0.9430, 2.5144, 3.3171, 3.1207, 3.5325], + device='cuda:3'), covar=tensor([0.1485, 0.2813, 0.2785, 0.1992, 0.0409, 0.0149, 0.0201, 0.0143], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0288, 0.0321, 0.0251, 0.0203, 0.0130, 0.0205, 0.0156], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:10:43,262 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=38837.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:10:46,234 INFO [train.py:903] (3/4) Epoch 6, batch 4700, loss[loss=0.3303, simple_loss=0.3768, pruned_loss=0.1419, over 19771.00 frames. ], tot_loss[loss=0.2711, simple_loss=0.3352, pruned_loss=0.1035, over 3808089.52 frames. ], batch size: 54, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:10:47,745 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=38841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:10:48,619 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.118e+02 6.701e+02 8.528e+02 1.085e+03 2.106e+03, threshold=1.706e+03, percent-clipped=3.0 +2023-04-01 08:10:52,862 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-01 08:11:00,086 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8447, 1.9104, 2.2180, 2.1521, 2.9514, 3.4914, 3.5546, 3.8572], + device='cuda:3'), covar=tensor([0.1490, 0.3274, 0.3142, 0.1799, 0.1095, 0.0479, 0.0271, 0.0175], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0290, 0.0323, 0.0253, 0.0204, 0.0131, 0.0206, 0.0156], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:11:07,237 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 08:11:46,444 INFO [train.py:903] (3/4) Epoch 6, batch 4750, loss[loss=0.3589, simple_loss=0.3885, pruned_loss=0.1647, over 13123.00 frames. ], tot_loss[loss=0.2702, simple_loss=0.3338, pruned_loss=0.1033, over 3807776.02 frames. ], batch size: 135, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:12:47,942 INFO [train.py:903] (3/4) Epoch 6, batch 4800, loss[loss=0.3569, simple_loss=0.3974, pruned_loss=0.1582, over 18853.00 frames. ], tot_loss[loss=0.2725, simple_loss=0.3355, pruned_loss=0.1048, over 3814070.43 frames. ], batch size: 74, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:12:52,285 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.945e+02 6.641e+02 7.688e+02 1.063e+03 2.770e+03, threshold=1.538e+03, percent-clipped=4.0 +2023-04-01 08:13:49,882 INFO [train.py:903] (3/4) Epoch 6, batch 4850, loss[loss=0.2589, simple_loss=0.3268, pruned_loss=0.09551, over 19592.00 frames. ], tot_loss[loss=0.2712, simple_loss=0.3344, pruned_loss=0.104, over 3832492.44 frames. ], batch size: 52, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:14:10,477 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5676, 1.4723, 1.3494, 1.6473, 1.5976, 1.5486, 1.3860, 1.6046], + device='cuda:3'), covar=tensor([0.0744, 0.1149, 0.1092, 0.0669, 0.0875, 0.0422, 0.0902, 0.0558], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0365, 0.0292, 0.0241, 0.0307, 0.0247, 0.0276, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:14:14,747 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 08:14:36,520 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 08:14:42,289 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 08:14:42,317 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 08:14:51,531 INFO [train.py:903] (3/4) Epoch 6, batch 4900, loss[loss=0.2729, simple_loss=0.3411, pruned_loss=0.1024, over 19722.00 frames. ], tot_loss[loss=0.2712, simple_loss=0.3347, pruned_loss=0.1039, over 3836131.31 frames. ], batch size: 51, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:14:51,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 08:14:55,128 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.255e+02 6.820e+02 8.455e+02 1.082e+03 3.554e+03, threshold=1.691e+03, percent-clipped=3.0 +2023-04-01 08:15:00,044 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3524, 3.9001, 2.4816, 3.5040, 1.1272, 3.3027, 3.5537, 3.7123], + device='cuda:3'), covar=tensor([0.0587, 0.1038, 0.1983, 0.0786, 0.3890, 0.1134, 0.0806, 0.0924], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0317, 0.0378, 0.0290, 0.0355, 0.0316, 0.0296, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 08:15:12,461 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 08:15:36,705 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 08:15:38,508 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1594, 5.4503, 2.8582, 4.7202, 1.2294, 5.2193, 5.3114, 5.5776], + device='cuda:3'), covar=tensor([0.0395, 0.1045, 0.1960, 0.0639, 0.4055, 0.0619, 0.0678, 0.0724], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0318, 0.0380, 0.0291, 0.0357, 0.0316, 0.0297, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 08:15:52,288 INFO [train.py:903] (3/4) Epoch 6, batch 4950, loss[loss=0.3279, simple_loss=0.3806, pruned_loss=0.1376, over 19491.00 frames. ], tot_loss[loss=0.2709, simple_loss=0.3343, pruned_loss=0.1037, over 3833972.41 frames. ], batch size: 64, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:15:56,167 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39093.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:16:11,566 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 08:16:27,973 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39118.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:16:34,793 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 08:16:54,839 INFO [train.py:903] (3/4) Epoch 6, batch 5000, loss[loss=0.1927, simple_loss=0.261, pruned_loss=0.06218, over 19723.00 frames. ], tot_loss[loss=0.2705, simple_loss=0.334, pruned_loss=0.1035, over 3842063.66 frames. ], batch size: 46, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:16:59,188 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.473e+02 6.699e+02 7.802e+02 1.019e+03 2.317e+03, threshold=1.560e+03, percent-clipped=4.0 +2023-04-01 08:17:05,624 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 08:17:14,765 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 08:17:19,804 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9396, 4.2825, 4.6680, 4.5630, 1.5442, 4.2682, 3.8513, 4.2112], + device='cuda:3'), covar=tensor([0.1071, 0.0685, 0.0467, 0.0417, 0.4628, 0.0430, 0.0469, 0.0983], + device='cuda:3'), in_proj_covar=tensor([0.0565, 0.0481, 0.0659, 0.0544, 0.0616, 0.0404, 0.0411, 0.0606], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 08:17:51,519 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=39185.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:17:57,105 INFO [train.py:903] (3/4) Epoch 6, batch 5050, loss[loss=0.2571, simple_loss=0.3378, pruned_loss=0.0882, over 19539.00 frames. ], tot_loss[loss=0.2708, simple_loss=0.3342, pruned_loss=0.1037, over 3816217.60 frames. ], batch size: 56, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:18:12,172 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39203.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:18:29,777 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 08:18:43,123 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 08:18:53,228 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2570, 1.2042, 1.5789, 1.1317, 2.6836, 3.4807, 3.3256, 3.7022], + device='cuda:3'), covar=tensor([0.1434, 0.3115, 0.2748, 0.1894, 0.0391, 0.0132, 0.0197, 0.0133], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0286, 0.0318, 0.0248, 0.0202, 0.0129, 0.0203, 0.0154], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:18:57,176 INFO [train.py:903] (3/4) Epoch 6, batch 5100, loss[loss=0.3236, simple_loss=0.3745, pruned_loss=0.1364, over 19536.00 frames. ], tot_loss[loss=0.2695, simple_loss=0.3333, pruned_loss=0.1029, over 3826191.33 frames. ], batch size: 56, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:19:00,439 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.152e+02 6.811e+02 8.395e+02 1.062e+03 1.934e+03, threshold=1.679e+03, percent-clipped=3.0 +2023-04-01 08:19:05,509 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 08:19:08,954 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 08:19:13,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 08:19:17,188 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.16 vs. limit=5.0 +2023-04-01 08:19:38,200 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.11 vs. limit=2.0 +2023-04-01 08:19:57,716 INFO [train.py:903] (3/4) Epoch 6, batch 5150, loss[loss=0.2437, simple_loss=0.3165, pruned_loss=0.08545, over 19571.00 frames. ], tot_loss[loss=0.2686, simple_loss=0.3324, pruned_loss=0.1024, over 3828764.81 frames. ], batch size: 52, lr: 1.35e-02, grad_scale: 8.0 +2023-04-01 08:20:07,874 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.78 vs. limit=5.0 +2023-04-01 08:20:09,380 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 08:20:11,764 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39300.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:20:13,884 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:20:43,355 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 08:21:00,829 INFO [train.py:903] (3/4) Epoch 6, batch 5200, loss[loss=0.255, simple_loss=0.3238, pruned_loss=0.09312, over 19694.00 frames. ], tot_loss[loss=0.2705, simple_loss=0.3341, pruned_loss=0.1034, over 3819256.28 frames. ], batch size: 53, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:21:04,235 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.495e+02 6.191e+02 7.598e+02 1.014e+03 2.218e+03, threshold=1.520e+03, percent-clipped=2.0 +2023-04-01 08:21:14,291 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 08:21:56,503 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 08:22:01,129 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0346, 5.4713, 2.9045, 4.6447, 1.3067, 5.2905, 5.2280, 5.6129], + device='cuda:3'), covar=tensor([0.0395, 0.0914, 0.1774, 0.0606, 0.3896, 0.0590, 0.0511, 0.0603], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0319, 0.0378, 0.0288, 0.0356, 0.0314, 0.0298, 0.0327], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:22:03,020 INFO [train.py:903] (3/4) Epoch 6, batch 5250, loss[loss=0.2625, simple_loss=0.3242, pruned_loss=0.1004, over 19727.00 frames. ], tot_loss[loss=0.2706, simple_loss=0.3343, pruned_loss=0.1034, over 3830871.60 frames. ], batch size: 51, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:22:48,197 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4239, 2.3629, 1.6175, 1.4138, 2.2797, 1.1062, 1.0579, 1.5110], + device='cuda:3'), covar=tensor([0.0971, 0.0595, 0.0933, 0.0665, 0.0373, 0.1117, 0.0811, 0.0500], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0282, 0.0317, 0.0235, 0.0226, 0.0309, 0.0283, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:23:05,661 INFO [train.py:903] (3/4) Epoch 6, batch 5300, loss[loss=0.2875, simple_loss=0.354, pruned_loss=0.1105, over 19502.00 frames. ], tot_loss[loss=0.2691, simple_loss=0.3334, pruned_loss=0.1024, over 3835088.67 frames. ], batch size: 64, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:23:10,376 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.313e+02 6.384e+02 7.596e+02 9.799e+02 2.153e+03, threshold=1.519e+03, percent-clipped=7.0 +2023-04-01 08:23:22,414 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 08:23:56,538 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39481.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:24:06,733 INFO [train.py:903] (3/4) Epoch 6, batch 5350, loss[loss=0.37, simple_loss=0.3922, pruned_loss=0.1739, over 12961.00 frames. ], tot_loss[loss=0.2691, simple_loss=0.3331, pruned_loss=0.1026, over 3838488.26 frames. ], batch size: 136, lr: 1.34e-02, grad_scale: 4.0 +2023-04-01 08:24:27,511 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6270, 3.9734, 4.1452, 4.1534, 1.4837, 3.8695, 3.4912, 3.7784], + device='cuda:3'), covar=tensor([0.0971, 0.0661, 0.0578, 0.0488, 0.4131, 0.0450, 0.0519, 0.1101], + device='cuda:3'), in_proj_covar=tensor([0.0564, 0.0477, 0.0660, 0.0538, 0.0613, 0.0404, 0.0413, 0.0610], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 08:24:41,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 08:25:05,950 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.82 vs. limit=5.0 +2023-04-01 08:25:07,430 INFO [train.py:903] (3/4) Epoch 6, batch 5400, loss[loss=0.262, simple_loss=0.3296, pruned_loss=0.0972, over 19675.00 frames. ], tot_loss[loss=0.2679, simple_loss=0.3321, pruned_loss=0.1019, over 3839277.72 frames. ], batch size: 53, lr: 1.34e-02, grad_scale: 4.0 +2023-04-01 08:25:15,189 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.754e+02 6.102e+02 7.285e+02 1.015e+03 2.320e+03, threshold=1.457e+03, percent-clipped=6.0 +2023-04-01 08:25:19,786 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=39547.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:25:30,292 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39556.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:26:01,275 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39581.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:26:09,084 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0997, 1.2115, 1.5749, 1.0680, 2.6935, 3.2749, 3.0541, 3.4792], + device='cuda:3'), covar=tensor([0.1616, 0.3325, 0.3035, 0.2057, 0.0400, 0.0174, 0.0218, 0.0148], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0283, 0.0315, 0.0245, 0.0198, 0.0129, 0.0201, 0.0151], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:26:13,204 INFO [train.py:903] (3/4) Epoch 6, batch 5450, loss[loss=0.3302, simple_loss=0.3812, pruned_loss=0.1396, over 13249.00 frames. ], tot_loss[loss=0.2676, simple_loss=0.3318, pruned_loss=0.1017, over 3816874.25 frames. ], batch size: 135, lr: 1.34e-02, grad_scale: 4.0 +2023-04-01 08:26:28,353 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1916, 1.1830, 1.5216, 1.3418, 2.2030, 1.9437, 2.3171, 0.7995], + device='cuda:3'), covar=tensor([0.1839, 0.3226, 0.1830, 0.1529, 0.1215, 0.1586, 0.1226, 0.3024], + device='cuda:3'), in_proj_covar=tensor([0.0444, 0.0518, 0.0501, 0.0406, 0.0566, 0.0448, 0.0630, 0.0454], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:27:13,644 INFO [train.py:903] (3/4) Epoch 6, batch 5500, loss[loss=0.2666, simple_loss=0.3353, pruned_loss=0.09892, over 19593.00 frames. ], tot_loss[loss=0.2687, simple_loss=0.3325, pruned_loss=0.1025, over 3827248.11 frames. ], batch size: 61, lr: 1.34e-02, grad_scale: 4.0 +2023-04-01 08:27:18,175 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.238e+02 6.796e+02 8.775e+02 1.086e+03 2.226e+03, threshold=1.755e+03, percent-clipped=13.0 +2023-04-01 08:27:20,739 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=39646.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:27:33,896 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39657.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 08:27:36,853 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 08:27:40,476 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39662.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:27:58,083 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4186, 2.0990, 1.8042, 1.7760, 1.5321, 1.8248, 0.3524, 1.1782], + device='cuda:3'), covar=tensor([0.0238, 0.0276, 0.0242, 0.0359, 0.0565, 0.0385, 0.0609, 0.0509], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0298, 0.0299, 0.0314, 0.0384, 0.0310, 0.0286, 0.0301], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:28:02,691 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.7942, 1.3471, 1.0314, 0.9116, 1.1906, 0.8956, 0.6593, 1.2520], + device='cuda:3'), covar=tensor([0.0536, 0.0604, 0.0858, 0.0488, 0.0350, 0.0926, 0.0591, 0.0308], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0276, 0.0309, 0.0233, 0.0219, 0.0304, 0.0282, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:28:14,677 INFO [train.py:903] (3/4) Epoch 6, batch 5550, loss[loss=0.2954, simple_loss=0.3464, pruned_loss=0.1222, over 19596.00 frames. ], tot_loss[loss=0.2689, simple_loss=0.3324, pruned_loss=0.1027, over 3836765.44 frames. ], batch size: 52, lr: 1.34e-02, grad_scale: 4.0 +2023-04-01 08:28:21,827 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 08:29:11,949 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 08:29:15,353 INFO [train.py:903] (3/4) Epoch 6, batch 5600, loss[loss=0.206, simple_loss=0.2808, pruned_loss=0.06558, over 19809.00 frames. ], tot_loss[loss=0.2686, simple_loss=0.332, pruned_loss=0.1025, over 3838482.61 frames. ], batch size: 49, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:29:20,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.863e+02 6.221e+02 7.656e+02 9.358e+02 1.388e+03, threshold=1.531e+03, percent-clipped=0.0 +2023-04-01 08:29:43,047 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39761.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:29:51,151 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1407, 1.1699, 1.6265, 1.3092, 2.7511, 2.0348, 2.8530, 1.0610], + device='cuda:3'), covar=tensor([0.2034, 0.3414, 0.1928, 0.1614, 0.1261, 0.1719, 0.1376, 0.3165], + device='cuda:3'), in_proj_covar=tensor([0.0441, 0.0518, 0.0503, 0.0406, 0.0565, 0.0447, 0.0628, 0.0454], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:30:01,391 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7890, 1.8696, 1.8385, 2.2031, 4.2742, 1.1915, 2.2356, 4.3480], + device='cuda:3'), covar=tensor([0.0294, 0.2256, 0.2346, 0.1270, 0.0566, 0.2296, 0.1231, 0.0268], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0317, 0.0319, 0.0295, 0.0319, 0.0317, 0.0292, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:30:12,402 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39785.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:30:19,676 INFO [train.py:903] (3/4) Epoch 6, batch 5650, loss[loss=0.2954, simple_loss=0.3578, pruned_loss=0.1165, over 19783.00 frames. ], tot_loss[loss=0.2688, simple_loss=0.3325, pruned_loss=0.1025, over 3844865.56 frames. ], batch size: 56, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:31:02,178 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=39825.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:31:08,826 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 08:31:13,807 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39833.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:31:21,779 INFO [train.py:903] (3/4) Epoch 6, batch 5700, loss[loss=0.2589, simple_loss=0.3307, pruned_loss=0.09358, over 19287.00 frames. ], tot_loss[loss=0.2689, simple_loss=0.3328, pruned_loss=0.1024, over 3844263.31 frames. ], batch size: 66, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:31:26,538 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.066e+02 6.807e+02 8.639e+02 1.032e+03 2.369e+03, threshold=1.728e+03, percent-clipped=2.0 +2023-04-01 08:32:05,581 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-01 08:32:21,661 INFO [train.py:903] (3/4) Epoch 6, batch 5750, loss[loss=0.2987, simple_loss=0.3468, pruned_loss=0.1253, over 19612.00 frames. ], tot_loss[loss=0.2693, simple_loss=0.3332, pruned_loss=0.1026, over 3842556.93 frames. ], batch size: 52, lr: 1.34e-02, grad_scale: 8.0 +2023-04-01 08:32:23,999 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 08:32:30,900 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 08:32:35,628 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 08:32:36,987 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7647, 1.7844, 1.9008, 2.5338, 1.6137, 2.2639, 2.3640, 1.8052], + device='cuda:3'), covar=tensor([0.2572, 0.2123, 0.1079, 0.1117, 0.2393, 0.0943, 0.2102, 0.1928], + device='cuda:3'), in_proj_covar=tensor([0.0690, 0.0684, 0.0588, 0.0834, 0.0714, 0.0598, 0.0725, 0.0629], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:32:58,465 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39918.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:33:23,258 INFO [train.py:903] (3/4) Epoch 6, batch 5800, loss[loss=0.3003, simple_loss=0.3598, pruned_loss=0.1204, over 19608.00 frames. ], tot_loss[loss=0.2681, simple_loss=0.3325, pruned_loss=0.1018, over 3852544.64 frames. ], batch size: 57, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:33:23,642 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39940.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:33:28,244 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39943.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:33:29,036 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.561e+02 6.519e+02 7.840e+02 9.394e+02 2.455e+03, threshold=1.568e+03, percent-clipped=4.0 +2023-04-01 08:34:26,987 INFO [train.py:903] (3/4) Epoch 6, batch 5850, loss[loss=0.2096, simple_loss=0.2854, pruned_loss=0.06693, over 19376.00 frames. ], tot_loss[loss=0.2672, simple_loss=0.332, pruned_loss=0.1012, over 3852366.78 frames. ], batch size: 48, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:34:40,718 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40001.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 08:34:59,424 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40017.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:35:29,013 INFO [train.py:903] (3/4) Epoch 6, batch 5900, loss[loss=0.2555, simple_loss=0.3184, pruned_loss=0.09627, over 19357.00 frames. ], tot_loss[loss=0.2692, simple_loss=0.3337, pruned_loss=0.1023, over 3853802.00 frames. ], batch size: 48, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:35:31,459 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 08:35:31,831 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40042.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:35:33,716 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.177e+02 6.123e+02 7.695e+02 9.772e+02 2.844e+03, threshold=1.539e+03, percent-clipped=4.0 +2023-04-01 08:35:53,051 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 08:36:30,104 INFO [train.py:903] (3/4) Epoch 6, batch 5950, loss[loss=0.2685, simple_loss=0.3419, pruned_loss=0.09759, over 19770.00 frames. ], tot_loss[loss=0.2685, simple_loss=0.3333, pruned_loss=0.1018, over 3853722.52 frames. ], batch size: 54, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:37:03,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40116.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 08:37:07,975 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-04-01 08:37:15,813 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9193, 4.3030, 4.6751, 4.6485, 1.5643, 4.2484, 3.7481, 4.2483], + device='cuda:3'), covar=tensor([0.1238, 0.0658, 0.0520, 0.0473, 0.4650, 0.0428, 0.0572, 0.1075], + device='cuda:3'), in_proj_covar=tensor([0.0565, 0.0481, 0.0658, 0.0542, 0.0614, 0.0408, 0.0417, 0.0605], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 08:37:18,155 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40129.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:37:30,832 INFO [train.py:903] (3/4) Epoch 6, batch 6000, loss[loss=0.2546, simple_loss=0.3313, pruned_loss=0.08899, over 19691.00 frames. ], tot_loss[loss=0.2697, simple_loss=0.3338, pruned_loss=0.1029, over 3835530.98 frames. ], batch size: 53, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:37:30,832 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 08:37:43,218 INFO [train.py:937] (3/4) Epoch 6, validation: loss=0.1955, simple_loss=0.2951, pruned_loss=0.04789, over 944034.00 frames. +2023-04-01 08:37:43,220 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 08:37:47,771 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.228e+02 6.298e+02 7.514e+02 9.544e+02 1.960e+03, threshold=1.503e+03, percent-clipped=1.0 +2023-04-01 08:38:29,639 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40177.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:38:44,848 INFO [train.py:903] (3/4) Epoch 6, batch 6050, loss[loss=0.2521, simple_loss=0.3074, pruned_loss=0.0984, over 19763.00 frames. ], tot_loss[loss=0.2698, simple_loss=0.3337, pruned_loss=0.1029, over 3833436.86 frames. ], batch size: 47, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:38:53,736 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40196.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:39:14,887 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40213.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:39:24,443 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40221.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:39:25,615 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4278, 1.1587, 1.7073, 1.2012, 2.7615, 3.4670, 3.2685, 3.6892], + device='cuda:3'), covar=tensor([0.1376, 0.3114, 0.2688, 0.1927, 0.0381, 0.0117, 0.0197, 0.0128], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0285, 0.0315, 0.0247, 0.0205, 0.0131, 0.0202, 0.0156], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:39:43,619 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40236.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:39:48,075 INFO [train.py:903] (3/4) Epoch 6, batch 6100, loss[loss=0.2442, simple_loss=0.3075, pruned_loss=0.09042, over 19596.00 frames. ], tot_loss[loss=0.2698, simple_loss=0.3336, pruned_loss=0.1029, over 3827950.80 frames. ], batch size: 52, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:39:54,264 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.428e+02 6.367e+02 7.728e+02 1.144e+03 2.582e+03, threshold=1.546e+03, percent-clipped=10.0 +2023-04-01 08:39:54,685 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40244.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:40:21,634 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2267, 2.2652, 1.6110, 1.3165, 2.1326, 1.1054, 1.0524, 1.6951], + device='cuda:3'), covar=tensor([0.0829, 0.0495, 0.0919, 0.0607, 0.0368, 0.1026, 0.0708, 0.0396], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0280, 0.0317, 0.0235, 0.0225, 0.0307, 0.0284, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:40:51,267 INFO [train.py:903] (3/4) Epoch 6, batch 6150, loss[loss=0.313, simple_loss=0.3692, pruned_loss=0.1284, over 17314.00 frames. ], tot_loss[loss=0.2699, simple_loss=0.3338, pruned_loss=0.103, over 3828279.41 frames. ], batch size: 101, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:40:51,571 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40290.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:40:54,085 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40292.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:41:19,452 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 08:41:52,310 INFO [train.py:903] (3/4) Epoch 6, batch 6200, loss[loss=0.2671, simple_loss=0.3188, pruned_loss=0.1077, over 19384.00 frames. ], tot_loss[loss=0.2686, simple_loss=0.3326, pruned_loss=0.1023, over 3839801.12 frames. ], batch size: 48, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:41:57,223 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.101e+02 6.810e+02 8.218e+02 1.008e+03 2.334e+03, threshold=1.644e+03, percent-clipped=5.0 +2023-04-01 08:42:01,860 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40348.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:42:15,832 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4839, 1.1906, 1.1339, 1.3127, 1.1062, 1.2774, 1.0810, 1.3474], + device='cuda:3'), covar=tensor([0.0918, 0.1153, 0.1397, 0.0866, 0.0998, 0.0584, 0.1206, 0.0715], + device='cuda:3'), in_proj_covar=tensor([0.0245, 0.0356, 0.0284, 0.0236, 0.0302, 0.0246, 0.0274, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:42:33,807 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40372.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 08:42:54,280 INFO [train.py:903] (3/4) Epoch 6, batch 6250, loss[loss=0.3249, simple_loss=0.3726, pruned_loss=0.1386, over 17372.00 frames. ], tot_loss[loss=0.2678, simple_loss=0.3324, pruned_loss=0.1016, over 3847923.47 frames. ], batch size: 100, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:43:04,720 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40397.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 08:43:05,883 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1815, 1.5598, 1.7125, 1.9296, 1.7542, 1.8300, 1.5968, 1.9783], + device='cuda:3'), covar=tensor([0.0729, 0.1501, 0.1231, 0.0973, 0.1239, 0.0460, 0.1093, 0.0609], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0361, 0.0286, 0.0239, 0.0305, 0.0247, 0.0276, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:43:25,233 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2289, 1.3795, 1.1242, 1.1040, 0.9963, 1.1855, 0.0343, 0.3907], + device='cuda:3'), covar=tensor([0.0325, 0.0307, 0.0197, 0.0243, 0.0670, 0.0262, 0.0566, 0.0493], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0299, 0.0295, 0.0313, 0.0385, 0.0310, 0.0283, 0.0297], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 08:43:27,265 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 08:43:58,137 INFO [train.py:903] (3/4) Epoch 6, batch 6300, loss[loss=0.2974, simple_loss=0.3567, pruned_loss=0.1191, over 18875.00 frames. ], tot_loss[loss=0.2686, simple_loss=0.3328, pruned_loss=0.1022, over 3828158.51 frames. ], batch size: 74, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:44:03,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.716e+02 6.389e+02 8.265e+02 1.061e+03 2.633e+03, threshold=1.653e+03, percent-clipped=7.0 +2023-04-01 08:45:00,754 INFO [train.py:903] (3/4) Epoch 6, batch 6350, loss[loss=0.2639, simple_loss=0.3356, pruned_loss=0.09611, over 18859.00 frames. ], tot_loss[loss=0.2672, simple_loss=0.3318, pruned_loss=0.1013, over 3821165.44 frames. ], batch size: 74, lr: 1.33e-02, grad_scale: 8.0 +2023-04-01 08:45:12,837 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40500.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:45:24,299 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:45:33,428 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8661, 1.7653, 2.2040, 2.7782, 2.6356, 2.4481, 2.2682, 3.0630], + device='cuda:3'), covar=tensor([0.0652, 0.1725, 0.1056, 0.0705, 0.0956, 0.0382, 0.0916, 0.0388], + device='cuda:3'), in_proj_covar=tensor([0.0246, 0.0362, 0.0285, 0.0239, 0.0301, 0.0246, 0.0275, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:45:33,996 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-04-01 08:45:44,982 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40525.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:46:02,599 INFO [train.py:903] (3/4) Epoch 6, batch 6400, loss[loss=0.3116, simple_loss=0.36, pruned_loss=0.1317, over 13356.00 frames. ], tot_loss[loss=0.2693, simple_loss=0.3332, pruned_loss=0.1027, over 3807533.53 frames. ], batch size: 135, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:46:07,241 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.716e+02 6.528e+02 8.039e+02 1.077e+03 1.980e+03, threshold=1.608e+03, percent-clipped=3.0 +2023-04-01 08:46:12,206 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40548.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:46:15,844 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-01 08:46:23,192 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40557.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:46:44,867 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40573.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:46:52,963 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40580.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:47:04,379 INFO [train.py:903] (3/4) Epoch 6, batch 6450, loss[loss=0.2053, simple_loss=0.2771, pruned_loss=0.06675, over 19745.00 frames. ], tot_loss[loss=0.2683, simple_loss=0.3326, pruned_loss=0.102, over 3816032.43 frames. ], batch size: 47, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:47:50,031 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 08:47:59,685 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40634.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:48:07,614 INFO [train.py:903] (3/4) Epoch 6, batch 6500, loss[loss=0.2381, simple_loss=0.3163, pruned_loss=0.07997, over 19781.00 frames. ], tot_loss[loss=0.2682, simple_loss=0.3324, pruned_loss=0.102, over 3813475.52 frames. ], batch size: 56, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:48:13,022 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.640e+02 6.039e+02 7.833e+02 1.001e+03 2.233e+03, threshold=1.567e+03, percent-clipped=5.0 +2023-04-01 08:48:15,449 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 08:48:44,346 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40669.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:48:47,957 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40672.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:48:54,075 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.03 vs. limit=5.0 +2023-04-01 08:49:12,715 INFO [train.py:903] (3/4) Epoch 6, batch 6550, loss[loss=0.2348, simple_loss=0.3034, pruned_loss=0.08314, over 19637.00 frames. ], tot_loss[loss=0.2699, simple_loss=0.3335, pruned_loss=0.1032, over 3799378.95 frames. ], batch size: 50, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:49:15,178 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40692.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:49:18,806 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40695.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:49:55,779 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-01 08:50:15,126 INFO [train.py:903] (3/4) Epoch 6, batch 6600, loss[loss=0.3052, simple_loss=0.3566, pruned_loss=0.1269, over 19841.00 frames. ], tot_loss[loss=0.2689, simple_loss=0.3326, pruned_loss=0.1026, over 3817058.88 frames. ], batch size: 52, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:50:19,765 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.418e+02 6.030e+02 7.716e+02 9.913e+02 2.888e+03, threshold=1.543e+03, percent-clipped=3.0 +2023-04-01 08:50:26,143 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40749.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:50:58,489 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.73 vs. limit=2.0 +2023-04-01 08:51:17,592 INFO [train.py:903] (3/4) Epoch 6, batch 6650, loss[loss=0.2784, simple_loss=0.3307, pruned_loss=0.1131, over 19786.00 frames. ], tot_loss[loss=0.2695, simple_loss=0.3331, pruned_loss=0.1029, over 3812268.39 frames. ], batch size: 47, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:51:19,261 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9002, 1.9834, 1.9958, 2.8031, 1.9106, 2.6353, 2.5360, 1.8208], + device='cuda:3'), covar=tensor([0.2592, 0.2053, 0.1035, 0.1279, 0.2382, 0.0880, 0.2041, 0.1897], + device='cuda:3'), in_proj_covar=tensor([0.0682, 0.0685, 0.0589, 0.0835, 0.0709, 0.0593, 0.0727, 0.0630], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:51:36,072 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0511, 1.0968, 1.2557, 1.4579, 2.6594, 0.8504, 1.7360, 2.6533], + device='cuda:3'), covar=tensor([0.0485, 0.2668, 0.2747, 0.1479, 0.0692, 0.2424, 0.1271, 0.0517], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0316, 0.0321, 0.0296, 0.0319, 0.0316, 0.0292, 0.0314], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:51:40,543 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40807.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:52:19,368 INFO [train.py:903] (3/4) Epoch 6, batch 6700, loss[loss=0.2728, simple_loss=0.3436, pruned_loss=0.1009, over 19795.00 frames. ], tot_loss[loss=0.2702, simple_loss=0.3335, pruned_loss=0.1035, over 3823689.72 frames. ], batch size: 56, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:52:24,134 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.981e+02 6.315e+02 8.385e+02 9.952e+02 2.559e+03, threshold=1.677e+03, percent-clipped=5.0 +2023-04-01 08:52:39,404 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=40854.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:53:13,390 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40883.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:53:21,366 INFO [train.py:903] (3/4) Epoch 6, batch 6750, loss[loss=0.2777, simple_loss=0.3368, pruned_loss=0.1093, over 18083.00 frames. ], tot_loss[loss=0.2697, simple_loss=0.3329, pruned_loss=0.1032, over 3798378.78 frames. ], batch size: 83, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:53:42,240 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8933, 1.5669, 1.5441, 1.9289, 1.8872, 1.7581, 1.6517, 1.8803], + device='cuda:3'), covar=tensor([0.0967, 0.1658, 0.1382, 0.0958, 0.1172, 0.0462, 0.1072, 0.0662], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0360, 0.0285, 0.0237, 0.0301, 0.0243, 0.0271, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:53:47,612 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3129, 1.2943, 1.3091, 1.4328, 2.8662, 0.9117, 2.1788, 3.1004], + device='cuda:3'), covar=tensor([0.0395, 0.2430, 0.2521, 0.1531, 0.0643, 0.2397, 0.1096, 0.0348], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0312, 0.0315, 0.0293, 0.0316, 0.0311, 0.0289, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:53:53,156 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40918.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:54:04,525 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40928.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:54:17,526 INFO [train.py:903] (3/4) Epoch 6, batch 6800, loss[loss=0.271, simple_loss=0.3369, pruned_loss=0.1026, over 19610.00 frames. ], tot_loss[loss=0.2694, simple_loss=0.3326, pruned_loss=0.1031, over 3783063.11 frames. ], batch size: 57, lr: 1.32e-02, grad_scale: 8.0 +2023-04-01 08:54:23,019 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.567e+02 5.879e+02 7.609e+02 1.019e+03 2.150e+03, threshold=1.522e+03, percent-clipped=4.0 +2023-04-01 08:54:31,411 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40951.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:54:33,681 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40953.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:55:04,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 08:55:04,906 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 08:55:08,786 INFO [train.py:903] (3/4) Epoch 7, batch 0, loss[loss=0.2661, simple_loss=0.3276, pruned_loss=0.1024, over 18110.00 frames. ], tot_loss[loss=0.2661, simple_loss=0.3276, pruned_loss=0.1024, over 18110.00 frames. ], batch size: 83, lr: 1.24e-02, grad_scale: 8.0 +2023-04-01 08:55:08,787 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 08:55:20,406 INFO [train.py:937] (3/4) Epoch 7, validation: loss=0.1957, simple_loss=0.2957, pruned_loss=0.04779, over 944034.00 frames. +2023-04-01 08:55:20,407 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 08:55:21,992 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:55:30,919 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40976.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:55:32,932 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 08:55:40,165 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6732, 1.7014, 1.8220, 2.4092, 1.5851, 2.1287, 2.2817, 1.7288], + device='cuda:3'), covar=tensor([0.2113, 0.1760, 0.0900, 0.0907, 0.1944, 0.0758, 0.1708, 0.1654], + device='cuda:3'), in_proj_covar=tensor([0.0682, 0.0675, 0.0584, 0.0829, 0.0702, 0.0591, 0.0726, 0.0625], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 08:56:04,695 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41005.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:56:16,007 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41013.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:56:19,148 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41015.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:56:22,009 INFO [train.py:903] (3/4) Epoch 7, batch 50, loss[loss=0.3228, simple_loss=0.3757, pruned_loss=0.1349, over 19759.00 frames. ], tot_loss[loss=0.2594, simple_loss=0.3272, pruned_loss=0.09575, over 872368.61 frames. ], batch size: 54, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 08:56:36,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41030.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:56:51,947 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.786e+02 6.089e+02 7.435e+02 1.027e+03 3.072e+03, threshold=1.487e+03, percent-clipped=7.0 +2023-04-01 08:56:56,507 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 08:57:17,985 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41063.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:57:18,119 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41063.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:57:23,453 INFO [train.py:903] (3/4) Epoch 7, batch 100, loss[loss=0.2567, simple_loss=0.3139, pruned_loss=0.09978, over 19741.00 frames. ], tot_loss[loss=0.2651, simple_loss=0.3307, pruned_loss=0.09973, over 1526881.21 frames. ], batch size: 47, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 08:57:34,743 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 08:57:41,763 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4081, 2.1650, 1.6188, 1.5010, 2.1415, 1.2414, 1.3629, 1.7414], + device='cuda:3'), covar=tensor([0.0700, 0.0531, 0.0865, 0.0558, 0.0342, 0.0928, 0.0531, 0.0363], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0279, 0.0313, 0.0237, 0.0226, 0.0308, 0.0282, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 08:57:46,202 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41088.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:58:24,484 INFO [train.py:903] (3/4) Epoch 7, batch 150, loss[loss=0.3033, simple_loss=0.3476, pruned_loss=0.1294, over 13341.00 frames. ], tot_loss[loss=0.2625, simple_loss=0.3283, pruned_loss=0.0984, over 2038139.81 frames. ], batch size: 137, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 08:58:26,830 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41120.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:58:36,400 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41128.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 08:58:56,748 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.181e+02 6.219e+02 8.190e+02 1.094e+03 2.901e+03, threshold=1.638e+03, percent-clipped=4.0 +2023-04-01 08:59:22,651 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 08:59:23,717 INFO [train.py:903] (3/4) Epoch 7, batch 200, loss[loss=0.286, simple_loss=0.3574, pruned_loss=0.1073, over 17440.00 frames. ], tot_loss[loss=0.2657, simple_loss=0.3302, pruned_loss=0.1006, over 2428710.03 frames. ], batch size: 101, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 08:59:44,428 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-01 09:00:27,762 INFO [train.py:903] (3/4) Epoch 7, batch 250, loss[loss=0.2334, simple_loss=0.2985, pruned_loss=0.08419, over 18178.00 frames. ], tot_loss[loss=0.2667, simple_loss=0.3307, pruned_loss=0.1013, over 2735295.66 frames. ], batch size: 40, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:00:38,045 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41225.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:00:40,035 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41227.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:00:59,441 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.194e+02 6.866e+02 8.894e+02 1.080e+03 3.290e+03, threshold=1.779e+03, percent-clipped=6.0 +2023-04-01 09:01:06,554 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:01:06,849 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.01 vs. limit=5.0 +2023-04-01 09:01:23,879 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41262.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:01:30,754 INFO [train.py:903] (3/4) Epoch 7, batch 300, loss[loss=0.2312, simple_loss=0.3027, pruned_loss=0.07986, over 19584.00 frames. ], tot_loss[loss=0.2669, simple_loss=0.3309, pruned_loss=0.1014, over 2982663.54 frames. ], batch size: 52, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:02:31,484 INFO [train.py:903] (3/4) Epoch 7, batch 350, loss[loss=0.3065, simple_loss=0.3616, pruned_loss=0.1257, over 19588.00 frames. ], tot_loss[loss=0.2655, simple_loss=0.3299, pruned_loss=0.1006, over 3179759.22 frames. ], batch size: 61, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:02:33,986 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 09:03:02,035 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41342.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:03:04,944 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.733e+02 5.797e+02 7.460e+02 9.435e+02 2.818e+03, threshold=1.492e+03, percent-clipped=3.0 +2023-04-01 09:03:22,758 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41359.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:03:32,954 INFO [train.py:903] (3/4) Epoch 7, batch 400, loss[loss=0.286, simple_loss=0.3591, pruned_loss=0.1065, over 17356.00 frames. ], tot_loss[loss=0.2654, simple_loss=0.33, pruned_loss=0.1004, over 3323086.70 frames. ], batch size: 101, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:03:44,421 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41377.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:03:54,273 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41384.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:04:20,763 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41407.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:04:23,209 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41409.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:04:29,177 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8578, 0.8975, 1.0798, 0.9428, 1.3179, 1.2810, 1.3585, 0.5245], + device='cuda:3'), covar=tensor([0.1229, 0.2150, 0.1240, 0.0973, 0.0831, 0.1111, 0.0794, 0.2145], + device='cuda:3'), in_proj_covar=tensor([0.0452, 0.0523, 0.0510, 0.0409, 0.0561, 0.0455, 0.0631, 0.0456], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 09:04:34,727 INFO [train.py:903] (3/4) Epoch 7, batch 450, loss[loss=0.2664, simple_loss=0.3392, pruned_loss=0.09682, over 18192.00 frames. ], tot_loss[loss=0.2638, simple_loss=0.3288, pruned_loss=0.09941, over 3438988.87 frames. ], batch size: 83, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:05:02,679 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 09:05:03,850 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 09:05:06,095 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.663e+02 5.824e+02 7.719e+02 9.807e+02 3.448e+03, threshold=1.544e+03, percent-clipped=7.0 +2023-04-01 09:05:31,861 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=41464.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:05:37,398 INFO [train.py:903] (3/4) Epoch 7, batch 500, loss[loss=0.2496, simple_loss=0.3259, pruned_loss=0.0866, over 17449.00 frames. ], tot_loss[loss=0.2662, simple_loss=0.331, pruned_loss=0.1007, over 3521688.80 frames. ], batch size: 101, lr: 1.23e-02, grad_scale: 16.0 +2023-04-01 09:05:44,925 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41474.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:06:38,626 INFO [train.py:903] (3/4) Epoch 7, batch 550, loss[loss=0.256, simple_loss=0.3241, pruned_loss=0.09396, over 19536.00 frames. ], tot_loss[loss=0.2664, simple_loss=0.3316, pruned_loss=0.1006, over 3592642.43 frames. ], batch size: 54, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:06:43,743 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41522.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:07:09,851 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.859e+02 6.032e+02 7.411e+02 9.275e+02 1.625e+03, threshold=1.482e+03, percent-clipped=1.0 +2023-04-01 09:07:37,922 INFO [train.py:903] (3/4) Epoch 7, batch 600, loss[loss=0.2798, simple_loss=0.348, pruned_loss=0.1058, over 19600.00 frames. ], tot_loss[loss=0.2676, simple_loss=0.3325, pruned_loss=0.1014, over 3641256.42 frames. ], batch size: 57, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:07:50,883 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41579.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:08:13,417 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 09:08:16,237 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41598.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:08:39,321 INFO [train.py:903] (3/4) Epoch 7, batch 650, loss[loss=0.2056, simple_loss=0.2879, pruned_loss=0.06167, over 19585.00 frames. ], tot_loss[loss=0.2666, simple_loss=0.3315, pruned_loss=0.1009, over 3684570.28 frames. ], batch size: 52, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:08:45,302 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41623.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:09:00,446 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41633.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:09:13,511 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.335e+02 5.719e+02 7.418e+02 1.065e+03 4.334e+03, threshold=1.484e+03, percent-clipped=7.0 +2023-04-01 09:09:28,650 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41658.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:09:40,773 INFO [train.py:903] (3/4) Epoch 7, batch 700, loss[loss=0.2556, simple_loss=0.3267, pruned_loss=0.09223, over 19537.00 frames. ], tot_loss[loss=0.2673, simple_loss=0.3323, pruned_loss=0.1011, over 3722284.27 frames. ], batch size: 54, lr: 1.23e-02, grad_scale: 8.0 +2023-04-01 09:09:48,605 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41672.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:10:30,227 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 09:10:45,013 INFO [train.py:903] (3/4) Epoch 7, batch 750, loss[loss=0.2412, simple_loss=0.3021, pruned_loss=0.09013, over 19747.00 frames. ], tot_loss[loss=0.2656, simple_loss=0.331, pruned_loss=0.1001, over 3753760.19 frames. ], batch size: 46, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:10:47,658 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9250, 3.5852, 2.1850, 3.2770, 1.1073, 3.3870, 3.2380, 3.3683], + device='cuda:3'), covar=tensor([0.0735, 0.1088, 0.2123, 0.0784, 0.3695, 0.0741, 0.0807, 0.0947], + device='cuda:3'), in_proj_covar=tensor([0.0378, 0.0334, 0.0378, 0.0300, 0.0356, 0.0312, 0.0304, 0.0338], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 09:10:59,392 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41730.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:11:15,966 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.723e+02 5.751e+02 6.886e+02 8.721e+02 1.519e+03, threshold=1.377e+03, percent-clipped=2.0 +2023-04-01 09:11:31,835 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41755.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:11:46,757 INFO [train.py:903] (3/4) Epoch 7, batch 800, loss[loss=0.2451, simple_loss=0.3238, pruned_loss=0.0832, over 19777.00 frames. ], tot_loss[loss=0.2642, simple_loss=0.3298, pruned_loss=0.09933, over 3778520.07 frames. ], batch size: 56, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:11:56,190 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 09:11:58,727 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41778.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:12:31,084 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41803.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:12:48,119 INFO [train.py:903] (3/4) Epoch 7, batch 850, loss[loss=0.2795, simple_loss=0.3374, pruned_loss=0.1107, over 19846.00 frames. ], tot_loss[loss=0.2651, simple_loss=0.3307, pruned_loss=0.09972, over 3791770.06 frames. ], batch size: 52, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:13:11,214 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41835.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:13:23,186 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.000e+02 6.514e+02 8.083e+02 9.646e+02 1.896e+03, threshold=1.617e+03, percent-clipped=5.0 +2023-04-01 09:13:38,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 09:13:40,992 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41860.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:13:50,191 INFO [train.py:903] (3/4) Epoch 7, batch 900, loss[loss=0.2414, simple_loss=0.3248, pruned_loss=0.07904, over 19556.00 frames. ], tot_loss[loss=0.263, simple_loss=0.3292, pruned_loss=0.09842, over 3813760.79 frames. ], batch size: 56, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:14:49,752 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0349, 1.1430, 1.3191, 1.4182, 2.6515, 0.9628, 2.1173, 2.7030], + device='cuda:3'), covar=tensor([0.0430, 0.2635, 0.2477, 0.1603, 0.0636, 0.2251, 0.0910, 0.0471], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0320, 0.0323, 0.0299, 0.0324, 0.0317, 0.0300, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:14:51,584 INFO [train.py:903] (3/4) Epoch 7, batch 950, loss[loss=0.3613, simple_loss=0.4084, pruned_loss=0.1571, over 19574.00 frames. ], tot_loss[loss=0.2625, simple_loss=0.3284, pruned_loss=0.09827, over 3824181.24 frames. ], batch size: 52, lr: 1.22e-02, grad_scale: 4.0 +2023-04-01 09:14:56,063 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 09:15:01,469 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41924.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:15:26,245 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.318e+02 6.505e+02 7.519e+02 9.487e+02 1.757e+03, threshold=1.504e+03, percent-clipped=1.0 +2023-04-01 09:15:54,009 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41966.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 09:15:55,890 INFO [train.py:903] (3/4) Epoch 7, batch 1000, loss[loss=0.2758, simple_loss=0.3447, pruned_loss=0.1035, over 19284.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3288, pruned_loss=0.09846, over 3812607.78 frames. ], batch size: 66, lr: 1.22e-02, grad_scale: 4.0 +2023-04-01 09:16:26,040 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-01 09:16:26,564 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2140, 2.0844, 1.7360, 1.5900, 1.5322, 1.6751, 0.2305, 1.0004], + device='cuda:3'), covar=tensor([0.0240, 0.0261, 0.0206, 0.0348, 0.0572, 0.0332, 0.0610, 0.0462], + device='cuda:3'), in_proj_covar=tensor([0.0298, 0.0306, 0.0297, 0.0324, 0.0391, 0.0317, 0.0290, 0.0305], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 09:16:48,569 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 09:16:56,960 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=42016.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:16:59,085 INFO [train.py:903] (3/4) Epoch 7, batch 1050, loss[loss=0.2458, simple_loss=0.3198, pruned_loss=0.08592, over 19671.00 frames. ], tot_loss[loss=0.2656, simple_loss=0.3309, pruned_loss=0.1002, over 3823863.30 frames. ], batch size: 53, lr: 1.22e-02, grad_scale: 4.0 +2023-04-01 09:17:29,376 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 09:17:33,748 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.193e+02 5.894e+02 7.129e+02 9.584e+02 2.561e+03, threshold=1.426e+03, percent-clipped=5.0 +2023-04-01 09:18:00,563 INFO [train.py:903] (3/4) Epoch 7, batch 1100, loss[loss=0.3934, simple_loss=0.4259, pruned_loss=0.1804, over 19697.00 frames. ], tot_loss[loss=0.2644, simple_loss=0.3297, pruned_loss=0.09955, over 3833734.83 frames. ], batch size: 59, lr: 1.22e-02, grad_scale: 4.0 +2023-04-01 09:19:03,370 INFO [train.py:903] (3/4) Epoch 7, batch 1150, loss[loss=0.2457, simple_loss=0.307, pruned_loss=0.09221, over 19442.00 frames. ], tot_loss[loss=0.2632, simple_loss=0.3286, pruned_loss=0.09888, over 3834517.93 frames. ], batch size: 48, lr: 1.22e-02, grad_scale: 4.0 +2023-04-01 09:19:21,146 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42131.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:19:33,492 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.5885, 0.8468, 0.6465, 0.6390, 0.7991, 0.5749, 0.6208, 0.7339], + device='cuda:3'), covar=tensor([0.0307, 0.0399, 0.0577, 0.0324, 0.0270, 0.0619, 0.0350, 0.0263], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0281, 0.0313, 0.0237, 0.0226, 0.0308, 0.0289, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:19:37,737 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.517e+02 5.892e+02 7.369e+02 1.011e+03 1.805e+03, threshold=1.474e+03, percent-clipped=4.0 +2023-04-01 09:19:51,767 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5447, 2.3899, 1.6798, 1.5560, 2.2601, 1.3193, 1.3577, 1.7526], + device='cuda:3'), covar=tensor([0.0638, 0.0440, 0.0693, 0.0524, 0.0318, 0.0805, 0.0589, 0.0380], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0280, 0.0313, 0.0237, 0.0225, 0.0308, 0.0287, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:20:05,772 INFO [train.py:903] (3/4) Epoch 7, batch 1200, loss[loss=0.2707, simple_loss=0.3371, pruned_loss=0.1021, over 19542.00 frames. ], tot_loss[loss=0.2646, simple_loss=0.3298, pruned_loss=0.09969, over 3827125.70 frames. ], batch size: 56, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:20:30,581 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 09:21:08,030 INFO [train.py:903] (3/4) Epoch 7, batch 1250, loss[loss=0.2695, simple_loss=0.3319, pruned_loss=0.1036, over 19765.00 frames. ], tot_loss[loss=0.2652, simple_loss=0.33, pruned_loss=0.1002, over 3811668.20 frames. ], batch size: 54, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:21:43,462 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.288e+02 6.237e+02 7.824e+02 1.034e+03 2.254e+03, threshold=1.565e+03, percent-clipped=6.0 +2023-04-01 09:22:09,630 INFO [train.py:903] (3/4) Epoch 7, batch 1300, loss[loss=0.2218, simple_loss=0.2868, pruned_loss=0.07842, over 19756.00 frames. ], tot_loss[loss=0.2638, simple_loss=0.329, pruned_loss=0.09937, over 3824146.00 frames. ], batch size: 47, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:22:09,798 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=42268.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:23:02,530 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=42310.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 09:23:12,310 INFO [train.py:903] (3/4) Epoch 7, batch 1350, loss[loss=0.2965, simple_loss=0.3598, pruned_loss=0.1166, over 18786.00 frames. ], tot_loss[loss=0.2633, simple_loss=0.3288, pruned_loss=0.09892, over 3840676.58 frames. ], batch size: 74, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:23:47,360 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.606e+02 6.578e+02 7.819e+02 1.024e+03 2.032e+03, threshold=1.564e+03, percent-clipped=3.0 +2023-04-01 09:23:50,218 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8905, 1.3660, 1.4453, 2.0961, 1.6173, 2.1659, 2.0437, 1.8851], + device='cuda:3'), covar=tensor([0.0697, 0.0966, 0.0994, 0.0851, 0.0856, 0.0618, 0.0812, 0.0607], + device='cuda:3'), in_proj_covar=tensor([0.0223, 0.0236, 0.0233, 0.0265, 0.0256, 0.0219, 0.0216, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 09:24:15,697 INFO [train.py:903] (3/4) Epoch 7, batch 1400, loss[loss=0.2312, simple_loss=0.2921, pruned_loss=0.08518, over 19409.00 frames. ], tot_loss[loss=0.2617, simple_loss=0.3276, pruned_loss=0.09794, over 3847704.61 frames. ], batch size: 48, lr: 1.22e-02, grad_scale: 8.0 +2023-04-01 09:24:16,606 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-04-01 09:24:34,429 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42383.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:24:39,201 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42387.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:24:50,674 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42397.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:24:52,468 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 09:25:10,403 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42412.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:25:12,408 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 09:25:17,935 INFO [train.py:903] (3/4) Epoch 7, batch 1450, loss[loss=0.2848, simple_loss=0.3497, pruned_loss=0.1099, over 19467.00 frames. ], tot_loss[loss=0.2601, simple_loss=0.3262, pruned_loss=0.09697, over 3850506.01 frames. ], batch size: 64, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:25:26,432 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42425.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 09:25:31,023 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0839, 1.1578, 1.5352, 0.8011, 2.4003, 2.9489, 2.6998, 3.1372], + device='cuda:3'), covar=tensor([0.1520, 0.3161, 0.2863, 0.2192, 0.0434, 0.0206, 0.0266, 0.0174], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0282, 0.0310, 0.0244, 0.0200, 0.0134, 0.0201, 0.0159], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 09:25:53,250 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.957e+02 5.854e+02 7.466e+02 9.791e+02 2.115e+03, threshold=1.493e+03, percent-clipped=4.0 +2023-04-01 09:26:18,636 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2827, 3.8685, 2.5962, 3.4974, 1.4990, 3.5052, 3.5192, 3.6785], + device='cuda:3'), covar=tensor([0.0689, 0.1082, 0.1745, 0.0647, 0.2925, 0.0781, 0.0681, 0.0984], + device='cuda:3'), in_proj_covar=tensor([0.0371, 0.0329, 0.0381, 0.0290, 0.0353, 0.0313, 0.0297, 0.0329], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 09:26:19,589 INFO [train.py:903] (3/4) Epoch 7, batch 1500, loss[loss=0.3009, simple_loss=0.366, pruned_loss=0.1179, over 19079.00 frames. ], tot_loss[loss=0.26, simple_loss=0.3262, pruned_loss=0.09686, over 3849928.90 frames. ], batch size: 69, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:27:20,485 INFO [train.py:903] (3/4) Epoch 7, batch 1550, loss[loss=0.2872, simple_loss=0.3399, pruned_loss=0.1173, over 19574.00 frames. ], tot_loss[loss=0.2623, simple_loss=0.3278, pruned_loss=0.09841, over 3841319.56 frames. ], batch size: 52, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:27:29,126 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42524.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:27:55,724 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.300e+02 6.427e+02 8.071e+02 9.919e+02 2.182e+03, threshold=1.614e+03, percent-clipped=7.0 +2023-04-01 09:28:23,010 INFO [train.py:903] (3/4) Epoch 7, batch 1600, loss[loss=0.2718, simple_loss=0.3417, pruned_loss=0.1009, over 19536.00 frames. ], tot_loss[loss=0.2644, simple_loss=0.3295, pruned_loss=0.0996, over 3834713.63 frames. ], batch size: 56, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:28:41,274 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 09:29:24,655 INFO [train.py:903] (3/4) Epoch 7, batch 1650, loss[loss=0.2558, simple_loss=0.3234, pruned_loss=0.09405, over 19840.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3281, pruned_loss=0.09887, over 3837546.71 frames. ], batch size: 52, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:29:50,171 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42639.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:29:59,743 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.059e+02 6.175e+02 7.945e+02 9.824e+02 2.630e+03, threshold=1.589e+03, percent-clipped=4.0 +2023-04-01 09:30:23,159 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42664.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:30:26,330 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42667.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:30:26,438 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1530, 2.1017, 1.5256, 1.5218, 1.3504, 1.6456, 0.4063, 1.0470], + device='cuda:3'), covar=tensor([0.0268, 0.0293, 0.0275, 0.0399, 0.0646, 0.0417, 0.0616, 0.0540], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0308, 0.0303, 0.0327, 0.0396, 0.0323, 0.0290, 0.0306], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 09:30:27,189 INFO [train.py:903] (3/4) Epoch 7, batch 1700, loss[loss=0.2914, simple_loss=0.3555, pruned_loss=0.1136, over 19530.00 frames. ], tot_loss[loss=0.2622, simple_loss=0.3276, pruned_loss=0.09839, over 3835582.27 frames. ], batch size: 64, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:30:42,167 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-01 09:30:43,931 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42681.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 09:31:02,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 09:31:15,455 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42706.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 09:31:29,060 INFO [train.py:903] (3/4) Epoch 7, batch 1750, loss[loss=0.2854, simple_loss=0.3537, pruned_loss=0.1086, over 19088.00 frames. ], tot_loss[loss=0.2631, simple_loss=0.3285, pruned_loss=0.09887, over 3844309.88 frames. ], batch size: 69, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:31:59,547 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=42741.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:32:05,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.512e+02 5.682e+02 7.177e+02 9.179e+02 1.731e+03, threshold=1.435e+03, percent-clipped=1.0 +2023-04-01 09:32:33,263 INFO [train.py:903] (3/4) Epoch 7, batch 1800, loss[loss=0.277, simple_loss=0.3442, pruned_loss=0.105, over 18827.00 frames. ], tot_loss[loss=0.2635, simple_loss=0.3286, pruned_loss=0.09922, over 3831760.99 frames. ], batch size: 74, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:33:27,619 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 09:33:35,067 INFO [train.py:903] (3/4) Epoch 7, batch 1850, loss[loss=0.3592, simple_loss=0.3984, pruned_loss=0.16, over 19328.00 frames. ], tot_loss[loss=0.2633, simple_loss=0.3285, pruned_loss=0.09907, over 3834955.38 frames. ], batch size: 66, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:34:04,823 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 09:34:09,013 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.550e+02 6.596e+02 7.909e+02 1.066e+03 2.536e+03, threshold=1.582e+03, percent-clipped=10.0 +2023-04-01 09:34:22,471 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42856.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:34:37,124 INFO [train.py:903] (3/4) Epoch 7, batch 1900, loss[loss=0.2427, simple_loss=0.3051, pruned_loss=0.09021, over 19389.00 frames. ], tot_loss[loss=0.2633, simple_loss=0.3287, pruned_loss=0.09897, over 3821520.77 frames. ], batch size: 48, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:34:37,304 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=42868.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:34:48,040 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42877.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:34:51,102 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 09:34:55,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 09:35:14,078 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42897.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:35:21,940 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 09:35:38,557 INFO [train.py:903] (3/4) Epoch 7, batch 1950, loss[loss=0.2999, simple_loss=0.3564, pruned_loss=0.1217, over 17365.00 frames. ], tot_loss[loss=0.2653, simple_loss=0.3305, pruned_loss=0.1, over 3812459.48 frames. ], batch size: 101, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:36:15,250 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.499e+02 6.682e+02 8.244e+02 9.665e+02 2.689e+03, threshold=1.649e+03, percent-clipped=3.0 +2023-04-01 09:36:23,015 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.81 vs. limit=5.0 +2023-04-01 09:36:41,126 INFO [train.py:903] (3/4) Epoch 7, batch 2000, loss[loss=0.289, simple_loss=0.3499, pruned_loss=0.114, over 19480.00 frames. ], tot_loss[loss=0.2646, simple_loss=0.3303, pruned_loss=0.09948, over 3812896.17 frames. ], batch size: 64, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:37:00,261 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42983.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:37:35,138 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=43011.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:37:36,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 09:37:43,658 INFO [train.py:903] (3/4) Epoch 7, batch 2050, loss[loss=0.2782, simple_loss=0.3469, pruned_loss=0.1048, over 19781.00 frames. ], tot_loss[loss=0.2656, simple_loss=0.3308, pruned_loss=0.1002, over 3803729.23 frames. ], batch size: 54, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:37:56,213 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 09:37:57,395 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 09:38:17,409 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.956e+02 6.149e+02 7.669e+02 9.586e+02 2.177e+03, threshold=1.534e+03, percent-clipped=1.0 +2023-04-01 09:38:18,641 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 09:38:38,206 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4221, 1.4861, 1.9119, 2.4352, 1.8495, 2.2326, 2.5430, 2.2709], + device='cuda:3'), covar=tensor([0.0753, 0.1207, 0.1085, 0.1101, 0.1080, 0.0928, 0.0937, 0.0748], + device='cuda:3'), in_proj_covar=tensor([0.0223, 0.0236, 0.0231, 0.0263, 0.0254, 0.0220, 0.0215, 0.0210], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 09:38:46,691 INFO [train.py:903] (3/4) Epoch 7, batch 2100, loss[loss=0.2616, simple_loss=0.3337, pruned_loss=0.0947, over 19777.00 frames. ], tot_loss[loss=0.2632, simple_loss=0.3288, pruned_loss=0.09885, over 3811216.35 frames. ], batch size: 56, lr: 1.21e-02, grad_scale: 8.0 +2023-04-01 09:38:47,710 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-01 09:39:13,061 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 09:39:28,856 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7771, 1.2241, 1.4061, 2.0680, 1.6368, 1.9138, 2.1600, 1.6131], + device='cuda:3'), covar=tensor([0.0757, 0.1030, 0.1068, 0.0819, 0.0854, 0.0711, 0.0724, 0.0688], + device='cuda:3'), in_proj_covar=tensor([0.0225, 0.0239, 0.0235, 0.0265, 0.0256, 0.0221, 0.0218, 0.0213], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 09:39:35,510 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 09:39:41,536 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43112.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:39:48,067 INFO [train.py:903] (3/4) Epoch 7, batch 2150, loss[loss=0.2388, simple_loss=0.2991, pruned_loss=0.08923, over 19419.00 frames. ], tot_loss[loss=0.2632, simple_loss=0.3284, pruned_loss=0.09902, over 3801983.11 frames. ], batch size: 48, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:39:57,668 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43126.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:40:13,322 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43137.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:40:23,158 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.978e+02 6.527e+02 7.673e+02 9.951e+02 2.226e+03, threshold=1.535e+03, percent-clipped=3.0 +2023-04-01 09:40:49,521 INFO [train.py:903] (3/4) Epoch 7, batch 2200, loss[loss=0.2694, simple_loss=0.3411, pruned_loss=0.09881, over 18784.00 frames. ], tot_loss[loss=0.2631, simple_loss=0.3287, pruned_loss=0.09878, over 3815674.55 frames. ], batch size: 74, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:41:53,419 INFO [train.py:903] (3/4) Epoch 7, batch 2250, loss[loss=0.274, simple_loss=0.3508, pruned_loss=0.09865, over 19639.00 frames. ], tot_loss[loss=0.2643, simple_loss=0.3295, pruned_loss=0.09961, over 3813482.01 frames. ], batch size: 60, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:41:57,907 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=43221.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:42:19,766 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43239.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:42:21,867 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=43241.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:42:27,691 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.233e+02 6.206e+02 7.577e+02 9.251e+02 2.641e+03, threshold=1.515e+03, percent-clipped=5.0 +2023-04-01 09:42:52,532 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43264.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:42:56,674 INFO [train.py:903] (3/4) Epoch 7, batch 2300, loss[loss=0.2484, simple_loss=0.3214, pruned_loss=0.08772, over 19537.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3281, pruned_loss=0.09887, over 3820291.26 frames. ], batch size: 54, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:43:03,293 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-01 09:43:10,466 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 09:43:59,264 INFO [train.py:903] (3/4) Epoch 7, batch 2350, loss[loss=0.301, simple_loss=0.3598, pruned_loss=0.1211, over 18754.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3284, pruned_loss=0.09876, over 3809552.62 frames. ], batch size: 74, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:44:22,139 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43336.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:44:32,288 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2563, 1.2721, 1.3970, 1.6192, 2.8091, 0.9972, 1.9305, 3.1072], + device='cuda:3'), covar=tensor([0.0458, 0.2522, 0.2512, 0.1549, 0.0716, 0.2357, 0.1213, 0.0366], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0318, 0.0326, 0.0297, 0.0323, 0.0319, 0.0302, 0.0319], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:44:34,240 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.645e+02 6.151e+02 7.487e+02 9.533e+02 1.563e+03, threshold=1.497e+03, percent-clipped=2.0 +2023-04-01 09:44:43,158 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 09:44:46,762 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43356.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:44:59,569 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 09:45:00,595 INFO [train.py:903] (3/4) Epoch 7, batch 2400, loss[loss=0.2603, simple_loss=0.318, pruned_loss=0.1013, over 19424.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3282, pruned_loss=0.09879, over 3812031.48 frames. ], batch size: 48, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:45:20,057 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43382.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:45:49,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43407.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:46:04,420 INFO [train.py:903] (3/4) Epoch 7, batch 2450, loss[loss=0.2224, simple_loss=0.2983, pruned_loss=0.07328, over 19731.00 frames. ], tot_loss[loss=0.2627, simple_loss=0.3285, pruned_loss=0.09846, over 3811837.19 frames. ], batch size: 51, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:46:38,155 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.412e+02 5.802e+02 7.669e+02 8.855e+02 2.284e+03, threshold=1.534e+03, percent-clipped=5.0 +2023-04-01 09:47:06,558 INFO [train.py:903] (3/4) Epoch 7, batch 2500, loss[loss=0.2496, simple_loss=0.3225, pruned_loss=0.08836, over 19766.00 frames. ], tot_loss[loss=0.2631, simple_loss=0.3287, pruned_loss=0.09873, over 3806261.91 frames. ], batch size: 54, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:47:53,176 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.96 vs. limit=5.0 +2023-04-01 09:48:09,517 INFO [train.py:903] (3/4) Epoch 7, batch 2550, loss[loss=0.266, simple_loss=0.3242, pruned_loss=0.1039, over 19387.00 frames. ], tot_loss[loss=0.2622, simple_loss=0.3282, pruned_loss=0.09812, over 3810595.48 frames. ], batch size: 48, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:48:25,886 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43532.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:48:44,536 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.143e+02 6.070e+02 7.288e+02 8.830e+02 1.707e+03, threshold=1.458e+03, percent-clipped=2.0 +2023-04-01 09:49:02,680 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7080, 3.1761, 3.2403, 3.2423, 1.1653, 2.9941, 2.6769, 2.9463], + device='cuda:3'), covar=tensor([0.1406, 0.0781, 0.0736, 0.0710, 0.4631, 0.0726, 0.0724, 0.1303], + device='cuda:3'), in_proj_covar=tensor([0.0579, 0.0505, 0.0681, 0.0560, 0.0644, 0.0431, 0.0433, 0.0633], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 09:49:05,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 09:49:10,238 INFO [train.py:903] (3/4) Epoch 7, batch 2600, loss[loss=0.2905, simple_loss=0.35, pruned_loss=0.1155, over 18754.00 frames. ], tot_loss[loss=0.2635, simple_loss=0.3293, pruned_loss=0.0988, over 3813721.67 frames. ], batch size: 74, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:49:41,324 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43592.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:49:51,827 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.69 vs. limit=5.0 +2023-04-01 09:50:02,800 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8337, 3.5324, 2.1796, 3.1252, 1.0481, 2.9991, 3.2119, 3.3758], + device='cuda:3'), covar=tensor([0.0595, 0.0915, 0.1964, 0.0694, 0.3447, 0.0863, 0.0681, 0.0731], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0336, 0.0394, 0.0300, 0.0365, 0.0325, 0.0309, 0.0345], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 09:50:06,464 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43612.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:50:09,209 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-01 09:50:12,284 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43617.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:50:12,994 INFO [train.py:903] (3/4) Epoch 7, batch 2650, loss[loss=0.243, simple_loss=0.3107, pruned_loss=0.08768, over 19748.00 frames. ], tot_loss[loss=0.2621, simple_loss=0.3282, pruned_loss=0.09807, over 3816501.83 frames. ], batch size: 51, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:50:23,872 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-01 09:50:35,408 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 09:50:38,365 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43637.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:50:49,495 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.166e+02 6.243e+02 7.316e+02 9.618e+02 1.411e+03, threshold=1.463e+03, percent-clipped=0.0 +2023-04-01 09:51:16,493 INFO [train.py:903] (3/4) Epoch 7, batch 2700, loss[loss=0.2284, simple_loss=0.2977, pruned_loss=0.0795, over 19397.00 frames. ], tot_loss[loss=0.2612, simple_loss=0.327, pruned_loss=0.09769, over 3818580.28 frames. ], batch size: 48, lr: 1.20e-02, grad_scale: 4.0 +2023-04-01 09:51:56,257 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.05 vs. limit=5.0 +2023-04-01 09:52:03,511 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9507, 1.2615, 1.4587, 1.5234, 2.6084, 1.0528, 1.9045, 2.7559], + device='cuda:3'), covar=tensor([0.0478, 0.2470, 0.2370, 0.1515, 0.0719, 0.2275, 0.1176, 0.0439], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0316, 0.0325, 0.0294, 0.0323, 0.0317, 0.0300, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:52:19,230 INFO [train.py:903] (3/4) Epoch 7, batch 2750, loss[loss=0.2967, simple_loss=0.3462, pruned_loss=0.1236, over 13829.00 frames. ], tot_loss[loss=0.2612, simple_loss=0.3269, pruned_loss=0.09777, over 3812310.69 frames. ], batch size: 137, lr: 1.20e-02, grad_scale: 4.0 +2023-04-01 09:52:55,475 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.581e+02 6.869e+02 8.299e+02 1.091e+03 2.331e+03, threshold=1.660e+03, percent-clipped=8.0 +2023-04-01 09:53:20,556 INFO [train.py:903] (3/4) Epoch 7, batch 2800, loss[loss=0.3116, simple_loss=0.3694, pruned_loss=0.1269, over 19568.00 frames. ], tot_loss[loss=0.2617, simple_loss=0.3274, pruned_loss=0.098, over 3815695.43 frames. ], batch size: 61, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:54:07,596 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43805.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:54:22,952 INFO [train.py:903] (3/4) Epoch 7, batch 2850, loss[loss=0.2469, simple_loss=0.3118, pruned_loss=0.09096, over 19746.00 frames. ], tot_loss[loss=0.2622, simple_loss=0.3276, pruned_loss=0.09844, over 3824441.08 frames. ], batch size: 51, lr: 1.20e-02, grad_scale: 8.0 +2023-04-01 09:54:59,106 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.801e+02 6.316e+02 8.520e+02 1.005e+03 1.613e+03, threshold=1.704e+03, percent-clipped=0.0 +2023-04-01 09:55:25,666 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 09:55:26,752 INFO [train.py:903] (3/4) Epoch 7, batch 2900, loss[loss=0.2307, simple_loss=0.3113, pruned_loss=0.07509, over 19521.00 frames. ], tot_loss[loss=0.2621, simple_loss=0.3279, pruned_loss=0.09818, over 3832154.51 frames. ], batch size: 54, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 09:55:36,426 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=43876.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:56:28,084 INFO [train.py:903] (3/4) Epoch 7, batch 2950, loss[loss=0.3309, simple_loss=0.3777, pruned_loss=0.1421, over 17318.00 frames. ], tot_loss[loss=0.2611, simple_loss=0.327, pruned_loss=0.09759, over 3838298.68 frames. ], batch size: 101, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 09:56:39,834 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43927.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:57:04,993 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.608e+02 5.923e+02 7.218e+02 9.278e+02 2.092e+03, threshold=1.444e+03, percent-clipped=1.0 +2023-04-01 09:57:25,768 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 09:57:30,518 INFO [train.py:903] (3/4) Epoch 7, batch 3000, loss[loss=0.2256, simple_loss=0.3104, pruned_loss=0.07037, over 19665.00 frames. ], tot_loss[loss=0.2597, simple_loss=0.3258, pruned_loss=0.09676, over 3845105.80 frames. ], batch size: 55, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 09:57:30,518 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 09:57:43,097 INFO [train.py:937] (3/4) Epoch 7, validation: loss=0.1917, simple_loss=0.2919, pruned_loss=0.04574, over 944034.00 frames. +2023-04-01 09:57:43,098 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 09:57:43,657 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7607, 1.8400, 1.8984, 2.6049, 1.6526, 2.2966, 2.3947, 1.8900], + device='cuda:3'), covar=tensor([0.2538, 0.2294, 0.1086, 0.1049, 0.2254, 0.0924, 0.2222, 0.1945], + device='cuda:3'), in_proj_covar=tensor([0.0700, 0.0704, 0.0596, 0.0834, 0.0719, 0.0613, 0.0733, 0.0639], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 09:57:49,866 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 09:57:54,981 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43977.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:57:58,832 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7791, 1.3949, 1.3444, 1.7913, 1.5782, 1.5480, 1.5219, 1.7195], + device='cuda:3'), covar=tensor([0.0883, 0.1569, 0.1420, 0.0905, 0.1141, 0.0538, 0.0997, 0.0640], + device='cuda:3'), in_proj_covar=tensor([0.0244, 0.0358, 0.0286, 0.0236, 0.0298, 0.0247, 0.0268, 0.0231], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 09:58:13,816 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43991.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 09:58:18,950 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8808, 1.9418, 1.9428, 2.9026, 1.8382, 2.6120, 2.5572, 1.9205], + device='cuda:3'), covar=tensor([0.2727, 0.2293, 0.1144, 0.1213, 0.2601, 0.0966, 0.2167, 0.1973], + device='cuda:3'), in_proj_covar=tensor([0.0694, 0.0701, 0.0594, 0.0834, 0.0716, 0.0611, 0.0730, 0.0637], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 09:58:31,227 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9750, 1.0315, 1.4236, 0.5405, 2.2398, 2.4031, 2.1617, 2.6173], + device='cuda:3'), covar=tensor([0.1474, 0.3395, 0.2907, 0.2212, 0.0440, 0.0233, 0.0383, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0284, 0.0313, 0.0245, 0.0203, 0.0133, 0.0204, 0.0162], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 09:58:46,650 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2275, 1.3078, 1.8193, 1.4426, 2.6906, 2.1410, 2.7308, 1.0728], + device='cuda:3'), covar=tensor([0.1990, 0.3426, 0.1855, 0.1585, 0.1236, 0.1680, 0.1390, 0.3161], + device='cuda:3'), in_proj_covar=tensor([0.0449, 0.0519, 0.0515, 0.0405, 0.0557, 0.0451, 0.0624, 0.0453], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 09:58:47,331 INFO [train.py:903] (3/4) Epoch 7, batch 3050, loss[loss=0.238, simple_loss=0.3133, pruned_loss=0.08132, over 19275.00 frames. ], tot_loss[loss=0.2605, simple_loss=0.3267, pruned_loss=0.09722, over 3848254.71 frames. ], batch size: 66, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 09:59:24,084 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.748e+02 5.923e+02 7.306e+02 8.819e+02 1.422e+03, threshold=1.461e+03, percent-clipped=0.0 +2023-04-01 09:59:50,239 INFO [train.py:903] (3/4) Epoch 7, batch 3100, loss[loss=0.2911, simple_loss=0.3524, pruned_loss=0.1149, over 19582.00 frames. ], tot_loss[loss=0.2594, simple_loss=0.3259, pruned_loss=0.09647, over 3851087.75 frames. ], batch size: 61, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:00:50,926 INFO [train.py:903] (3/4) Epoch 7, batch 3150, loss[loss=0.2379, simple_loss=0.2958, pruned_loss=0.08997, over 19486.00 frames. ], tot_loss[loss=0.2607, simple_loss=0.3266, pruned_loss=0.09741, over 3844128.53 frames. ], batch size: 49, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:00:53,389 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0829, 1.1721, 1.5959, 0.8917, 2.3774, 2.9723, 2.7298, 3.1530], + device='cuda:3'), covar=tensor([0.1484, 0.3176, 0.2771, 0.2191, 0.0460, 0.0191, 0.0272, 0.0179], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0286, 0.0316, 0.0250, 0.0205, 0.0134, 0.0207, 0.0163], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 10:00:56,512 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2083, 1.1821, 1.8529, 1.5964, 2.8967, 4.1959, 4.1882, 4.6340], + device='cuda:3'), covar=tensor([0.1592, 0.3309, 0.2832, 0.1808, 0.0510, 0.0194, 0.0182, 0.0107], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0286, 0.0316, 0.0250, 0.0205, 0.0134, 0.0207, 0.0163], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 10:01:18,567 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 10:01:26,087 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.269e+02 6.470e+02 8.010e+02 1.018e+03 2.357e+03, threshold=1.602e+03, percent-clipped=4.0 +2023-04-01 10:01:29,402 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=44149.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:01:51,333 INFO [train.py:903] (3/4) Epoch 7, batch 3200, loss[loss=0.2544, simple_loss=0.3289, pruned_loss=0.08991, over 19523.00 frames. ], tot_loss[loss=0.2607, simple_loss=0.3266, pruned_loss=0.09736, over 3842384.17 frames. ], batch size: 54, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:02:51,383 INFO [train.py:903] (3/4) Epoch 7, batch 3250, loss[loss=0.2707, simple_loss=0.3418, pruned_loss=0.09981, over 19601.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3288, pruned_loss=0.09852, over 3834080.44 frames. ], batch size: 57, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:03:05,873 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 10:03:25,181 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-01 10:03:27,995 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.707e+02 6.408e+02 8.261e+02 1.024e+03 1.757e+03, threshold=1.652e+03, percent-clipped=4.0 +2023-04-01 10:03:28,442 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44247.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:03:49,224 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44264.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:03:53,343 INFO [train.py:903] (3/4) Epoch 7, batch 3300, loss[loss=0.2471, simple_loss=0.3185, pruned_loss=0.08788, over 19600.00 frames. ], tot_loss[loss=0.2624, simple_loss=0.328, pruned_loss=0.09843, over 3821227.54 frames. ], batch size: 52, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:03:59,186 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=44271.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:04:00,238 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 10:04:00,641 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44272.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:04:57,179 INFO [train.py:903] (3/4) Epoch 7, batch 3350, loss[loss=0.2585, simple_loss=0.3276, pruned_loss=0.09476, over 19661.00 frames. ], tot_loss[loss=0.263, simple_loss=0.3285, pruned_loss=0.09876, over 3829074.28 frames. ], batch size: 55, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:05:01,676 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=44321.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:05:32,247 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.423e+02 6.124e+02 7.865e+02 1.014e+03 2.362e+03, threshold=1.573e+03, percent-clipped=3.0 +2023-04-01 10:05:58,464 INFO [train.py:903] (3/4) Epoch 7, batch 3400, loss[loss=0.2956, simple_loss=0.3604, pruned_loss=0.1154, over 19478.00 frames. ], tot_loss[loss=0.2619, simple_loss=0.3274, pruned_loss=0.09817, over 3831533.59 frames. ], batch size: 64, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:06:20,848 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44386.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:07:01,357 INFO [train.py:903] (3/4) Epoch 7, batch 3450, loss[loss=0.1982, simple_loss=0.2683, pruned_loss=0.06408, over 18783.00 frames. ], tot_loss[loss=0.2614, simple_loss=0.3269, pruned_loss=0.09798, over 3830003.01 frames. ], batch size: 41, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:07:06,915 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 10:07:08,515 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2308, 1.2409, 1.6311, 1.4593, 2.4264, 2.0681, 2.6008, 1.0555], + device='cuda:3'), covar=tensor([0.2207, 0.3771, 0.2040, 0.1796, 0.1335, 0.1837, 0.1399, 0.3188], + device='cuda:3'), in_proj_covar=tensor([0.0458, 0.0529, 0.0522, 0.0414, 0.0572, 0.0463, 0.0638, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 10:07:25,050 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44436.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:07:39,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.210e+02 5.624e+02 7.209e+02 9.228e+02 1.461e+03, threshold=1.442e+03, percent-clipped=0.0 +2023-04-01 10:08:03,657 INFO [train.py:903] (3/4) Epoch 7, batch 3500, loss[loss=0.2012, simple_loss=0.2691, pruned_loss=0.06665, over 19755.00 frames. ], tot_loss[loss=0.2624, simple_loss=0.328, pruned_loss=0.09836, over 3828871.90 frames. ], batch size: 46, lr: 1.19e-02, grad_scale: 4.0 +2023-04-01 10:09:07,866 INFO [train.py:903] (3/4) Epoch 7, batch 3550, loss[loss=0.286, simple_loss=0.3538, pruned_loss=0.1091, over 19602.00 frames. ], tot_loss[loss=0.2644, simple_loss=0.3299, pruned_loss=0.09948, over 3820967.22 frames. ], batch size: 61, lr: 1.19e-02, grad_scale: 4.0 +2023-04-01 10:09:10,474 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44520.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:09:37,863 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-01 10:09:40,619 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44545.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:09:40,732 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44545.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:09:43,701 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.445e+02 5.736e+02 7.487e+02 9.426e+02 2.431e+03, threshold=1.497e+03, percent-clipped=5.0 +2023-04-01 10:10:10,018 INFO [train.py:903] (3/4) Epoch 7, batch 3600, loss[loss=0.3166, simple_loss=0.3664, pruned_loss=0.1334, over 13405.00 frames. ], tot_loss[loss=0.2652, simple_loss=0.3301, pruned_loss=0.1001, over 3804605.13 frames. ], batch size: 136, lr: 1.19e-02, grad_scale: 8.0 +2023-04-01 10:10:10,278 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44568.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:11:12,085 INFO [train.py:903] (3/4) Epoch 7, batch 3650, loss[loss=0.2465, simple_loss=0.3221, pruned_loss=0.08542, over 19658.00 frames. ], tot_loss[loss=0.2651, simple_loss=0.33, pruned_loss=0.1001, over 3795291.33 frames. ], batch size: 60, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:11:16,252 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.76 vs. limit=5.0 +2023-04-01 10:11:43,265 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44642.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:11:49,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.538e+02 6.222e+02 7.769e+02 9.647e+02 2.431e+03, threshold=1.554e+03, percent-clipped=4.0 +2023-04-01 10:12:12,918 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44667.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:12:13,723 INFO [train.py:903] (3/4) Epoch 7, batch 3700, loss[loss=0.2316, simple_loss=0.3107, pruned_loss=0.07622, over 19736.00 frames. ], tot_loss[loss=0.2639, simple_loss=0.329, pruned_loss=0.0994, over 3803788.21 frames. ], batch size: 63, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:12:45,621 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44692.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:12:59,043 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7622, 1.6319, 1.2833, 1.7199, 1.7989, 1.4622, 1.3383, 1.5828], + device='cuda:3'), covar=tensor([0.0965, 0.1671, 0.1796, 0.1083, 0.1444, 0.1048, 0.1439, 0.1020], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0358, 0.0288, 0.0236, 0.0303, 0.0244, 0.0270, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 10:13:16,193 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44717.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:13:16,967 INFO [train.py:903] (3/4) Epoch 7, batch 3750, loss[loss=0.2613, simple_loss=0.3359, pruned_loss=0.09341, over 19651.00 frames. ], tot_loss[loss=0.2648, simple_loss=0.3298, pruned_loss=0.09987, over 3802710.45 frames. ], batch size: 55, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:13:53,509 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.417e+02 6.149e+02 7.382e+02 9.468e+02 1.650e+03, threshold=1.476e+03, percent-clipped=2.0 +2023-04-01 10:14:18,149 INFO [train.py:903] (3/4) Epoch 7, batch 3800, loss[loss=0.2246, simple_loss=0.2921, pruned_loss=0.0786, over 19787.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3281, pruned_loss=0.0989, over 3813545.40 frames. ], batch size: 49, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:14:50,705 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 10:15:19,313 INFO [train.py:903] (3/4) Epoch 7, batch 3850, loss[loss=0.2735, simple_loss=0.3477, pruned_loss=0.09967, over 19377.00 frames. ], tot_loss[loss=0.2628, simple_loss=0.3284, pruned_loss=0.09857, over 3812248.17 frames. ], batch size: 70, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:15:57,068 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.826e+02 6.658e+02 8.285e+02 1.095e+03 3.075e+03, threshold=1.657e+03, percent-clipped=10.0 +2023-04-01 10:16:20,985 INFO [train.py:903] (3/4) Epoch 7, batch 3900, loss[loss=0.291, simple_loss=0.3499, pruned_loss=0.116, over 19786.00 frames. ], tot_loss[loss=0.2622, simple_loss=0.3276, pruned_loss=0.09841, over 3813636.73 frames. ], batch size: 56, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:16:48,347 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=44889.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:17:15,639 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=44912.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:17:24,388 INFO [train.py:903] (3/4) Epoch 7, batch 3950, loss[loss=0.257, simple_loss=0.3157, pruned_loss=0.09917, over 19723.00 frames. ], tot_loss[loss=0.2615, simple_loss=0.3272, pruned_loss=0.09791, over 3816449.26 frames. ], batch size: 51, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:17:29,135 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 10:18:00,740 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.332e+02 5.655e+02 7.377e+02 9.527e+02 2.304e+03, threshold=1.475e+03, percent-clipped=3.0 +2023-04-01 10:18:26,675 INFO [train.py:903] (3/4) Epoch 7, batch 4000, loss[loss=0.2598, simple_loss=0.3323, pruned_loss=0.09364, over 19579.00 frames. ], tot_loss[loss=0.2618, simple_loss=0.3278, pruned_loss=0.09793, over 3818841.39 frames. ], batch size: 61, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:19:11,998 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45004.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:19:15,206 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 10:19:22,027 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-04-01 10:19:27,786 INFO [train.py:903] (3/4) Epoch 7, batch 4050, loss[loss=0.2621, simple_loss=0.3312, pruned_loss=0.09656, over 18336.00 frames. ], tot_loss[loss=0.261, simple_loss=0.3274, pruned_loss=0.09732, over 3819761.71 frames. ], batch size: 84, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:19:39,412 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45027.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:20:05,076 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.186e+02 6.019e+02 6.941e+02 8.524e+02 1.564e+03, threshold=1.388e+03, percent-clipped=1.0 +2023-04-01 10:20:28,972 INFO [train.py:903] (3/4) Epoch 7, batch 4100, loss[loss=0.2829, simple_loss=0.3503, pruned_loss=0.1077, over 17284.00 frames. ], tot_loss[loss=0.2607, simple_loss=0.3268, pruned_loss=0.09724, over 3826195.67 frames. ], batch size: 101, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:20:36,306 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45074.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:20:54,203 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8476, 4.2317, 4.5241, 4.4797, 1.7518, 4.1558, 3.7131, 4.1159], + device='cuda:3'), covar=tensor([0.1108, 0.0699, 0.0505, 0.0458, 0.4229, 0.0468, 0.0526, 0.1025], + device='cuda:3'), in_proj_covar=tensor([0.0589, 0.0512, 0.0696, 0.0570, 0.0647, 0.0437, 0.0441, 0.0650], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 10:21:05,613 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 10:21:07,164 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45099.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:21:30,596 INFO [train.py:903] (3/4) Epoch 7, batch 4150, loss[loss=0.2188, simple_loss=0.2826, pruned_loss=0.0775, over 19751.00 frames. ], tot_loss[loss=0.2619, simple_loss=0.3275, pruned_loss=0.09816, over 3827442.65 frames. ], batch size: 46, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:21:41,539 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-01 10:22:07,171 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.699e+02 6.430e+02 8.235e+02 1.004e+03 2.607e+03, threshold=1.647e+03, percent-clipped=6.0 +2023-04-01 10:22:32,968 INFO [train.py:903] (3/4) Epoch 7, batch 4200, loss[loss=0.2421, simple_loss=0.3039, pruned_loss=0.09014, over 19629.00 frames. ], tot_loss[loss=0.2616, simple_loss=0.3271, pruned_loss=0.09806, over 3817766.28 frames. ], batch size: 50, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:22:38,329 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 10:23:33,351 INFO [train.py:903] (3/4) Epoch 7, batch 4250, loss[loss=0.2138, simple_loss=0.2808, pruned_loss=0.07338, over 19089.00 frames. ], tot_loss[loss=0.2612, simple_loss=0.3268, pruned_loss=0.09784, over 3816865.81 frames. ], batch size: 42, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:23:49,621 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 10:23:53,861 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-04-01 10:24:01,333 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 10:24:12,489 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.934e+02 6.110e+02 8.094e+02 1.072e+03 2.309e+03, threshold=1.619e+03, percent-clipped=5.0 +2023-04-01 10:24:27,363 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45260.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:24:37,358 INFO [train.py:903] (3/4) Epoch 7, batch 4300, loss[loss=0.2615, simple_loss=0.3321, pruned_loss=0.09546, over 19530.00 frames. ], tot_loss[loss=0.2609, simple_loss=0.3269, pruned_loss=0.09743, over 3821595.67 frames. ], batch size: 54, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:24:57,141 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45283.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:25:00,280 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45285.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:25:03,710 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4479, 1.1578, 1.0969, 1.3839, 1.1870, 1.2870, 1.1918, 1.3470], + device='cuda:3'), covar=tensor([0.0956, 0.1213, 0.1407, 0.0868, 0.0955, 0.0545, 0.1054, 0.0691], + device='cuda:3'), in_proj_covar=tensor([0.0244, 0.0350, 0.0282, 0.0234, 0.0297, 0.0241, 0.0265, 0.0228], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 10:25:27,074 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45308.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:25:32,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 10:25:39,641 INFO [train.py:903] (3/4) Epoch 7, batch 4350, loss[loss=0.2919, simple_loss=0.3508, pruned_loss=0.1165, over 19405.00 frames. ], tot_loss[loss=0.2617, simple_loss=0.3274, pruned_loss=0.09794, over 3823411.32 frames. ], batch size: 70, lr: 1.18e-02, grad_scale: 8.0 +2023-04-01 10:26:16,052 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.214e+02 6.479e+02 7.493e+02 1.002e+03 2.532e+03, threshold=1.499e+03, percent-clipped=5.0 +2023-04-01 10:26:34,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2787, 3.0798, 1.8038, 2.3796, 1.7905, 2.3234, 0.8057, 2.1093], + device='cuda:3'), covar=tensor([0.0413, 0.0400, 0.0497, 0.0652, 0.0835, 0.0801, 0.0845, 0.0750], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0313, 0.0302, 0.0333, 0.0408, 0.0325, 0.0289, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 10:26:39,945 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45366.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 10:26:42,860 INFO [train.py:903] (3/4) Epoch 7, batch 4400, loss[loss=0.2339, simple_loss=0.3054, pruned_loss=0.08117, over 19689.00 frames. ], tot_loss[loss=0.2629, simple_loss=0.3284, pruned_loss=0.09866, over 3817085.43 frames. ], batch size: 53, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:26:48,513 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-04-01 10:27:06,062 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 10:27:14,901 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 10:27:44,672 INFO [train.py:903] (3/4) Epoch 7, batch 4450, loss[loss=0.2199, simple_loss=0.2905, pruned_loss=0.07465, over 19425.00 frames. ], tot_loss[loss=0.2615, simple_loss=0.3272, pruned_loss=0.0979, over 3829022.00 frames. ], batch size: 48, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:27:44,841 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=45418.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:28:16,084 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=45443.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:28:22,922 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.630e+02 6.151e+02 7.769e+02 9.586e+02 4.695e+03, threshold=1.554e+03, percent-clipped=4.0 +2023-04-01 10:28:46,110 INFO [train.py:903] (3/4) Epoch 7, batch 4500, loss[loss=0.304, simple_loss=0.3649, pruned_loss=0.1216, over 19662.00 frames. ], tot_loss[loss=0.2617, simple_loss=0.3273, pruned_loss=0.09811, over 3823307.71 frames. ], batch size: 60, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:29:09,233 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.67 vs. limit=5.0 +2023-04-01 10:29:48,542 INFO [train.py:903] (3/4) Epoch 7, batch 4550, loss[loss=0.2995, simple_loss=0.3579, pruned_loss=0.1205, over 19050.00 frames. ], tot_loss[loss=0.261, simple_loss=0.3264, pruned_loss=0.09776, over 3841467.89 frames. ], batch size: 75, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:30:00,876 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 10:30:09,149 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45533.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:30:23,718 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 10:30:27,041 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.783e+02 6.251e+02 7.662e+02 9.726e+02 2.453e+03, threshold=1.532e+03, percent-clipped=6.0 +2023-04-01 10:30:38,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45558.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:30:51,041 INFO [train.py:903] (3/4) Epoch 7, batch 4600, loss[loss=0.2429, simple_loss=0.3173, pruned_loss=0.08426, over 19657.00 frames. ], tot_loss[loss=0.2623, simple_loss=0.3276, pruned_loss=0.0985, over 3824957.03 frames. ], batch size: 55, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:31:54,318 INFO [train.py:903] (3/4) Epoch 7, batch 4650, loss[loss=0.2182, simple_loss=0.2926, pruned_loss=0.07193, over 19396.00 frames. ], tot_loss[loss=0.2603, simple_loss=0.3265, pruned_loss=0.09703, over 3831204.08 frames. ], batch size: 48, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:32:08,739 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3605, 1.4800, 2.0474, 1.5407, 3.2543, 2.7363, 3.5292, 1.4656], + device='cuda:3'), covar=tensor([0.1967, 0.3488, 0.1998, 0.1603, 0.1331, 0.1536, 0.1401, 0.3130], + device='cuda:3'), in_proj_covar=tensor([0.0451, 0.0530, 0.0517, 0.0412, 0.0564, 0.0458, 0.0625, 0.0453], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 10:32:11,395 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-04-01 10:32:11,865 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 10:32:24,512 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 10:32:33,274 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.198e+02 5.859e+02 7.403e+02 8.847e+02 2.429e+03, threshold=1.481e+03, percent-clipped=2.0 +2023-04-01 10:32:55,952 INFO [train.py:903] (3/4) Epoch 7, batch 4700, loss[loss=0.2043, simple_loss=0.2784, pruned_loss=0.06515, over 19775.00 frames. ], tot_loss[loss=0.2609, simple_loss=0.3269, pruned_loss=0.09738, over 3810246.06 frames. ], batch size: 47, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:33:14,950 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4239, 1.5521, 2.0541, 1.6294, 3.2281, 2.8288, 3.5037, 1.4577], + device='cuda:3'), covar=tensor([0.1934, 0.3265, 0.2015, 0.1532, 0.1398, 0.1458, 0.1406, 0.3024], + device='cuda:3'), in_proj_covar=tensor([0.0460, 0.0537, 0.0525, 0.0418, 0.0573, 0.0465, 0.0633, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 10:33:18,900 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 10:33:47,817 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=45710.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 10:33:51,389 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2345, 1.3226, 1.7933, 1.7127, 3.0165, 4.5824, 4.5635, 5.0482], + device='cuda:3'), covar=tensor([0.1504, 0.2948, 0.2746, 0.1729, 0.0466, 0.0163, 0.0154, 0.0087], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0285, 0.0316, 0.0246, 0.0205, 0.0135, 0.0204, 0.0166], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 10:33:58,624 INFO [train.py:903] (3/4) Epoch 7, batch 4750, loss[loss=0.3292, simple_loss=0.3831, pruned_loss=0.1377, over 19736.00 frames. ], tot_loss[loss=0.2614, simple_loss=0.3273, pruned_loss=0.09777, over 3810044.75 frames. ], batch size: 63, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:34:01,234 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45720.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:34:36,017 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.623e+02 6.576e+02 8.008e+02 9.322e+02 2.457e+03, threshold=1.602e+03, percent-clipped=6.0 +2023-04-01 10:35:01,472 INFO [train.py:903] (3/4) Epoch 7, batch 4800, loss[loss=0.282, simple_loss=0.3414, pruned_loss=0.1114, over 19682.00 frames. ], tot_loss[loss=0.261, simple_loss=0.327, pruned_loss=0.0975, over 3807302.08 frames. ], batch size: 53, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:35:19,785 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-01 10:35:26,502 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45789.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:35:29,646 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45792.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:35:58,127 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45814.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:35:58,157 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45814.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:36:02,102 INFO [train.py:903] (3/4) Epoch 7, batch 4850, loss[loss=0.2598, simple_loss=0.3253, pruned_loss=0.09721, over 19533.00 frames. ], tot_loss[loss=0.2623, simple_loss=0.3279, pruned_loss=0.09832, over 3807979.00 frames. ], batch size: 56, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:36:10,624 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45825.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 10:36:25,112 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 10:36:27,424 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45839.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:36:38,173 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-01 10:36:40,693 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.989e+02 6.823e+02 8.718e+02 1.115e+03 2.265e+03, threshold=1.744e+03, percent-clipped=6.0 +2023-04-01 10:36:43,213 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6624, 1.3494, 1.4263, 1.9317, 1.3575, 1.8156, 1.8441, 1.8303], + device='cuda:3'), covar=tensor([0.0762, 0.1014, 0.1009, 0.0859, 0.0970, 0.0711, 0.0917, 0.0620], + device='cuda:3'), in_proj_covar=tensor([0.0219, 0.0238, 0.0233, 0.0265, 0.0254, 0.0220, 0.0219, 0.0215], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005, 0.0004], + device='cuda:3') +2023-04-01 10:36:48,782 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 10:36:54,244 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 10:36:54,266 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 10:37:03,557 INFO [train.py:903] (3/4) Epoch 7, batch 4900, loss[loss=0.2243, simple_loss=0.2924, pruned_loss=0.0781, over 19392.00 frames. ], tot_loss[loss=0.2619, simple_loss=0.3275, pruned_loss=0.09812, over 3819349.28 frames. ], batch size: 48, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:37:04,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 10:37:14,481 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45877.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:37:19,166 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.11 vs. limit=2.0 +2023-04-01 10:37:26,126 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 10:37:31,690 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-04-01 10:38:05,107 INFO [train.py:903] (3/4) Epoch 7, batch 4950, loss[loss=0.3168, simple_loss=0.3704, pruned_loss=0.1316, over 17343.00 frames. ], tot_loss[loss=0.2605, simple_loss=0.3267, pruned_loss=0.09709, over 3809831.01 frames. ], batch size: 101, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:38:24,761 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 10:38:44,269 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.370e+02 6.277e+02 7.738e+02 9.356e+02 2.304e+03, threshold=1.548e+03, percent-clipped=1.0 +2023-04-01 10:38:47,738 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 10:39:09,684 INFO [train.py:903] (3/4) Epoch 7, batch 5000, loss[loss=0.2329, simple_loss=0.3108, pruned_loss=0.07757, over 19597.00 frames. ], tot_loss[loss=0.2607, simple_loss=0.327, pruned_loss=0.09717, over 3820765.71 frames. ], batch size: 61, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:39:20,561 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 10:39:30,931 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 10:40:12,335 INFO [train.py:903] (3/4) Epoch 7, batch 5050, loss[loss=0.2425, simple_loss=0.3158, pruned_loss=0.08463, over 19763.00 frames. ], tot_loss[loss=0.2601, simple_loss=0.3267, pruned_loss=0.09677, over 3816265.52 frames. ], batch size: 54, lr: 1.17e-02, grad_scale: 8.0 +2023-04-01 10:40:49,620 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 10:40:51,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.897e+02 6.208e+02 7.894e+02 9.516e+02 2.052e+03, threshold=1.579e+03, percent-clipped=3.0 +2023-04-01 10:40:58,131 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 10:41:03,344 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5435, 1.5950, 1.6615, 2.0518, 1.3396, 1.6723, 2.0865, 1.5687], + device='cuda:3'), covar=tensor([0.2307, 0.1681, 0.1051, 0.1010, 0.1898, 0.0998, 0.2188, 0.1863], + device='cuda:3'), in_proj_covar=tensor([0.0710, 0.0712, 0.0604, 0.0854, 0.0724, 0.0628, 0.0746, 0.0651], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 10:41:08,568 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46064.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:41:13,139 INFO [train.py:903] (3/4) Epoch 7, batch 5100, loss[loss=0.2283, simple_loss=0.3036, pruned_loss=0.07646, over 19656.00 frames. ], tot_loss[loss=0.2603, simple_loss=0.3267, pruned_loss=0.097, over 3808652.05 frames. ], batch size: 55, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:41:24,899 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 10:41:28,149 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 10:41:28,568 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46081.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 10:41:32,522 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 10:42:00,060 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46106.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 10:42:13,344 INFO [train.py:903] (3/4) Epoch 7, batch 5150, loss[loss=0.2401, simple_loss=0.2987, pruned_loss=0.09072, over 19740.00 frames. ], tot_loss[loss=0.26, simple_loss=0.3264, pruned_loss=0.09683, over 3810571.53 frames. ], batch size: 47, lr: 1.17e-02, grad_scale: 4.0 +2023-04-01 10:42:26,334 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 10:42:37,815 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46136.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:42:53,640 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.315e+02 6.102e+02 7.228e+02 8.853e+02 1.806e+03, threshold=1.446e+03, percent-clipped=2.0 +2023-04-01 10:43:00,542 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 10:43:15,665 INFO [train.py:903] (3/4) Epoch 7, batch 5200, loss[loss=0.2964, simple_loss=0.3719, pruned_loss=0.1105, over 18311.00 frames. ], tot_loss[loss=0.26, simple_loss=0.3265, pruned_loss=0.09678, over 3808288.67 frames. ], batch size: 83, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:43:30,896 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46179.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:43:31,689 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 10:44:18,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 10:44:19,700 INFO [train.py:903] (3/4) Epoch 7, batch 5250, loss[loss=0.2882, simple_loss=0.353, pruned_loss=0.1117, over 19392.00 frames. ], tot_loss[loss=0.2603, simple_loss=0.3271, pruned_loss=0.09675, over 3813935.87 frames. ], batch size: 70, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:44:23,438 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46221.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:44:32,721 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3726, 1.5051, 1.8113, 1.3805, 2.4027, 2.6481, 2.6089, 2.8230], + device='cuda:3'), covar=tensor([0.1204, 0.2245, 0.2105, 0.1880, 0.0752, 0.0577, 0.0222, 0.0204], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0284, 0.0316, 0.0245, 0.0203, 0.0133, 0.0202, 0.0165], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 10:44:58,428 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.945e+02 5.998e+02 7.245e+02 8.826e+02 1.462e+03, threshold=1.449e+03, percent-clipped=1.0 +2023-04-01 10:45:00,951 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46251.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:45:21,290 INFO [train.py:903] (3/4) Epoch 7, batch 5300, loss[loss=0.255, simple_loss=0.3219, pruned_loss=0.09405, over 19470.00 frames. ], tot_loss[loss=0.2607, simple_loss=0.3275, pruned_loss=0.0969, over 3811466.93 frames. ], batch size: 49, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:45:39,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 10:45:54,314 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46294.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:46:22,054 INFO [train.py:903] (3/4) Epoch 7, batch 5350, loss[loss=0.2377, simple_loss=0.2983, pruned_loss=0.08857, over 19726.00 frames. ], tot_loss[loss=0.2611, simple_loss=0.328, pruned_loss=0.09713, over 3825308.93 frames. ], batch size: 45, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:46:47,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46336.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:47:00,316 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 10:47:03,769 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.963e+02 5.681e+02 7.157e+02 9.578e+02 3.754e+03, threshold=1.431e+03, percent-clipped=4.0 +2023-04-01 10:47:26,532 INFO [train.py:903] (3/4) Epoch 7, batch 5400, loss[loss=0.2042, simple_loss=0.276, pruned_loss=0.06625, over 19724.00 frames. ], tot_loss[loss=0.2596, simple_loss=0.3264, pruned_loss=0.09643, over 3822885.16 frames. ], batch size: 51, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:47:58,983 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2195, 1.1859, 1.3935, 1.2948, 1.7833, 1.7808, 1.9486, 0.5688], + device='cuda:3'), covar=tensor([0.1830, 0.2947, 0.1705, 0.1585, 0.1077, 0.1608, 0.0959, 0.2841], + device='cuda:3'), in_proj_covar=tensor([0.0454, 0.0530, 0.0521, 0.0414, 0.0569, 0.0459, 0.0636, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 10:48:29,980 INFO [train.py:903] (3/4) Epoch 7, batch 5450, loss[loss=0.2838, simple_loss=0.347, pruned_loss=0.1103, over 19778.00 frames. ], tot_loss[loss=0.2612, simple_loss=0.3274, pruned_loss=0.09754, over 3824012.16 frames. ], batch size: 54, lr: 1.16e-02, grad_scale: 4.0 +2023-04-01 10:48:49,860 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46435.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:49:10,731 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.600e+02 7.102e+02 8.361e+02 1.036e+03 2.875e+03, threshold=1.672e+03, percent-clipped=7.0 +2023-04-01 10:49:23,157 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46460.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:49:31,707 INFO [train.py:903] (3/4) Epoch 7, batch 5500, loss[loss=0.2664, simple_loss=0.337, pruned_loss=0.0979, over 19550.00 frames. ], tot_loss[loss=0.2613, simple_loss=0.3276, pruned_loss=0.09757, over 3832371.91 frames. ], batch size: 61, lr: 1.16e-02, grad_scale: 4.0 +2023-04-01 10:49:58,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 10:50:20,878 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46507.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:50:33,173 INFO [train.py:903] (3/4) Epoch 7, batch 5550, loss[loss=0.2473, simple_loss=0.3213, pruned_loss=0.08661, over 19687.00 frames. ], tot_loss[loss=0.2603, simple_loss=0.3273, pruned_loss=0.09668, over 3832189.05 frames. ], batch size: 60, lr: 1.16e-02, grad_scale: 4.0 +2023-04-01 10:50:43,598 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 10:50:51,465 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46532.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:51:15,056 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.541e+02 5.815e+02 7.044e+02 9.146e+02 3.032e+03, threshold=1.409e+03, percent-clipped=3.0 +2023-04-01 10:51:21,515 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.10 vs. limit=2.0 +2023-04-01 10:51:30,166 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6862, 1.3082, 1.4139, 1.7597, 3.1530, 1.0979, 2.2483, 3.3673], + device='cuda:3'), covar=tensor([0.0372, 0.2638, 0.2689, 0.1558, 0.0705, 0.2451, 0.1137, 0.0362], + device='cuda:3'), in_proj_covar=tensor([0.0315, 0.0319, 0.0329, 0.0299, 0.0328, 0.0322, 0.0297, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 10:51:31,092 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 10:51:35,755 INFO [train.py:903] (3/4) Epoch 7, batch 5600, loss[loss=0.2489, simple_loss=0.3198, pruned_loss=0.08901, over 19656.00 frames. ], tot_loss[loss=0.2608, simple_loss=0.3278, pruned_loss=0.09688, over 3829846.04 frames. ], batch size: 60, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:52:06,750 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46592.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:52:32,980 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46613.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:52:39,566 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46617.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:52:40,323 INFO [train.py:903] (3/4) Epoch 7, batch 5650, loss[loss=0.212, simple_loss=0.2818, pruned_loss=0.07115, over 19468.00 frames. ], tot_loss[loss=0.2603, simple_loss=0.3275, pruned_loss=0.09652, over 3832297.56 frames. ], batch size: 49, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:52:42,901 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46620.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:53:03,553 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46638.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:53:12,134 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-01 10:53:20,361 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.719e+02 5.976e+02 7.674e+02 9.559e+02 1.706e+03, threshold=1.535e+03, percent-clipped=4.0 +2023-04-01 10:53:29,236 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 10:53:42,092 INFO [train.py:903] (3/4) Epoch 7, batch 5700, loss[loss=0.303, simple_loss=0.3691, pruned_loss=0.1185, over 19672.00 frames. ], tot_loss[loss=0.2583, simple_loss=0.3256, pruned_loss=0.09547, over 3827280.29 frames. ], batch size: 58, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:53:58,471 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4852, 1.0728, 1.2723, 1.1114, 2.1222, 0.8629, 1.7514, 2.2462], + device='cuda:3'), covar=tensor([0.0635, 0.2532, 0.2424, 0.1591, 0.0868, 0.2097, 0.0987, 0.0505], + device='cuda:3'), in_proj_covar=tensor([0.0314, 0.0316, 0.0325, 0.0297, 0.0326, 0.0321, 0.0296, 0.0315], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 10:54:43,277 INFO [train.py:903] (3/4) Epoch 7, batch 5750, loss[loss=0.2439, simple_loss=0.3244, pruned_loss=0.08171, over 19652.00 frames. ], tot_loss[loss=0.2578, simple_loss=0.325, pruned_loss=0.09529, over 3822899.76 frames. ], batch size: 58, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:54:45,649 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 10:54:55,172 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 10:54:59,793 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 10:55:01,307 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-01 10:55:25,496 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.168e+02 6.148e+02 7.021e+02 8.191e+02 1.564e+03, threshold=1.404e+03, percent-clipped=1.0 +2023-04-01 10:55:28,135 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46753.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:55:45,400 INFO [train.py:903] (3/4) Epoch 7, batch 5800, loss[loss=0.2534, simple_loss=0.3257, pruned_loss=0.09058, over 19539.00 frames. ], tot_loss[loss=0.2573, simple_loss=0.3243, pruned_loss=0.09512, over 3811809.15 frames. ], batch size: 54, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:55:47,400 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 10:56:49,194 INFO [train.py:903] (3/4) Epoch 7, batch 5850, loss[loss=0.2641, simple_loss=0.3252, pruned_loss=0.1015, over 19604.00 frames. ], tot_loss[loss=0.2569, simple_loss=0.3238, pruned_loss=0.09498, over 3812214.92 frames. ], batch size: 52, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:57:02,093 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46828.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:57:09,250 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2786, 1.0714, 1.2056, 1.3804, 0.8856, 1.3808, 1.3646, 1.2484], + device='cuda:3'), covar=tensor([0.0869, 0.1157, 0.1111, 0.0708, 0.0989, 0.0832, 0.0822, 0.0775], + device='cuda:3'), in_proj_covar=tensor([0.0218, 0.0237, 0.0233, 0.0263, 0.0256, 0.0218, 0.0213, 0.0212], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 10:57:09,591 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.55 vs. limit=5.0 +2023-04-01 10:57:26,224 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0463, 2.2774, 2.4101, 2.4108, 1.0793, 2.2133, 2.0601, 2.1946], + device='cuda:3'), covar=tensor([0.1086, 0.1991, 0.0618, 0.0638, 0.3468, 0.0882, 0.0577, 0.1000], + device='cuda:3'), in_proj_covar=tensor([0.0579, 0.0508, 0.0694, 0.0571, 0.0640, 0.0442, 0.0434, 0.0649], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 10:57:29,389 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.050e+02 5.967e+02 7.374e+02 9.303e+02 2.879e+03, threshold=1.475e+03, percent-clipped=6.0 +2023-04-01 10:57:51,519 INFO [train.py:903] (3/4) Epoch 7, batch 5900, loss[loss=0.3024, simple_loss=0.3596, pruned_loss=0.1226, over 17272.00 frames. ], tot_loss[loss=0.2592, simple_loss=0.3257, pruned_loss=0.09633, over 3800687.10 frames. ], batch size: 101, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:57:57,275 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 10:58:16,952 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 10:58:52,939 INFO [train.py:903] (3/4) Epoch 7, batch 5950, loss[loss=0.2575, simple_loss=0.3286, pruned_loss=0.09322, over 19357.00 frames. ], tot_loss[loss=0.2599, simple_loss=0.3264, pruned_loss=0.09674, over 3789189.30 frames. ], batch size: 70, lr: 1.16e-02, grad_scale: 8.0 +2023-04-01 10:59:05,782 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8679, 4.3524, 4.7016, 4.6163, 1.7111, 4.3286, 3.8050, 4.2561], + device='cuda:3'), covar=tensor([0.1172, 0.0683, 0.0505, 0.0458, 0.4637, 0.0540, 0.0536, 0.1130], + device='cuda:3'), in_proj_covar=tensor([0.0584, 0.0514, 0.0695, 0.0574, 0.0648, 0.0446, 0.0440, 0.0654], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 10:59:34,053 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.877e+02 6.644e+02 7.805e+02 9.690e+02 1.794e+03, threshold=1.561e+03, percent-clipped=4.0 +2023-04-01 10:59:40,942 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46957.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:59:48,569 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=46964.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 10:59:52,804 INFO [train.py:903] (3/4) Epoch 7, batch 6000, loss[loss=0.2287, simple_loss=0.2952, pruned_loss=0.08107, over 19394.00 frames. ], tot_loss[loss=0.2592, simple_loss=0.3256, pruned_loss=0.09634, over 3786457.69 frames. ], batch size: 48, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 10:59:52,805 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 11:00:03,700 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9857, 1.1899, 1.3072, 1.4468, 2.5524, 0.9095, 2.0004, 2.7919], + device='cuda:3'), covar=tensor([0.0407, 0.2725, 0.2817, 0.1820, 0.0660, 0.2563, 0.0964, 0.0373], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0317, 0.0324, 0.0297, 0.0326, 0.0321, 0.0296, 0.0315], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:00:05,284 INFO [train.py:937] (3/4) Epoch 7, validation: loss=0.1903, simple_loss=0.2902, pruned_loss=0.04516, over 944034.00 frames. +2023-04-01 11:00:05,284 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 11:00:41,388 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1521, 3.5905, 2.0288, 2.1114, 3.1763, 1.8012, 1.1551, 1.9034], + device='cuda:3'), covar=tensor([0.0947, 0.0410, 0.0837, 0.0580, 0.0324, 0.0870, 0.0881, 0.0543], + device='cuda:3'), in_proj_covar=tensor([0.0284, 0.0283, 0.0316, 0.0238, 0.0225, 0.0309, 0.0286, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:00:57,569 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47009.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:01:09,016 INFO [train.py:903] (3/4) Epoch 7, batch 6050, loss[loss=0.201, simple_loss=0.273, pruned_loss=0.06453, over 19303.00 frames. ], tot_loss[loss=0.258, simple_loss=0.3246, pruned_loss=0.09567, over 3785162.88 frames. ], batch size: 44, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:01:30,905 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47034.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:01:49,944 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.317e+02 6.018e+02 7.577e+02 1.005e+03 2.434e+03, threshold=1.515e+03, percent-clipped=7.0 +2023-04-01 11:02:13,573 INFO [train.py:903] (3/4) Epoch 7, batch 6100, loss[loss=0.2737, simple_loss=0.3355, pruned_loss=0.106, over 19584.00 frames. ], tot_loss[loss=0.2578, simple_loss=0.3245, pruned_loss=0.09551, over 3788822.62 frames. ], batch size: 61, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:02:15,643 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-04-01 11:02:18,752 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47072.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:02:26,705 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47079.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:03:15,748 INFO [train.py:903] (3/4) Epoch 7, batch 6150, loss[loss=0.2711, simple_loss=0.3428, pruned_loss=0.09971, over 18154.00 frames. ], tot_loss[loss=0.2572, simple_loss=0.3243, pruned_loss=0.09504, over 3800621.67 frames. ], batch size: 83, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:03:16,150 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1185, 3.6296, 2.1518, 2.0643, 2.9997, 1.7694, 1.1834, 1.9974], + device='cuda:3'), covar=tensor([0.0887, 0.0327, 0.0694, 0.0559, 0.0412, 0.0834, 0.0848, 0.0538], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0284, 0.0314, 0.0239, 0.0226, 0.0312, 0.0288, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:03:41,641 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 11:03:56,565 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.178e+02 5.986e+02 7.602e+02 9.492e+02 2.168e+03, threshold=1.520e+03, percent-clipped=5.0 +2023-04-01 11:04:15,665 INFO [train.py:903] (3/4) Epoch 7, batch 6200, loss[loss=0.2457, simple_loss=0.3057, pruned_loss=0.09281, over 19400.00 frames. ], tot_loss[loss=0.2584, simple_loss=0.3253, pruned_loss=0.09576, over 3811272.05 frames. ], batch size: 48, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:04:20,365 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=47172.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:04:38,672 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2056, 1.2433, 1.6300, 1.4058, 2.3128, 1.9769, 2.3881, 0.7951], + device='cuda:3'), covar=tensor([0.1901, 0.3302, 0.1730, 0.1583, 0.1058, 0.1597, 0.1134, 0.3058], + device='cuda:3'), in_proj_covar=tensor([0.0451, 0.0536, 0.0527, 0.0421, 0.0570, 0.0465, 0.0633, 0.0463], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:05:17,526 INFO [train.py:903] (3/4) Epoch 7, batch 6250, loss[loss=0.251, simple_loss=0.3272, pruned_loss=0.0874, over 19668.00 frames. ], tot_loss[loss=0.2585, simple_loss=0.3257, pruned_loss=0.09563, over 3810782.29 frames. ], batch size: 53, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:05:47,059 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 11:05:57,377 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.252e+02 6.379e+02 7.672e+02 9.726e+02 2.182e+03, threshold=1.534e+03, percent-clipped=2.0 +2023-04-01 11:06:08,352 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47259.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:06:19,225 INFO [train.py:903] (3/4) Epoch 7, batch 6300, loss[loss=0.2623, simple_loss=0.3371, pruned_loss=0.09379, over 19539.00 frames. ], tot_loss[loss=0.2598, simple_loss=0.3269, pruned_loss=0.09638, over 3819449.99 frames. ], batch size: 54, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:06:43,162 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47287.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:06:49,173 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2510, 1.3321, 1.6877, 1.4661, 2.4428, 2.0993, 2.6289, 1.0490], + device='cuda:3'), covar=tensor([0.1967, 0.3433, 0.1890, 0.1642, 0.1341, 0.1721, 0.1460, 0.3090], + device='cuda:3'), in_proj_covar=tensor([0.0450, 0.0538, 0.0527, 0.0420, 0.0571, 0.0463, 0.0630, 0.0461], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:07:21,231 INFO [train.py:903] (3/4) Epoch 7, batch 6350, loss[loss=0.2701, simple_loss=0.3448, pruned_loss=0.09766, over 19702.00 frames. ], tot_loss[loss=0.2595, simple_loss=0.3264, pruned_loss=0.09628, over 3826622.54 frames. ], batch size: 59, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:07:26,539 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.05 vs. limit=5.0 +2023-04-01 11:07:33,241 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47328.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:07:42,294 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47335.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:08:02,956 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.061e+02 5.885e+02 7.911e+02 1.072e+03 3.322e+03, threshold=1.582e+03, percent-clipped=7.0 +2023-04-01 11:08:05,624 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47353.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:08:13,943 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47360.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:08:22,753 INFO [train.py:903] (3/4) Epoch 7, batch 6400, loss[loss=0.2596, simple_loss=0.3329, pruned_loss=0.09313, over 19617.00 frames. ], tot_loss[loss=0.26, simple_loss=0.3269, pruned_loss=0.09654, over 3822292.70 frames. ], batch size: 57, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:09:00,242 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47397.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:09:25,579 INFO [train.py:903] (3/4) Epoch 7, batch 6450, loss[loss=0.2115, simple_loss=0.2807, pruned_loss=0.07119, over 19751.00 frames. ], tot_loss[loss=0.2595, simple_loss=0.3269, pruned_loss=0.0961, over 3809683.13 frames. ], batch size: 45, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:10:05,688 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.425e+02 6.283e+02 7.588e+02 1.008e+03 1.535e+03, threshold=1.518e+03, percent-clipped=0.0 +2023-04-01 11:10:08,042 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 11:10:26,631 INFO [train.py:903] (3/4) Epoch 7, batch 6500, loss[loss=0.3641, simple_loss=0.3972, pruned_loss=0.1655, over 13450.00 frames. ], tot_loss[loss=0.2584, simple_loss=0.3258, pruned_loss=0.09552, over 3817426.80 frames. ], batch size: 135, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:10:29,888 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 11:11:27,816 INFO [train.py:903] (3/4) Epoch 7, batch 6550, loss[loss=0.2678, simple_loss=0.3171, pruned_loss=0.1092, over 18592.00 frames. ], tot_loss[loss=0.2587, simple_loss=0.3257, pruned_loss=0.09588, over 3821423.48 frames. ], batch size: 41, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:11:39,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7246, 1.3261, 1.4979, 1.6092, 3.2051, 1.0792, 2.2351, 3.4376], + device='cuda:3'), covar=tensor([0.0416, 0.2588, 0.2549, 0.1639, 0.0641, 0.2513, 0.1201, 0.0365], + device='cuda:3'), in_proj_covar=tensor([0.0315, 0.0317, 0.0322, 0.0298, 0.0324, 0.0320, 0.0297, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:11:58,295 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47543.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:12:06,377 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7101, 1.7595, 1.8533, 2.5554, 1.6704, 2.2069, 2.2915, 1.8362], + device='cuda:3'), covar=tensor([0.2671, 0.2301, 0.1167, 0.1140, 0.2434, 0.1072, 0.2399, 0.2071], + device='cuda:3'), in_proj_covar=tensor([0.0714, 0.0721, 0.0605, 0.0852, 0.0730, 0.0632, 0.0746, 0.0655], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:12:10,157 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.342e+02 6.347e+02 8.071e+02 1.082e+03 2.174e+03, threshold=1.614e+03, percent-clipped=4.0 +2023-04-01 11:12:16,324 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4612, 1.2377, 1.1311, 1.3679, 1.1373, 1.2763, 1.1469, 1.2938], + device='cuda:3'), covar=tensor([0.0973, 0.1193, 0.1381, 0.0831, 0.1083, 0.0565, 0.1140, 0.0790], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0358, 0.0288, 0.0239, 0.0304, 0.0243, 0.0273, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:12:29,849 INFO [train.py:903] (3/4) Epoch 7, batch 6600, loss[loss=0.2263, simple_loss=0.3044, pruned_loss=0.07409, over 19668.00 frames. ], tot_loss[loss=0.26, simple_loss=0.3268, pruned_loss=0.09661, over 3819017.16 frames. ], batch size: 53, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:12:30,269 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47568.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:13:13,417 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=47603.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:13:31,701 INFO [train.py:903] (3/4) Epoch 7, batch 6650, loss[loss=0.2296, simple_loss=0.2971, pruned_loss=0.0811, over 19481.00 frames. ], tot_loss[loss=0.2595, simple_loss=0.326, pruned_loss=0.09645, over 3823469.01 frames. ], batch size: 49, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:13:37,027 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0083, 1.7083, 1.5892, 2.1422, 1.7916, 1.8249, 1.6342, 1.9863], + device='cuda:3'), covar=tensor([0.0916, 0.1695, 0.1423, 0.0963, 0.1374, 0.0481, 0.1146, 0.0642], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0359, 0.0289, 0.0240, 0.0305, 0.0245, 0.0275, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:14:11,833 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.526e+02 6.649e+02 7.986e+02 1.008e+03 1.623e+03, threshold=1.597e+03, percent-clipped=0.0 +2023-04-01 11:14:31,907 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-01 11:14:32,253 INFO [train.py:903] (3/4) Epoch 7, batch 6700, loss[loss=0.2079, simple_loss=0.2766, pruned_loss=0.06961, over 19720.00 frames. ], tot_loss[loss=0.2588, simple_loss=0.3251, pruned_loss=0.09623, over 3822541.72 frames. ], batch size: 45, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:15:21,500 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9190, 4.3129, 4.6473, 4.6523, 1.4819, 4.3377, 3.8017, 4.2115], + device='cuda:3'), covar=tensor([0.1249, 0.0699, 0.0589, 0.0494, 0.5040, 0.0401, 0.0592, 0.1142], + device='cuda:3'), in_proj_covar=tensor([0.0589, 0.0510, 0.0699, 0.0580, 0.0643, 0.0446, 0.0443, 0.0650], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 11:15:30,804 INFO [train.py:903] (3/4) Epoch 7, batch 6750, loss[loss=0.3119, simple_loss=0.3524, pruned_loss=0.1357, over 19726.00 frames. ], tot_loss[loss=0.2575, simple_loss=0.3239, pruned_loss=0.09555, over 3813703.43 frames. ], batch size: 51, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:15:31,133 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47718.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:15:58,011 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=47741.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:16:09,116 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.932e+02 5.831e+02 7.200e+02 8.445e+02 1.747e+03, threshold=1.440e+03, percent-clipped=3.0 +2023-04-01 11:16:13,239 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.42 vs. limit=5.0 +2023-04-01 11:16:15,241 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47756.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:16:28,688 INFO [train.py:903] (3/4) Epoch 7, batch 6800, loss[loss=0.2421, simple_loss=0.3183, pruned_loss=0.08288, over 19498.00 frames. ], tot_loss[loss=0.2571, simple_loss=0.324, pruned_loss=0.09516, over 3816739.93 frames. ], batch size: 64, lr: 1.15e-02, grad_scale: 8.0 +2023-04-01 11:17:14,509 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 11:17:15,587 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 11:17:18,393 INFO [train.py:903] (3/4) Epoch 8, batch 0, loss[loss=0.2208, simple_loss=0.2918, pruned_loss=0.07491, over 19834.00 frames. ], tot_loss[loss=0.2208, simple_loss=0.2918, pruned_loss=0.07491, over 19834.00 frames. ], batch size: 52, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:17:18,394 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 11:17:30,978 INFO [train.py:937] (3/4) Epoch 8, validation: loss=0.1916, simple_loss=0.2915, pruned_loss=0.0458, over 944034.00 frames. +2023-04-01 11:17:30,979 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 11:17:41,962 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 11:17:43,324 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47806.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:18:07,512 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47827.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:18:32,756 INFO [train.py:903] (3/4) Epoch 8, batch 50, loss[loss=0.2007, simple_loss=0.2809, pruned_loss=0.06027, over 19742.00 frames. ], tot_loss[loss=0.2538, simple_loss=0.3222, pruned_loss=0.09273, over 861178.19 frames. ], batch size: 51, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:18:38,613 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.560e+02 5.925e+02 7.453e+02 9.478e+02 2.348e+03, threshold=1.491e+03, percent-clipped=8.0 +2023-04-01 11:18:44,831 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47856.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:18:52,168 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.85 vs. limit=5.0 +2023-04-01 11:19:06,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 11:19:32,493 INFO [train.py:903] (3/4) Epoch 8, batch 100, loss[loss=0.2381, simple_loss=0.3008, pruned_loss=0.0877, over 19779.00 frames. ], tot_loss[loss=0.255, simple_loss=0.3231, pruned_loss=0.09345, over 1529607.93 frames. ], batch size: 47, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:19:42,616 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 11:20:32,672 INFO [train.py:903] (3/4) Epoch 8, batch 150, loss[loss=0.2516, simple_loss=0.3301, pruned_loss=0.08652, over 18712.00 frames. ], tot_loss[loss=0.2553, simple_loss=0.3228, pruned_loss=0.09389, over 2050463.92 frames. ], batch size: 74, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:20:38,416 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.629e+02 6.208e+02 7.626e+02 9.440e+02 2.273e+03, threshold=1.525e+03, percent-clipped=3.0 +2023-04-01 11:20:38,917 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2187, 2.9202, 1.9619, 2.0020, 1.8447, 2.4139, 0.6849, 2.0908], + device='cuda:3'), covar=tensor([0.0351, 0.0307, 0.0411, 0.0557, 0.0608, 0.0526, 0.0723, 0.0603], + device='cuda:3'), in_proj_covar=tensor([0.0311, 0.0314, 0.0307, 0.0325, 0.0398, 0.0322, 0.0285, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:21:07,734 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47974.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:21:29,238 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 11:21:32,698 INFO [train.py:903] (3/4) Epoch 8, batch 200, loss[loss=0.2641, simple_loss=0.332, pruned_loss=0.09815, over 19691.00 frames. ], tot_loss[loss=0.2539, simple_loss=0.3219, pruned_loss=0.09291, over 2444428.72 frames. ], batch size: 58, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:21:36,577 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47999.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:22:15,569 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2336, 1.2307, 1.6343, 1.2172, 2.7936, 3.5292, 3.3229, 3.7542], + device='cuda:3'), covar=tensor([0.1448, 0.3124, 0.2901, 0.2027, 0.0458, 0.0171, 0.0213, 0.0141], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0284, 0.0317, 0.0247, 0.0208, 0.0136, 0.0204, 0.0172], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:22:35,823 INFO [train.py:903] (3/4) Epoch 8, batch 250, loss[loss=0.2343, simple_loss=0.3124, pruned_loss=0.07807, over 19656.00 frames. ], tot_loss[loss=0.2554, simple_loss=0.323, pruned_loss=0.09385, over 2744417.64 frames. ], batch size: 60, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:22:42,343 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.557e+02 6.073e+02 7.344e+02 8.973e+02 2.163e+03, threshold=1.469e+03, percent-clipped=4.0 +2023-04-01 11:23:30,381 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48090.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:23:31,782 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2932, 1.3191, 1.5775, 1.4879, 2.1592, 2.1363, 2.3256, 0.7378], + device='cuda:3'), covar=tensor([0.1849, 0.3316, 0.1881, 0.1512, 0.1241, 0.1632, 0.1111, 0.3248], + device='cuda:3'), in_proj_covar=tensor([0.0455, 0.0533, 0.0530, 0.0416, 0.0572, 0.0467, 0.0624, 0.0459], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:23:36,834 INFO [train.py:903] (3/4) Epoch 8, batch 300, loss[loss=0.2649, simple_loss=0.3277, pruned_loss=0.1011, over 18695.00 frames. ], tot_loss[loss=0.257, simple_loss=0.3242, pruned_loss=0.09489, over 2981124.51 frames. ], batch size: 74, lr: 1.08e-02, grad_scale: 8.0 +2023-04-01 11:23:41,649 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48100.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:23:55,944 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48112.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:23:59,582 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2345, 1.3114, 1.6968, 1.4643, 2.4698, 2.3312, 2.6426, 0.9561], + device='cuda:3'), covar=tensor([0.1953, 0.3449, 0.1937, 0.1514, 0.1377, 0.1558, 0.1456, 0.3288], + device='cuda:3'), in_proj_covar=tensor([0.0458, 0.0536, 0.0532, 0.0418, 0.0575, 0.0468, 0.0629, 0.0461], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:24:14,361 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8157, 1.8629, 1.9135, 2.6948, 1.7705, 2.4192, 2.3206, 1.7348], + device='cuda:3'), covar=tensor([0.2720, 0.2231, 0.1136, 0.1186, 0.2509, 0.1016, 0.2426, 0.2165], + device='cuda:3'), in_proj_covar=tensor([0.0721, 0.0723, 0.0607, 0.0852, 0.0734, 0.0632, 0.0748, 0.0657], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:24:27,823 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48137.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:24:37,683 INFO [train.py:903] (3/4) Epoch 8, batch 350, loss[loss=0.208, simple_loss=0.2772, pruned_loss=0.06943, over 19406.00 frames. ], tot_loss[loss=0.2556, simple_loss=0.3227, pruned_loss=0.09429, over 3182747.90 frames. ], batch size: 48, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:24:39,944 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 11:24:42,120 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48150.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:24:43,059 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.994e+02 6.192e+02 7.149e+02 9.949e+02 1.629e+03, threshold=1.430e+03, percent-clipped=6.0 +2023-04-01 11:25:09,480 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48171.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:25:22,169 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8168, 1.2974, 1.3845, 2.1231, 1.5901, 1.7542, 1.9890, 1.6831], + device='cuda:3'), covar=tensor([0.0811, 0.1323, 0.1200, 0.0904, 0.0907, 0.0927, 0.1035, 0.0847], + device='cuda:3'), in_proj_covar=tensor([0.0217, 0.0235, 0.0232, 0.0262, 0.0246, 0.0217, 0.0211, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 11:25:33,669 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4685, 1.2290, 1.7561, 1.2111, 2.8718, 3.5987, 3.4383, 3.8755], + device='cuda:3'), covar=tensor([0.1347, 0.3285, 0.2841, 0.1974, 0.0413, 0.0172, 0.0197, 0.0145], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0284, 0.0313, 0.0244, 0.0207, 0.0136, 0.0202, 0.0171], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:25:37,717 INFO [train.py:903] (3/4) Epoch 8, batch 400, loss[loss=0.2785, simple_loss=0.3553, pruned_loss=0.1008, over 19104.00 frames. ], tot_loss[loss=0.2544, simple_loss=0.3216, pruned_loss=0.09359, over 3328440.38 frames. ], batch size: 69, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:25:57,727 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3992, 2.1863, 1.7402, 1.6262, 1.4636, 1.6904, 0.4939, 1.2557], + device='cuda:3'), covar=tensor([0.0286, 0.0323, 0.0283, 0.0446, 0.0673, 0.0451, 0.0641, 0.0533], + device='cuda:3'), in_proj_covar=tensor([0.0309, 0.0311, 0.0306, 0.0324, 0.0397, 0.0318, 0.0285, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:26:03,330 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9334, 5.3311, 2.9381, 4.6798, 0.9427, 5.0511, 5.1492, 5.4003], + device='cuda:3'), covar=tensor([0.0449, 0.0890, 0.1673, 0.0533, 0.4088, 0.0487, 0.0599, 0.0671], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0340, 0.0395, 0.0297, 0.0364, 0.0322, 0.0316, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 11:26:03,485 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48215.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:26:13,177 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8218, 1.9710, 2.0015, 2.7913, 1.8866, 2.5451, 2.5448, 1.9086], + device='cuda:3'), covar=tensor([0.2897, 0.2355, 0.1128, 0.1370, 0.2678, 0.1105, 0.2425, 0.2063], + device='cuda:3'), in_proj_covar=tensor([0.0726, 0.0730, 0.0612, 0.0861, 0.0737, 0.0638, 0.0756, 0.0660], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:26:14,354 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0080, 2.1477, 2.1112, 3.1622, 2.2897, 3.3558, 2.9479, 2.1686], + device='cuda:3'), covar=tensor([0.2924, 0.2260, 0.1134, 0.1519, 0.2691, 0.0905, 0.2164, 0.1879], + device='cuda:3'), in_proj_covar=tensor([0.0726, 0.0730, 0.0612, 0.0861, 0.0737, 0.0638, 0.0756, 0.0660], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:26:21,225 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48230.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:26:38,937 INFO [train.py:903] (3/4) Epoch 8, batch 450, loss[loss=0.2239, simple_loss=0.2951, pruned_loss=0.07637, over 19777.00 frames. ], tot_loss[loss=0.2537, simple_loss=0.3213, pruned_loss=0.09304, over 3435343.15 frames. ], batch size: 54, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:26:45,636 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.339e+02 5.626e+02 7.034e+02 8.568e+02 1.629e+03, threshold=1.407e+03, percent-clipped=1.0 +2023-04-01 11:27:03,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48265.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:27:11,113 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 11:27:12,252 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 11:27:27,229 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48286.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:27:41,081 INFO [train.py:903] (3/4) Epoch 8, batch 500, loss[loss=0.2317, simple_loss=0.2987, pruned_loss=0.08242, over 19622.00 frames. ], tot_loss[loss=0.2544, simple_loss=0.3218, pruned_loss=0.09345, over 3514613.50 frames. ], batch size: 50, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:28:37,114 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48342.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:28:41,346 INFO [train.py:903] (3/4) Epoch 8, batch 550, loss[loss=0.2378, simple_loss=0.3003, pruned_loss=0.08761, over 19394.00 frames. ], tot_loss[loss=0.2527, simple_loss=0.3204, pruned_loss=0.09249, over 3591974.37 frames. ], batch size: 48, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:28:47,050 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.262e+02 6.228e+02 7.563e+02 9.337e+02 1.593e+03, threshold=1.513e+03, percent-clipped=3.0 +2023-04-01 11:28:47,316 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8983, 4.3711, 2.7290, 3.8324, 1.1551, 4.1544, 4.1316, 4.2653], + device='cuda:3'), covar=tensor([0.0551, 0.1134, 0.1844, 0.0741, 0.3610, 0.0655, 0.0604, 0.0787], + device='cuda:3'), in_proj_covar=tensor([0.0393, 0.0344, 0.0398, 0.0302, 0.0366, 0.0325, 0.0318, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 11:29:23,958 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1991, 1.3308, 1.1394, 0.9371, 1.0427, 1.0846, 0.0403, 0.3179], + device='cuda:3'), covar=tensor([0.0388, 0.0434, 0.0252, 0.0363, 0.0806, 0.0327, 0.0645, 0.0644], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0313, 0.0309, 0.0328, 0.0401, 0.0321, 0.0289, 0.0312], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:29:42,861 INFO [train.py:903] (3/4) Epoch 8, batch 600, loss[loss=0.2922, simple_loss=0.3519, pruned_loss=0.1162, over 19783.00 frames. ], tot_loss[loss=0.2521, simple_loss=0.3198, pruned_loss=0.09217, over 3654967.62 frames. ], batch size: 56, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:30:05,020 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48415.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:30:26,080 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 11:30:31,029 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48434.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:30:39,411 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48441.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:30:44,670 INFO [train.py:903] (3/4) Epoch 8, batch 650, loss[loss=0.2461, simple_loss=0.3221, pruned_loss=0.08511, over 19507.00 frames. ], tot_loss[loss=0.2525, simple_loss=0.3204, pruned_loss=0.09232, over 3698724.77 frames. ], batch size: 56, lr: 1.07e-02, grad_scale: 16.0 +2023-04-01 11:30:50,369 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.061e+02 6.183e+02 7.491e+02 9.829e+02 2.830e+03, threshold=1.498e+03, percent-clipped=3.0 +2023-04-01 11:31:17,364 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48471.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:31:45,461 INFO [train.py:903] (3/4) Epoch 8, batch 700, loss[loss=0.3213, simple_loss=0.374, pruned_loss=0.1343, over 13315.00 frames. ], tot_loss[loss=0.2533, simple_loss=0.3208, pruned_loss=0.09284, over 3718906.87 frames. ], batch size: 135, lr: 1.07e-02, grad_scale: 16.0 +2023-04-01 11:31:45,799 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48496.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:32:17,670 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48521.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:32:43,992 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48542.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:32:48,872 INFO [train.py:903] (3/4) Epoch 8, batch 750, loss[loss=0.2574, simple_loss=0.3253, pruned_loss=0.09473, over 19523.00 frames. ], tot_loss[loss=0.254, simple_loss=0.3214, pruned_loss=0.09325, over 3751301.86 frames. ], batch size: 54, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:32:49,333 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48546.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:32:52,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48549.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:32:55,668 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.581e+02 6.054e+02 7.636e+02 9.380e+02 1.990e+03, threshold=1.527e+03, percent-clipped=3.0 +2023-04-01 11:33:13,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48567.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:33:21,448 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48574.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:33:37,263 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1836, 1.2964, 1.6539, 1.2409, 2.6517, 3.3546, 3.1517, 3.6186], + device='cuda:3'), covar=tensor([0.1544, 0.3165, 0.2867, 0.2050, 0.0513, 0.0226, 0.0219, 0.0162], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0285, 0.0316, 0.0246, 0.0209, 0.0135, 0.0202, 0.0172], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:33:38,455 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5165, 1.8201, 1.9700, 2.5928, 2.2763, 2.2417, 2.0325, 2.5367], + device='cuda:3'), covar=tensor([0.0833, 0.1950, 0.1355, 0.0871, 0.1296, 0.0470, 0.1126, 0.0596], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0358, 0.0285, 0.0238, 0.0303, 0.0243, 0.0273, 0.0231], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:33:49,551 INFO [train.py:903] (3/4) Epoch 8, batch 800, loss[loss=0.2402, simple_loss=0.3023, pruned_loss=0.08906, over 19601.00 frames. ], tot_loss[loss=0.2561, simple_loss=0.3232, pruned_loss=0.09451, over 3769814.46 frames. ], batch size: 52, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:34:02,660 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 11:34:34,446 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8035, 4.3183, 2.4270, 3.7786, 1.2353, 3.9850, 3.9999, 4.2366], + device='cuda:3'), covar=tensor([0.0639, 0.1173, 0.2234, 0.0764, 0.3773, 0.0803, 0.0754, 0.0932], + device='cuda:3'), in_proj_covar=tensor([0.0394, 0.0344, 0.0401, 0.0302, 0.0366, 0.0328, 0.0318, 0.0355], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 11:34:51,087 INFO [train.py:903] (3/4) Epoch 8, batch 850, loss[loss=0.2933, simple_loss=0.3458, pruned_loss=0.1204, over 12857.00 frames. ], tot_loss[loss=0.2561, simple_loss=0.3232, pruned_loss=0.09453, over 3770418.73 frames. ], batch size: 135, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:34:57,942 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.799e+02 5.922e+02 7.936e+02 9.993e+02 1.897e+03, threshold=1.587e+03, percent-clipped=5.0 +2023-04-01 11:35:39,758 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 11:35:39,901 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48686.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:35:43,344 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48689.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:35:50,882 INFO [train.py:903] (3/4) Epoch 8, batch 900, loss[loss=0.2595, simple_loss=0.3343, pruned_loss=0.09237, over 19525.00 frames. ], tot_loss[loss=0.2567, simple_loss=0.3238, pruned_loss=0.09482, over 3756761.20 frames. ], batch size: 56, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:36:35,812 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0918, 1.1760, 1.0092, 0.8129, 0.9059, 0.9512, 0.0599, 0.3438], + device='cuda:3'), covar=tensor([0.0295, 0.0308, 0.0198, 0.0279, 0.0545, 0.0265, 0.0581, 0.0466], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0313, 0.0311, 0.0329, 0.0402, 0.0324, 0.0291, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:36:54,666 INFO [train.py:903] (3/4) Epoch 8, batch 950, loss[loss=0.2455, simple_loss=0.322, pruned_loss=0.08448, over 19667.00 frames. ], tot_loss[loss=0.256, simple_loss=0.3235, pruned_loss=0.09427, over 3772431.26 frames. ], batch size: 58, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:36:56,558 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 11:37:03,057 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.996e+02 5.931e+02 7.048e+02 8.289e+02 1.665e+03, threshold=1.410e+03, percent-clipped=1.0 +2023-04-01 11:37:11,362 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48759.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:37:42,413 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=48785.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:37:57,228 INFO [train.py:903] (3/4) Epoch 8, batch 1000, loss[loss=0.2396, simple_loss=0.3053, pruned_loss=0.08694, over 19659.00 frames. ], tot_loss[loss=0.2531, simple_loss=0.3209, pruned_loss=0.0926, over 3791011.99 frames. ], batch size: 53, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:38:03,560 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48801.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:38:08,276 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48805.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:38:41,025 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48830.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:38:50,181 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 11:38:51,478 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48839.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:38:59,277 INFO [train.py:903] (3/4) Epoch 8, batch 1050, loss[loss=0.2891, simple_loss=0.3532, pruned_loss=0.1125, over 19603.00 frames. ], tot_loss[loss=0.2538, simple_loss=0.3215, pruned_loss=0.09309, over 3799384.15 frames. ], batch size: 61, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:39:06,227 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.557e+02 5.500e+02 6.561e+02 8.180e+02 1.521e+03, threshold=1.312e+03, percent-clipped=1.0 +2023-04-01 11:39:11,120 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48856.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:39:31,755 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 11:39:34,235 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48874.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:39:45,603 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8650, 4.2363, 4.4890, 4.4733, 1.4389, 4.1868, 3.5876, 4.1045], + device='cuda:3'), covar=tensor([0.1180, 0.0699, 0.0544, 0.0496, 0.4848, 0.0495, 0.0602, 0.1069], + device='cuda:3'), in_proj_covar=tensor([0.0584, 0.0514, 0.0691, 0.0581, 0.0643, 0.0439, 0.0442, 0.0642], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 11:39:45,753 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2585, 2.1234, 1.7479, 1.7060, 1.4805, 1.7498, 0.5614, 1.0895], + device='cuda:3'), covar=tensor([0.0306, 0.0337, 0.0250, 0.0414, 0.0693, 0.0429, 0.0629, 0.0609], + device='cuda:3'), in_proj_covar=tensor([0.0311, 0.0312, 0.0309, 0.0329, 0.0401, 0.0320, 0.0289, 0.0313], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 11:39:58,991 INFO [train.py:903] (3/4) Epoch 8, batch 1100, loss[loss=0.1996, simple_loss=0.2755, pruned_loss=0.06189, over 19734.00 frames. ], tot_loss[loss=0.2543, simple_loss=0.3217, pruned_loss=0.09343, over 3804189.25 frames. ], batch size: 46, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:40:04,002 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48900.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:41:00,124 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48945.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 11:41:00,833 INFO [train.py:903] (3/4) Epoch 8, batch 1150, loss[loss=0.3345, simple_loss=0.381, pruned_loss=0.144, over 13633.00 frames. ], tot_loss[loss=0.2546, simple_loss=0.3222, pruned_loss=0.09352, over 3797150.98 frames. ], batch size: 136, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:41:09,120 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.525e+02 5.943e+02 6.952e+02 8.882e+02 1.618e+03, threshold=1.390e+03, percent-clipped=5.0 +2023-04-01 11:41:30,913 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48970.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:42:04,113 INFO [train.py:903] (3/4) Epoch 8, batch 1200, loss[loss=0.2381, simple_loss=0.3116, pruned_loss=0.08236, over 19502.00 frames. ], tot_loss[loss=0.254, simple_loss=0.3217, pruned_loss=0.09311, over 3803188.81 frames. ], batch size: 64, lr: 1.07e-02, grad_scale: 8.0 +2023-04-01 11:42:32,672 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 11:43:05,009 INFO [train.py:903] (3/4) Epoch 8, batch 1250, loss[loss=0.2216, simple_loss=0.2862, pruned_loss=0.07845, over 19730.00 frames. ], tot_loss[loss=0.2541, simple_loss=0.3215, pruned_loss=0.09339, over 3811794.07 frames. ], batch size: 45, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:43:11,745 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.529e+02 6.811e+02 8.521e+02 1.008e+03 2.064e+03, threshold=1.704e+03, percent-clipped=4.0 +2023-04-01 11:43:18,050 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49057.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:43:39,591 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.10 vs. limit=2.0 +2023-04-01 11:43:50,337 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49082.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:44:06,035 INFO [train.py:903] (3/4) Epoch 8, batch 1300, loss[loss=0.2751, simple_loss=0.3447, pruned_loss=0.1027, over 19349.00 frames. ], tot_loss[loss=0.254, simple_loss=0.3217, pruned_loss=0.09312, over 3827370.62 frames. ], batch size: 70, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:44:49,033 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49130.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:45:07,322 INFO [train.py:903] (3/4) Epoch 8, batch 1350, loss[loss=0.2529, simple_loss=0.324, pruned_loss=0.09094, over 19449.00 frames. ], tot_loss[loss=0.2551, simple_loss=0.3224, pruned_loss=0.09388, over 3842093.94 frames. ], batch size: 64, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:45:08,146 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-04-01 11:45:16,536 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.771e+02 5.835e+02 7.092e+02 8.908e+02 2.388e+03, threshold=1.418e+03, percent-clipped=3.0 +2023-04-01 11:45:21,321 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49155.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:45:22,444 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49156.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:45:41,563 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4160, 2.2988, 1.5421, 1.5315, 2.0817, 1.2003, 1.1717, 1.7101], + device='cuda:3'), covar=tensor([0.0889, 0.0502, 0.0921, 0.0605, 0.0421, 0.0968, 0.0684, 0.0442], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0290, 0.0315, 0.0243, 0.0232, 0.0309, 0.0284, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:45:42,695 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6630, 1.1684, 1.4556, 1.5643, 3.2145, 1.1597, 2.0655, 3.3994], + device='cuda:3'), covar=tensor([0.0425, 0.2599, 0.2555, 0.1619, 0.0620, 0.2261, 0.1313, 0.0343], + device='cuda:3'), in_proj_covar=tensor([0.0313, 0.0317, 0.0325, 0.0298, 0.0325, 0.0319, 0.0297, 0.0321], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:45:52,074 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49181.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:45:54,143 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=49183.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:46:10,832 INFO [train.py:903] (3/4) Epoch 8, batch 1400, loss[loss=0.2083, simple_loss=0.2775, pruned_loss=0.06955, over 19792.00 frames. ], tot_loss[loss=0.2534, simple_loss=0.3211, pruned_loss=0.09284, over 3830824.36 frames. ], batch size: 49, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:46:17,523 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=49200.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:47:12,311 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 11:47:13,439 INFO [train.py:903] (3/4) Epoch 8, batch 1450, loss[loss=0.2911, simple_loss=0.3602, pruned_loss=0.111, over 17313.00 frames. ], tot_loss[loss=0.254, simple_loss=0.3213, pruned_loss=0.09338, over 3811195.55 frames. ], batch size: 101, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:47:19,909 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.194e+02 6.215e+02 8.146e+02 9.729e+02 2.293e+03, threshold=1.629e+03, percent-clipped=2.0 +2023-04-01 11:47:22,668 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49254.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:48:14,512 INFO [train.py:903] (3/4) Epoch 8, batch 1500, loss[loss=0.2532, simple_loss=0.3131, pruned_loss=0.09665, over 19375.00 frames. ], tot_loss[loss=0.253, simple_loss=0.3208, pruned_loss=0.09261, over 3806014.53 frames. ], batch size: 47, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:48:16,763 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49298.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:48:38,601 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49315.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:49:05,504 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8377, 1.4928, 1.3589, 1.7162, 1.6745, 1.3241, 1.2106, 1.6301], + device='cuda:3'), covar=tensor([0.0924, 0.1483, 0.1470, 0.1031, 0.1199, 0.0785, 0.1397, 0.0752], + device='cuda:3'), in_proj_covar=tensor([0.0241, 0.0346, 0.0278, 0.0232, 0.0290, 0.0238, 0.0266, 0.0225], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 11:49:14,520 INFO [train.py:903] (3/4) Epoch 8, batch 1550, loss[loss=0.2375, simple_loss=0.3105, pruned_loss=0.08219, over 19771.00 frames. ], tot_loss[loss=0.2533, simple_loss=0.3213, pruned_loss=0.09261, over 3818923.12 frames. ], batch size: 54, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:49:23,152 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.230e+02 6.392e+02 7.844e+02 9.464e+02 1.840e+03, threshold=1.569e+03, percent-clipped=1.0 +2023-04-01 11:49:42,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3394, 3.9389, 2.5540, 3.5531, 1.1419, 3.6274, 3.6854, 3.8025], + device='cuda:3'), covar=tensor([0.0613, 0.1066, 0.1964, 0.0703, 0.3734, 0.0814, 0.0741, 0.0888], + device='cuda:3'), in_proj_covar=tensor([0.0393, 0.0341, 0.0399, 0.0298, 0.0368, 0.0326, 0.0318, 0.0355], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 11:50:17,463 INFO [train.py:903] (3/4) Epoch 8, batch 1600, loss[loss=0.2289, simple_loss=0.2937, pruned_loss=0.08209, over 19809.00 frames. ], tot_loss[loss=0.2538, simple_loss=0.3217, pruned_loss=0.09294, over 3821125.97 frames. ], batch size: 49, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:50:25,527 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7678, 4.1817, 4.5005, 4.4523, 1.8581, 4.1828, 3.6209, 4.1066], + device='cuda:3'), covar=tensor([0.1113, 0.0774, 0.0525, 0.0488, 0.4289, 0.0486, 0.0586, 0.1008], + device='cuda:3'), in_proj_covar=tensor([0.0600, 0.0533, 0.0717, 0.0598, 0.0667, 0.0458, 0.0454, 0.0663], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 11:50:38,465 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 11:51:12,154 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49440.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:51:20,003 INFO [train.py:903] (3/4) Epoch 8, batch 1650, loss[loss=0.2999, simple_loss=0.3599, pruned_loss=0.1199, over 18166.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3224, pruned_loss=0.09334, over 3818882.94 frames. ], batch size: 83, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:51:24,778 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49450.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:51:26,748 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.384e+02 6.230e+02 7.478e+02 9.229e+02 3.510e+03, threshold=1.496e+03, percent-clipped=3.0 +2023-04-01 11:52:21,801 INFO [train.py:903] (3/4) Epoch 8, batch 1700, loss[loss=0.2559, simple_loss=0.3099, pruned_loss=0.101, over 19768.00 frames. ], tot_loss[loss=0.2528, simple_loss=0.321, pruned_loss=0.09234, over 3827985.70 frames. ], batch size: 47, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:53:02,521 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 11:53:23,396 INFO [train.py:903] (3/4) Epoch 8, batch 1750, loss[loss=0.2332, simple_loss=0.3008, pruned_loss=0.08283, over 19394.00 frames. ], tot_loss[loss=0.2543, simple_loss=0.3223, pruned_loss=0.09315, over 3824647.28 frames. ], batch size: 48, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:53:31,458 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.581e+02 6.147e+02 7.390e+02 1.012e+03 1.809e+03, threshold=1.478e+03, percent-clipped=6.0 +2023-04-01 11:53:35,079 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49554.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:53:56,608 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49571.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:53:58,160 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.76 vs. limit=2.0 +2023-04-01 11:54:06,059 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49579.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:54:09,487 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3011, 3.8938, 2.5456, 3.4760, 1.1442, 3.5031, 3.5505, 3.7090], + device='cuda:3'), covar=tensor([0.0769, 0.1123, 0.2103, 0.0873, 0.4062, 0.0983, 0.0931, 0.1031], + device='cuda:3'), in_proj_covar=tensor([0.0396, 0.0343, 0.0401, 0.0299, 0.0366, 0.0325, 0.0316, 0.0356], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 11:54:27,937 INFO [train.py:903] (3/4) Epoch 8, batch 1800, loss[loss=0.2255, simple_loss=0.3059, pruned_loss=0.0725, over 18269.00 frames. ], tot_loss[loss=0.2537, simple_loss=0.322, pruned_loss=0.09269, over 3817562.04 frames. ], batch size: 83, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:54:28,357 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49596.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:54:31,445 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=49598.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:54:51,842 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7790, 4.1836, 4.4868, 4.4576, 1.8960, 4.1347, 3.6115, 4.0893], + device='cuda:3'), covar=tensor([0.1133, 0.0931, 0.0503, 0.0468, 0.4434, 0.0585, 0.0612, 0.1027], + device='cuda:3'), in_proj_covar=tensor([0.0606, 0.0533, 0.0718, 0.0602, 0.0669, 0.0461, 0.0454, 0.0665], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 11:55:25,416 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 11:55:26,233 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.44 vs. limit=5.0 +2023-04-01 11:55:30,988 INFO [train.py:903] (3/4) Epoch 8, batch 1850, loss[loss=0.2361, simple_loss=0.3145, pruned_loss=0.07883, over 19766.00 frames. ], tot_loss[loss=0.2535, simple_loss=0.3216, pruned_loss=0.09271, over 3803262.74 frames. ], batch size: 54, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:55:38,000 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.438e+02 5.845e+02 7.519e+02 8.649e+02 2.522e+03, threshold=1.504e+03, percent-clipped=4.0 +2023-04-01 11:56:02,474 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 11:56:31,217 INFO [train.py:903] (3/4) Epoch 8, batch 1900, loss[loss=0.2758, simple_loss=0.353, pruned_loss=0.09927, over 19607.00 frames. ], tot_loss[loss=0.256, simple_loss=0.3233, pruned_loss=0.09436, over 3805185.93 frames. ], batch size: 57, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:56:33,998 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3298, 1.4846, 1.7006, 1.5849, 2.6131, 2.2330, 2.7660, 0.9960], + device='cuda:3'), covar=tensor([0.1860, 0.3208, 0.1898, 0.1512, 0.1150, 0.1594, 0.1187, 0.3069], + device='cuda:3'), in_proj_covar=tensor([0.0457, 0.0535, 0.0532, 0.0415, 0.0571, 0.0467, 0.0627, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 11:56:48,645 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 11:56:52,394 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49713.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:56:54,388 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 11:57:18,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 11:57:33,017 INFO [train.py:903] (3/4) Epoch 8, batch 1950, loss[loss=0.2773, simple_loss=0.3486, pruned_loss=0.103, over 19315.00 frames. ], tot_loss[loss=0.2547, simple_loss=0.3225, pruned_loss=0.0935, over 3803030.46 frames. ], batch size: 66, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:57:40,107 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.492e+02 5.402e+02 6.624e+02 8.916e+02 2.925e+03, threshold=1.325e+03, percent-clipped=4.0 +2023-04-01 11:58:20,355 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=49784.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:58:20,517 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49784.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 11:58:32,991 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=49794.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 11:58:35,910 INFO [train.py:903] (3/4) Epoch 8, batch 2000, loss[loss=0.2399, simple_loss=0.3155, pruned_loss=0.08209, over 18135.00 frames. ], tot_loss[loss=0.2549, simple_loss=0.3229, pruned_loss=0.09348, over 3823311.05 frames. ], batch size: 83, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:59:34,101 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.94 vs. limit=5.0 +2023-04-01 11:59:35,611 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 11:59:38,067 INFO [train.py:903] (3/4) Epoch 8, batch 2050, loss[loss=0.2364, simple_loss=0.3122, pruned_loss=0.08031, over 19674.00 frames. ], tot_loss[loss=0.2559, simple_loss=0.324, pruned_loss=0.09388, over 3821523.75 frames. ], batch size: 59, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 11:59:45,925 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.480e+02 5.411e+02 7.156e+02 9.098e+02 3.444e+03, threshold=1.431e+03, percent-clipped=9.0 +2023-04-01 11:59:53,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 11:59:55,109 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 12:00:17,465 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 12:00:39,978 INFO [train.py:903] (3/4) Epoch 8, batch 2100, loss[loss=0.3198, simple_loss=0.3622, pruned_loss=0.1387, over 12910.00 frames. ], tot_loss[loss=0.2564, simple_loss=0.3244, pruned_loss=0.09418, over 3820167.30 frames. ], batch size: 135, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 12:00:43,786 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49899.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:00:55,216 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49909.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:01:10,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 12:01:32,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 12:01:40,057 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9272, 2.0111, 2.0895, 2.9115, 1.8386, 2.5661, 2.6757, 1.9614], + device='cuda:3'), covar=tensor([0.2983, 0.2465, 0.1173, 0.1456, 0.2896, 0.1178, 0.2439, 0.2189], + device='cuda:3'), in_proj_covar=tensor([0.0724, 0.0734, 0.0612, 0.0859, 0.0731, 0.0636, 0.0760, 0.0658], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 12:01:41,833 INFO [train.py:903] (3/4) Epoch 8, batch 2150, loss[loss=0.2166, simple_loss=0.2885, pruned_loss=0.07239, over 19591.00 frames. ], tot_loss[loss=0.2564, simple_loss=0.3244, pruned_loss=0.09423, over 3818387.76 frames. ], batch size: 52, lr: 1.06e-02, grad_scale: 8.0 +2023-04-01 12:01:48,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.004e+02 5.997e+02 7.086e+02 8.659e+02 2.224e+03, threshold=1.417e+03, percent-clipped=8.0 +2023-04-01 12:01:49,703 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0312, 2.0379, 1.6656, 1.5576, 1.4513, 1.6512, 0.3541, 0.8565], + device='cuda:3'), covar=tensor([0.0292, 0.0285, 0.0218, 0.0329, 0.0641, 0.0358, 0.0592, 0.0556], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0307, 0.0308, 0.0327, 0.0397, 0.0319, 0.0288, 0.0306], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 12:02:06,582 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2849, 1.3696, 1.7023, 1.4888, 2.4818, 2.2935, 2.5529, 0.9695], + device='cuda:3'), covar=tensor([0.1838, 0.3216, 0.1900, 0.1505, 0.1088, 0.1439, 0.1195, 0.2947], + device='cuda:3'), in_proj_covar=tensor([0.0454, 0.0530, 0.0526, 0.0411, 0.0567, 0.0461, 0.0628, 0.0459], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 12:02:12,069 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:02:41,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49994.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:02:43,587 INFO [train.py:903] (3/4) Epoch 8, batch 2200, loss[loss=0.2407, simple_loss=0.3127, pruned_loss=0.08436, over 19755.00 frames. ], tot_loss[loss=0.2569, simple_loss=0.3246, pruned_loss=0.09457, over 3806462.62 frames. ], batch size: 54, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:03:48,613 INFO [train.py:903] (3/4) Epoch 8, batch 2250, loss[loss=0.2341, simple_loss=0.2971, pruned_loss=0.08552, over 19794.00 frames. ], tot_loss[loss=0.2576, simple_loss=0.3253, pruned_loss=0.09492, over 3819770.54 frames. ], batch size: 49, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:03:51,223 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1252, 5.5011, 2.9400, 4.8580, 1.1840, 5.3156, 5.4088, 5.5625], + device='cuda:3'), covar=tensor([0.0374, 0.0812, 0.1825, 0.0608, 0.3949, 0.0661, 0.0553, 0.0911], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0338, 0.0398, 0.0294, 0.0360, 0.0323, 0.0312, 0.0353], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 12:03:55,466 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.410e+02 6.238e+02 7.806e+02 1.014e+03 2.092e+03, threshold=1.561e+03, percent-clipped=8.0 +2023-04-01 12:03:58,802 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2725, 3.8425, 2.4529, 3.4797, 1.3326, 3.5384, 3.6004, 3.6765], + device='cuda:3'), covar=tensor([0.0676, 0.1139, 0.2130, 0.0740, 0.3536, 0.0921, 0.0745, 0.1047], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0339, 0.0398, 0.0294, 0.0360, 0.0323, 0.0313, 0.0353], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 12:04:51,010 INFO [train.py:903] (3/4) Epoch 8, batch 2300, loss[loss=0.2246, simple_loss=0.3058, pruned_loss=0.07176, over 19535.00 frames. ], tot_loss[loss=0.2574, simple_loss=0.3252, pruned_loss=0.09476, over 3805349.94 frames. ], batch size: 54, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:05:03,056 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2859, 0.9256, 1.1628, 2.0568, 1.6773, 1.1425, 1.8291, 1.2328], + device='cuda:3'), covar=tensor([0.1241, 0.1930, 0.1511, 0.1043, 0.1072, 0.1590, 0.1160, 0.1247], + device='cuda:3'), in_proj_covar=tensor([0.0217, 0.0232, 0.0229, 0.0258, 0.0245, 0.0217, 0.0208, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 12:05:05,019 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 12:05:11,322 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2475, 2.9546, 2.1650, 2.1623, 1.8220, 2.4983, 0.7865, 1.9928], + device='cuda:3'), covar=tensor([0.0313, 0.0330, 0.0366, 0.0615, 0.0725, 0.0571, 0.0763, 0.0648], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0312, 0.0312, 0.0328, 0.0403, 0.0324, 0.0293, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 12:05:31,277 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=50128.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:05:37,140 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2888, 1.4109, 1.9566, 1.6142, 2.9628, 4.6168, 4.5822, 4.9797], + device='cuda:3'), covar=tensor([0.1485, 0.3082, 0.2709, 0.1841, 0.0485, 0.0142, 0.0152, 0.0088], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0284, 0.0312, 0.0247, 0.0208, 0.0135, 0.0204, 0.0171], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 12:05:52,929 INFO [train.py:903] (3/4) Epoch 8, batch 2350, loss[loss=0.2518, simple_loss=0.3248, pruned_loss=0.08942, over 17380.00 frames. ], tot_loss[loss=0.2559, simple_loss=0.324, pruned_loss=0.09394, over 3810719.14 frames. ], batch size: 101, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:05:53,263 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50146.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:05:59,868 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.785e+02 5.938e+02 7.771e+02 9.106e+02 1.869e+03, threshold=1.554e+03, percent-clipped=2.0 +2023-04-01 12:06:03,818 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50155.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:06:18,093 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50165.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:06:36,731 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50180.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:06:38,694 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 12:06:47,997 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50190.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:06:54,684 INFO [train.py:903] (3/4) Epoch 8, batch 2400, loss[loss=0.2596, simple_loss=0.3227, pruned_loss=0.09823, over 19736.00 frames. ], tot_loss[loss=0.255, simple_loss=0.3236, pruned_loss=0.09319, over 3824262.52 frames. ], batch size: 63, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:06:54,690 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 12:07:20,948 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50215.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:07:54,905 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50243.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:07:58,092 INFO [train.py:903] (3/4) Epoch 8, batch 2450, loss[loss=0.223, simple_loss=0.2955, pruned_loss=0.07523, over 19853.00 frames. ], tot_loss[loss=0.2551, simple_loss=0.3233, pruned_loss=0.09346, over 3812428.15 frames. ], batch size: 52, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:08:05,246 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.846e+02 6.354e+02 7.339e+02 9.160e+02 2.255e+03, threshold=1.468e+03, percent-clipped=3.0 +2023-04-01 12:09:00,867 INFO [train.py:903] (3/4) Epoch 8, batch 2500, loss[loss=0.3078, simple_loss=0.3667, pruned_loss=0.1245, over 18816.00 frames. ], tot_loss[loss=0.2553, simple_loss=0.3234, pruned_loss=0.09355, over 3809628.63 frames. ], batch size: 74, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:10:03,021 INFO [train.py:903] (3/4) Epoch 8, batch 2550, loss[loss=0.2596, simple_loss=0.3363, pruned_loss=0.09142, over 19734.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3229, pruned_loss=0.09303, over 3808662.65 frames. ], batch size: 63, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:10:09,520 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.329e+02 5.479e+02 6.803e+02 8.076e+02 1.672e+03, threshold=1.361e+03, percent-clipped=2.0 +2023-04-01 12:10:39,026 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-01 12:10:59,277 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 12:11:05,094 INFO [train.py:903] (3/4) Epoch 8, batch 2600, loss[loss=0.2042, simple_loss=0.2725, pruned_loss=0.06796, over 19293.00 frames. ], tot_loss[loss=0.255, simple_loss=0.3231, pruned_loss=0.0934, over 3821546.85 frames. ], batch size: 44, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:11:45,197 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 12:12:09,318 INFO [train.py:903] (3/4) Epoch 8, batch 2650, loss[loss=0.2075, simple_loss=0.2772, pruned_loss=0.06896, over 19799.00 frames. ], tot_loss[loss=0.2539, simple_loss=0.3221, pruned_loss=0.09288, over 3822465.88 frames. ], batch size: 47, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:12:15,976 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.098e+02 6.836e+02 8.198e+02 1.046e+03 1.620e+03, threshold=1.640e+03, percent-clipped=8.0 +2023-04-01 12:12:27,568 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 12:13:04,663 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=50490.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:13:11,332 INFO [train.py:903] (3/4) Epoch 8, batch 2700, loss[loss=0.2635, simple_loss=0.3189, pruned_loss=0.104, over 18594.00 frames. ], tot_loss[loss=0.2546, simple_loss=0.3225, pruned_loss=0.09334, over 3818908.51 frames. ], batch size: 41, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:13:16,521 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50499.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:13:47,859 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50524.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:14:15,291 INFO [train.py:903] (3/4) Epoch 8, batch 2750, loss[loss=0.275, simple_loss=0.3349, pruned_loss=0.1075, over 19669.00 frames. ], tot_loss[loss=0.2547, simple_loss=0.3224, pruned_loss=0.09355, over 3836363.12 frames. ], batch size: 58, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:14:23,786 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.883e+02 5.912e+02 7.181e+02 9.047e+02 1.864e+03, threshold=1.436e+03, percent-clipped=1.0 +2023-04-01 12:14:30,883 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=50559.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:14:37,797 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50564.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:15:18,012 INFO [train.py:903] (3/4) Epoch 8, batch 2800, loss[loss=0.2132, simple_loss=0.2914, pruned_loss=0.0675, over 19587.00 frames. ], tot_loss[loss=0.2536, simple_loss=0.3216, pruned_loss=0.09282, over 3844056.59 frames. ], batch size: 52, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:15:29,739 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50605.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:16:22,000 INFO [train.py:903] (3/4) Epoch 8, batch 2850, loss[loss=0.2137, simple_loss=0.2818, pruned_loss=0.07282, over 19624.00 frames. ], tot_loss[loss=0.2525, simple_loss=0.3204, pruned_loss=0.09236, over 3835853.25 frames. ], batch size: 50, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:16:31,184 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.810e+02 5.617e+02 7.073e+02 8.787e+02 1.544e+03, threshold=1.415e+03, percent-clipped=2.0 +2023-04-01 12:16:57,660 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50674.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:17:25,977 INFO [train.py:903] (3/4) Epoch 8, batch 2900, loss[loss=0.2999, simple_loss=0.3539, pruned_loss=0.1229, over 19540.00 frames. ], tot_loss[loss=0.2527, simple_loss=0.3208, pruned_loss=0.09234, over 3833467.87 frames. ], batch size: 56, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:17:26,041 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 12:18:29,363 INFO [train.py:903] (3/4) Epoch 8, batch 2950, loss[loss=0.1871, simple_loss=0.2679, pruned_loss=0.0531, over 19717.00 frames. ], tot_loss[loss=0.2525, simple_loss=0.321, pruned_loss=0.09205, over 3835410.90 frames. ], batch size: 45, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:18:37,518 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.796e+02 6.074e+02 7.951e+02 1.027e+03 2.467e+03, threshold=1.590e+03, percent-clipped=7.0 +2023-04-01 12:19:31,423 INFO [train.py:903] (3/4) Epoch 8, batch 3000, loss[loss=0.3039, simple_loss=0.3578, pruned_loss=0.125, over 18279.00 frames. ], tot_loss[loss=0.2546, simple_loss=0.3222, pruned_loss=0.0935, over 3811301.21 frames. ], batch size: 83, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:19:31,424 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 12:19:44,058 INFO [train.py:937] (3/4) Epoch 8, validation: loss=0.1875, simple_loss=0.2879, pruned_loss=0.04358, over 944034.00 frames. +2023-04-01 12:19:44,059 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 12:19:46,436 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 12:19:47,996 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9791, 1.7602, 1.5622, 1.9589, 1.9003, 1.7778, 1.4935, 1.8648], + device='cuda:3'), covar=tensor([0.0857, 0.1349, 0.1318, 0.0897, 0.0993, 0.0482, 0.1147, 0.0585], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0355, 0.0286, 0.0241, 0.0297, 0.0239, 0.0272, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 12:19:49,169 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50800.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:20:21,829 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50826.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:20:45,719 INFO [train.py:903] (3/4) Epoch 8, batch 3050, loss[loss=0.2367, simple_loss=0.3172, pruned_loss=0.0781, over 18630.00 frames. ], tot_loss[loss=0.256, simple_loss=0.3235, pruned_loss=0.09424, over 3820975.31 frames. ], batch size: 74, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:20:55,179 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.189e+02 5.734e+02 7.199e+02 9.163e+02 1.650e+03, threshold=1.440e+03, percent-clipped=2.0 +2023-04-01 12:21:05,857 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3503, 1.2596, 1.6225, 1.4251, 3.1912, 2.8641, 3.3298, 1.2772], + device='cuda:3'), covar=tensor([0.1863, 0.3316, 0.2115, 0.1501, 0.1180, 0.1305, 0.1359, 0.3083], + device='cuda:3'), in_proj_covar=tensor([0.0469, 0.0540, 0.0538, 0.0422, 0.0577, 0.0472, 0.0638, 0.0469], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 12:21:06,915 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50861.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:21:36,839 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50886.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:21:49,959 INFO [train.py:903] (3/4) Epoch 8, batch 3100, loss[loss=0.2549, simple_loss=0.3267, pruned_loss=0.09157, over 19512.00 frames. ], tot_loss[loss=0.2556, simple_loss=0.3233, pruned_loss=0.09396, over 3831143.51 frames. ], batch size: 54, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:22:01,846 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-01 12:22:04,805 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=50908.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:22:32,040 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50930.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:22:52,356 INFO [train.py:903] (3/4) Epoch 8, batch 3150, loss[loss=0.2047, simple_loss=0.2685, pruned_loss=0.07043, over 19342.00 frames. ], tot_loss[loss=0.256, simple_loss=0.3237, pruned_loss=0.09414, over 3815577.85 frames. ], batch size: 44, lr: 1.05e-02, grad_scale: 8.0 +2023-04-01 12:23:00,491 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.154e+02 5.915e+02 7.023e+02 8.955e+02 1.571e+03, threshold=1.405e+03, percent-clipped=4.0 +2023-04-01 12:23:03,162 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50955.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:23:14,726 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50964.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:23:20,357 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 12:23:54,308 INFO [train.py:903] (3/4) Epoch 8, batch 3200, loss[loss=0.2581, simple_loss=0.3228, pruned_loss=0.09669, over 17972.00 frames. ], tot_loss[loss=0.2563, simple_loss=0.3245, pruned_loss=0.09412, over 3828679.05 frames. ], batch size: 83, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:24:30,164 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51023.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:24:57,557 INFO [train.py:903] (3/4) Epoch 8, batch 3250, loss[loss=0.2352, simple_loss=0.2961, pruned_loss=0.08715, over 19739.00 frames. ], tot_loss[loss=0.2549, simple_loss=0.323, pruned_loss=0.09341, over 3831126.02 frames. ], batch size: 46, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:25:05,775 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.083e+02 6.087e+02 7.883e+02 9.942e+02 3.174e+03, threshold=1.577e+03, percent-clipped=7.0 +2023-04-01 12:25:40,485 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51080.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:26:00,812 INFO [train.py:903] (3/4) Epoch 8, batch 3300, loss[loss=0.2499, simple_loss=0.3266, pruned_loss=0.08654, over 19486.00 frames. ], tot_loss[loss=0.2553, simple_loss=0.3232, pruned_loss=0.09366, over 3825524.33 frames. ], batch size: 64, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:26:08,880 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 12:27:02,450 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51144.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:27:04,793 INFO [train.py:903] (3/4) Epoch 8, batch 3350, loss[loss=0.2292, simple_loss=0.3042, pruned_loss=0.07709, over 19687.00 frames. ], tot_loss[loss=0.2563, simple_loss=0.3241, pruned_loss=0.09426, over 3822716.99 frames. ], batch size: 53, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:27:12,718 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.478e+02 5.942e+02 7.377e+02 9.279e+02 2.136e+03, threshold=1.475e+03, percent-clipped=2.0 +2023-04-01 12:27:34,026 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51170.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:27:47,826 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2103, 1.3246, 1.1879, 1.0492, 1.0672, 1.0496, 0.0398, 0.3863], + device='cuda:3'), covar=tensor([0.0336, 0.0359, 0.0224, 0.0271, 0.0773, 0.0301, 0.0644, 0.0583], + device='cuda:3'), in_proj_covar=tensor([0.0321, 0.0318, 0.0315, 0.0331, 0.0408, 0.0333, 0.0294, 0.0314], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 12:28:00,625 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51190.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:28:07,186 INFO [train.py:903] (3/4) Epoch 8, batch 3400, loss[loss=0.3194, simple_loss=0.3653, pruned_loss=0.1367, over 13235.00 frames. ], tot_loss[loss=0.2558, simple_loss=0.3235, pruned_loss=0.09408, over 3818489.49 frames. ], batch size: 136, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:29:05,386 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-01 12:29:10,668 INFO [train.py:903] (3/4) Epoch 8, batch 3450, loss[loss=0.2476, simple_loss=0.318, pruned_loss=0.08866, over 19667.00 frames. ], tot_loss[loss=0.2539, simple_loss=0.322, pruned_loss=0.09286, over 3832422.56 frames. ], batch size: 55, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:29:12,351 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.67 vs. limit=2.0 +2023-04-01 12:29:16,192 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 12:29:18,574 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.243e+02 6.280e+02 7.477e+02 9.686e+02 1.820e+03, threshold=1.495e+03, percent-clipped=3.0 +2023-04-01 12:29:27,099 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51259.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:29:52,061 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51279.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:30:00,108 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51285.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:30:12,377 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 12:30:12,716 INFO [train.py:903] (3/4) Epoch 8, batch 3500, loss[loss=0.319, simple_loss=0.3722, pruned_loss=0.1329, over 19332.00 frames. ], tot_loss[loss=0.2549, simple_loss=0.3226, pruned_loss=0.09365, over 3799123.73 frames. ], batch size: 66, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:30:25,524 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51304.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:30:30,163 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51308.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:30:31,845 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-04-01 12:31:19,057 INFO [train.py:903] (3/4) Epoch 8, batch 3550, loss[loss=0.2934, simple_loss=0.3486, pruned_loss=0.1191, over 19664.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3225, pruned_loss=0.09324, over 3809304.62 frames. ], batch size: 55, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:31:27,383 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.001e+02 5.209e+02 6.396e+02 8.561e+02 1.899e+03, threshold=1.279e+03, percent-clipped=3.0 +2023-04-01 12:31:47,285 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9096, 1.9391, 2.0648, 2.8425, 1.8257, 2.6013, 2.5487, 1.8941], + device='cuda:3'), covar=tensor([0.3081, 0.2471, 0.1135, 0.1447, 0.2893, 0.1116, 0.2448, 0.2204], + device='cuda:3'), in_proj_covar=tensor([0.0726, 0.0734, 0.0612, 0.0858, 0.0731, 0.0641, 0.0749, 0.0662], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 12:32:21,072 INFO [train.py:903] (3/4) Epoch 8, batch 3600, loss[loss=0.2947, simple_loss=0.3577, pruned_loss=0.1158, over 17193.00 frames. ], tot_loss[loss=0.2548, simple_loss=0.3227, pruned_loss=0.09345, over 3815202.16 frames. ], batch size: 101, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:32:26,096 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51400.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:32:38,822 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51410.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:32:55,884 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51423.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:32:56,816 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51424.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:33:22,342 INFO [train.py:903] (3/4) Epoch 8, batch 3650, loss[loss=0.2954, simple_loss=0.3522, pruned_loss=0.1193, over 19570.00 frames. ], tot_loss[loss=0.2554, simple_loss=0.3233, pruned_loss=0.09369, over 3809807.61 frames. ], batch size: 61, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:33:31,510 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.306e+02 6.415e+02 7.779e+02 9.918e+02 2.619e+03, threshold=1.556e+03, percent-clipped=14.0 +2023-04-01 12:34:24,456 INFO [train.py:903] (3/4) Epoch 8, batch 3700, loss[loss=0.2439, simple_loss=0.328, pruned_loss=0.07995, over 18796.00 frames. ], tot_loss[loss=0.2553, simple_loss=0.3233, pruned_loss=0.0936, over 3812293.52 frames. ], batch size: 74, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:34:46,878 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51513.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:34:49,322 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51515.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:35:12,270 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51534.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:35:18,982 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51539.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:35:20,156 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51540.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:35:22,213 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51541.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:35:28,602 INFO [train.py:903] (3/4) Epoch 8, batch 3750, loss[loss=0.2456, simple_loss=0.3251, pruned_loss=0.08308, over 18050.00 frames. ], tot_loss[loss=0.2542, simple_loss=0.3227, pruned_loss=0.09287, over 3816174.53 frames. ], batch size: 83, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:35:36,648 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.985e+02 5.907e+02 7.282e+02 9.270e+02 2.268e+03, threshold=1.456e+03, percent-clipped=4.0 +2023-04-01 12:35:52,365 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51566.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:36:30,186 INFO [train.py:903] (3/4) Epoch 8, batch 3800, loss[loss=0.2144, simple_loss=0.2856, pruned_loss=0.07158, over 19829.00 frames. ], tot_loss[loss=0.2533, simple_loss=0.3218, pruned_loss=0.09234, over 3831364.90 frames. ], batch size: 52, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:37:02,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 12:37:31,373 INFO [train.py:903] (3/4) Epoch 8, batch 3850, loss[loss=0.3165, simple_loss=0.3608, pruned_loss=0.1361, over 13287.00 frames. ], tot_loss[loss=0.2532, simple_loss=0.3215, pruned_loss=0.09246, over 3826946.98 frames. ], batch size: 136, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:37:35,041 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51649.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:37:40,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.568e+02 6.155e+02 7.716e+02 1.023e+03 2.199e+03, threshold=1.543e+03, percent-clipped=8.0 +2023-04-01 12:38:13,479 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51679.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:38:33,777 INFO [train.py:903] (3/4) Epoch 8, batch 3900, loss[loss=0.2974, simple_loss=0.3548, pruned_loss=0.12, over 13190.00 frames. ], tot_loss[loss=0.2523, simple_loss=0.3209, pruned_loss=0.09183, over 3829126.69 frames. ], batch size: 136, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:38:46,001 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51704.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:39:33,761 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51744.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:39:37,153 INFO [train.py:903] (3/4) Epoch 8, batch 3950, loss[loss=0.2502, simple_loss=0.3146, pruned_loss=0.09286, over 19835.00 frames. ], tot_loss[loss=0.2519, simple_loss=0.3201, pruned_loss=0.09184, over 3827554.08 frames. ], batch size: 52, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:39:41,712 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 12:39:45,226 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.212e+02 5.815e+02 7.280e+02 9.203e+02 2.422e+03, threshold=1.456e+03, percent-clipped=4.0 +2023-04-01 12:39:46,690 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51754.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:39:59,579 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51765.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:40:37,996 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51795.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:40:38,724 INFO [train.py:903] (3/4) Epoch 8, batch 4000, loss[loss=0.2246, simple_loss=0.286, pruned_loss=0.08165, over 19727.00 frames. ], tot_loss[loss=0.2514, simple_loss=0.3195, pruned_loss=0.09169, over 3824096.44 frames. ], batch size: 46, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:41:09,883 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51820.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:41:27,824 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 12:41:41,697 INFO [train.py:903] (3/4) Epoch 8, batch 4050, loss[loss=0.2638, simple_loss=0.3369, pruned_loss=0.09538, over 18755.00 frames. ], tot_loss[loss=0.2507, simple_loss=0.3192, pruned_loss=0.09107, over 3829237.44 frames. ], batch size: 74, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:41:50,762 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.576e+02 5.742e+02 7.614e+02 9.901e+02 2.045e+03, threshold=1.523e+03, percent-clipped=5.0 +2023-04-01 12:41:56,460 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=51857.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:41:59,731 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51859.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:42:11,969 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51869.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:42:43,874 INFO [train.py:903] (3/4) Epoch 8, batch 4100, loss[loss=0.256, simple_loss=0.3219, pruned_loss=0.09506, over 19525.00 frames. ], tot_loss[loss=0.2498, simple_loss=0.3184, pruned_loss=0.09062, over 3836181.99 frames. ], batch size: 54, lr: 1.04e-02, grad_scale: 8.0 +2023-04-01 12:42:56,252 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51905.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:43:21,454 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 12:43:26,474 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51930.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 12:43:39,203 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-04-01 12:43:47,787 INFO [train.py:903] (3/4) Epoch 8, batch 4150, loss[loss=0.2369, simple_loss=0.3152, pruned_loss=0.07934, over 19616.00 frames. ], tot_loss[loss=0.2506, simple_loss=0.3191, pruned_loss=0.09109, over 3828599.97 frames. ], batch size: 61, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:43:56,803 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.129e+02 6.343e+02 7.798e+02 9.790e+02 2.215e+03, threshold=1.560e+03, percent-clipped=4.0 +2023-04-01 12:44:19,707 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51972.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:44:50,672 INFO [train.py:903] (3/4) Epoch 8, batch 4200, loss[loss=0.2626, simple_loss=0.3307, pruned_loss=0.09723, over 19499.00 frames. ], tot_loss[loss=0.2511, simple_loss=0.3195, pruned_loss=0.09138, over 3819427.85 frames. ], batch size: 64, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:44:57,654 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 12:45:04,800 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1810, 4.2827, 4.7501, 4.7013, 2.6929, 4.3976, 4.0452, 4.4321], + device='cuda:3'), covar=tensor([0.0946, 0.1936, 0.0434, 0.0461, 0.3339, 0.0527, 0.0437, 0.0800], + device='cuda:3'), in_proj_covar=tensor([0.0600, 0.0526, 0.0712, 0.0601, 0.0666, 0.0461, 0.0450, 0.0661], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 12:45:53,324 INFO [train.py:903] (3/4) Epoch 8, batch 4250, loss[loss=0.2319, simple_loss=0.3162, pruned_loss=0.07378, over 19550.00 frames. ], tot_loss[loss=0.2517, simple_loss=0.3205, pruned_loss=0.09144, over 3832911.08 frames. ], batch size: 56, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:45:57,238 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1444, 1.8169, 1.3658, 1.1919, 1.6121, 0.9873, 1.0920, 1.7131], + device='cuda:3'), covar=tensor([0.0629, 0.0533, 0.0886, 0.0595, 0.0384, 0.1013, 0.0546, 0.0314], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0290, 0.0314, 0.0241, 0.0226, 0.0308, 0.0284, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 12:46:01,321 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.676e+02 5.455e+02 6.472e+02 8.916e+02 2.597e+03, threshold=1.294e+03, percent-clipped=4.0 +2023-04-01 12:46:08,494 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52058.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:46:11,508 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 12:46:18,993 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-01 12:46:21,912 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 12:46:55,487 INFO [train.py:903] (3/4) Epoch 8, batch 4300, loss[loss=0.2447, simple_loss=0.3149, pruned_loss=0.08724, over 19626.00 frames. ], tot_loss[loss=0.2503, simple_loss=0.3194, pruned_loss=0.09064, over 3836145.90 frames. ], batch size: 50, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:47:15,125 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=52109.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:47:22,353 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52115.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:47:31,689 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5895, 1.9695, 2.3981, 2.5819, 2.2780, 2.2852, 2.2911, 2.5459], + device='cuda:3'), covar=tensor([0.0657, 0.1513, 0.0912, 0.0795, 0.1051, 0.0382, 0.0801, 0.0478], + device='cuda:3'), in_proj_covar=tensor([0.0247, 0.0350, 0.0282, 0.0234, 0.0295, 0.0238, 0.0267, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 12:47:34,156 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52125.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:47:50,787 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 12:47:53,339 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52140.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:48:00,441 INFO [train.py:903] (3/4) Epoch 8, batch 4350, loss[loss=0.2315, simple_loss=0.2963, pruned_loss=0.08339, over 19387.00 frames. ], tot_loss[loss=0.2504, simple_loss=0.3194, pruned_loss=0.09071, over 3823020.87 frames. ], batch size: 48, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:48:06,452 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52150.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:48:09,411 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.508e+02 5.672e+02 7.291e+02 9.101e+02 1.997e+03, threshold=1.458e+03, percent-clipped=8.0 +2023-04-01 12:48:48,791 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-04-01 12:49:03,039 INFO [train.py:903] (3/4) Epoch 8, batch 4400, loss[loss=0.2423, simple_loss=0.3078, pruned_loss=0.08842, over 19695.00 frames. ], tot_loss[loss=0.2489, simple_loss=0.3184, pruned_loss=0.08972, over 3839684.29 frames. ], batch size: 53, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:49:21,182 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52211.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:49:26,762 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 12:49:38,219 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 12:49:38,563 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52224.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:49:43,333 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52228.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:50:01,068 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-04-01 12:50:05,247 INFO [train.py:903] (3/4) Epoch 8, batch 4450, loss[loss=0.2502, simple_loss=0.3231, pruned_loss=0.08867, over 19754.00 frames. ], tot_loss[loss=0.2509, simple_loss=0.3199, pruned_loss=0.09094, over 3831205.49 frames. ], batch size: 63, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:50:13,311 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.227e+02 5.880e+02 7.086e+02 8.839e+02 1.936e+03, threshold=1.417e+03, percent-clipped=3.0 +2023-04-01 12:50:13,759 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52253.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:50:14,760 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0314, 1.8902, 1.9531, 2.2430, 2.1917, 1.9170, 2.0439, 2.0939], + device='cuda:3'), covar=tensor([0.0730, 0.1298, 0.0980, 0.0629, 0.0811, 0.0412, 0.0782, 0.0508], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0351, 0.0284, 0.0233, 0.0298, 0.0241, 0.0270, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 12:51:06,390 INFO [train.py:903] (3/4) Epoch 8, batch 4500, loss[loss=0.3025, simple_loss=0.3593, pruned_loss=0.1228, over 18800.00 frames. ], tot_loss[loss=0.2524, simple_loss=0.3214, pruned_loss=0.0917, over 3825991.00 frames. ], batch size: 74, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:51:50,145 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52330.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:52:10,352 INFO [train.py:903] (3/4) Epoch 8, batch 4550, loss[loss=0.2656, simple_loss=0.3379, pruned_loss=0.09669, over 19663.00 frames. ], tot_loss[loss=0.2519, simple_loss=0.321, pruned_loss=0.09144, over 3815897.95 frames. ], batch size: 58, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:52:10,749 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7528, 1.3777, 1.2758, 1.6252, 1.4784, 1.5187, 1.3263, 1.5632], + device='cuda:3'), covar=tensor([0.0866, 0.1351, 0.1316, 0.0856, 0.1027, 0.0511, 0.1088, 0.0670], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0351, 0.0286, 0.0234, 0.0297, 0.0241, 0.0270, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 12:52:18,694 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.675e+02 5.922e+02 7.010e+02 8.869e+02 1.679e+03, threshold=1.402e+03, percent-clipped=2.0 +2023-04-01 12:52:18,730 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 12:52:27,451 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-01 12:52:41,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 12:53:11,441 INFO [train.py:903] (3/4) Epoch 8, batch 4600, loss[loss=0.2813, simple_loss=0.3499, pruned_loss=0.1063, over 18690.00 frames. ], tot_loss[loss=0.2535, simple_loss=0.3222, pruned_loss=0.09247, over 3811273.26 frames. ], batch size: 74, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:53:18,628 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=52402.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:53:19,886 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2256, 1.3116, 1.1812, 1.0118, 1.0048, 1.0354, 0.0273, 0.3587], + device='cuda:3'), covar=tensor([0.0354, 0.0349, 0.0217, 0.0314, 0.0772, 0.0292, 0.0699, 0.0591], + device='cuda:3'), in_proj_covar=tensor([0.0322, 0.0321, 0.0316, 0.0337, 0.0413, 0.0331, 0.0301, 0.0317], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 12:54:12,886 INFO [train.py:903] (3/4) Epoch 8, batch 4650, loss[loss=0.3253, simple_loss=0.3771, pruned_loss=0.1367, over 13882.00 frames. ], tot_loss[loss=0.2525, simple_loss=0.3212, pruned_loss=0.09186, over 3823372.27 frames. ], batch size: 135, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:54:21,262 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.655e+02 5.664e+02 6.903e+02 8.285e+02 1.576e+03, threshold=1.381e+03, percent-clipped=2.0 +2023-04-01 12:54:26,501 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8514, 1.8904, 1.9813, 2.4335, 1.6767, 2.2134, 2.3250, 1.9354], + device='cuda:3'), covar=tensor([0.2807, 0.2154, 0.1181, 0.1296, 0.2486, 0.1151, 0.2697, 0.2067], + device='cuda:3'), in_proj_covar=tensor([0.0722, 0.0734, 0.0607, 0.0861, 0.0726, 0.0636, 0.0754, 0.0659], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 12:54:30,512 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 12:54:42,678 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 12:54:56,945 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52480.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:55:15,249 INFO [train.py:903] (3/4) Epoch 8, batch 4700, loss[loss=0.2028, simple_loss=0.2733, pruned_loss=0.0661, over 19270.00 frames. ], tot_loss[loss=0.2524, simple_loss=0.3212, pruned_loss=0.09183, over 3820790.26 frames. ], batch size: 44, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:55:27,909 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52505.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:55:28,210 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.68 vs. limit=5.0 +2023-04-01 12:55:39,741 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 12:55:41,526 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 12:55:43,531 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52517.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:56:18,361 INFO [train.py:903] (3/4) Epoch 8, batch 4750, loss[loss=0.2589, simple_loss=0.3406, pruned_loss=0.08856, over 19541.00 frames. ], tot_loss[loss=0.2523, simple_loss=0.3212, pruned_loss=0.09168, over 3824044.56 frames. ], batch size: 54, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:56:29,688 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.555e+02 6.316e+02 7.348e+02 9.529e+02 1.491e+03, threshold=1.470e+03, percent-clipped=3.0 +2023-04-01 12:56:31,064 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=52555.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:56:52,284 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52573.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 12:57:22,250 INFO [train.py:903] (3/4) Epoch 8, batch 4800, loss[loss=0.2219, simple_loss=0.2836, pruned_loss=0.08009, over 19728.00 frames. ], tot_loss[loss=0.2526, simple_loss=0.3209, pruned_loss=0.09213, over 3826476.79 frames. ], batch size: 46, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:58:22,769 INFO [train.py:903] (3/4) Epoch 8, batch 4850, loss[loss=0.234, simple_loss=0.3021, pruned_loss=0.08297, over 19469.00 frames. ], tot_loss[loss=0.2513, simple_loss=0.3197, pruned_loss=0.09143, over 3826450.83 frames. ], batch size: 49, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:58:32,081 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.583e+02 6.021e+02 7.604e+02 9.872e+02 2.114e+03, threshold=1.521e+03, percent-clipped=8.0 +2023-04-01 12:58:45,976 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 12:58:52,949 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52670.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:58:58,365 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=52674.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 12:59:08,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 12:59:14,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 12:59:14,127 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 12:59:23,225 INFO [train.py:903] (3/4) Epoch 8, batch 4900, loss[loss=0.2864, simple_loss=0.3512, pruned_loss=0.1108, over 18159.00 frames. ], tot_loss[loss=0.2507, simple_loss=0.3194, pruned_loss=0.09098, over 3824671.03 frames. ], batch size: 83, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 12:59:24,407 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 12:59:44,292 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 12:59:49,955 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8036, 4.4020, 2.5797, 3.9548, 1.1213, 4.0234, 4.1372, 4.2107], + device='cuda:3'), covar=tensor([0.0583, 0.1069, 0.1969, 0.0759, 0.3790, 0.0745, 0.0714, 0.0841], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0349, 0.0407, 0.0302, 0.0367, 0.0333, 0.0327, 0.0364], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:00:22,661 INFO [train.py:903] (3/4) Epoch 8, batch 4950, loss[loss=0.3017, simple_loss=0.3585, pruned_loss=0.1225, over 13132.00 frames. ], tot_loss[loss=0.2526, simple_loss=0.321, pruned_loss=0.09212, over 3819884.60 frames. ], batch size: 136, lr: 1.03e-02, grad_scale: 8.0 +2023-04-01 13:00:25,766 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2032, 3.8143, 2.5482, 3.4707, 1.0277, 3.5229, 3.5789, 3.7028], + device='cuda:3'), covar=tensor([0.0691, 0.1027, 0.1952, 0.0728, 0.3917, 0.0871, 0.0753, 0.1029], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0346, 0.0406, 0.0300, 0.0365, 0.0331, 0.0325, 0.0362], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:00:35,702 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.348e+02 6.374e+02 8.178e+02 1.048e+03 2.702e+03, threshold=1.636e+03, percent-clipped=11.0 +2023-04-01 13:00:40,340 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 13:00:53,324 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52770.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:00:56,957 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52773.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:01:04,931 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 13:01:17,721 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52789.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:01:26,623 INFO [train.py:903] (3/4) Epoch 8, batch 5000, loss[loss=0.2499, simple_loss=0.3239, pruned_loss=0.08799, over 19764.00 frames. ], tot_loss[loss=0.2537, simple_loss=0.3223, pruned_loss=0.09257, over 3826205.65 frames. ], batch size: 56, lr: 1.03e-02, grad_scale: 4.0 +2023-04-01 13:01:29,351 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52798.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:01:35,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 13:01:42,313 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-04-01 13:01:47,281 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 13:01:51,287 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7161, 3.1429, 3.2247, 3.1986, 1.2028, 3.0176, 2.7135, 2.9607], + device='cuda:3'), covar=tensor([0.1345, 0.0815, 0.0784, 0.0841, 0.4517, 0.0716, 0.0709, 0.1263], + device='cuda:3'), in_proj_covar=tensor([0.0597, 0.0524, 0.0715, 0.0601, 0.0666, 0.0464, 0.0449, 0.0661], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 13:02:28,532 INFO [train.py:903] (3/4) Epoch 8, batch 5050, loss[loss=0.2546, simple_loss=0.3359, pruned_loss=0.08664, over 19527.00 frames. ], tot_loss[loss=0.2529, simple_loss=0.3217, pruned_loss=0.09203, over 3829073.44 frames. ], batch size: 56, lr: 1.03e-02, grad_scale: 4.0 +2023-04-01 13:02:39,034 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.635e+02 5.648e+02 7.062e+02 8.811e+02 1.795e+03, threshold=1.412e+03, percent-clipped=2.0 +2023-04-01 13:02:51,282 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.78 vs. limit=5.0 +2023-04-01 13:03:05,006 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 13:03:15,919 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.74 vs. limit=5.0 +2023-04-01 13:03:30,516 INFO [train.py:903] (3/4) Epoch 8, batch 5100, loss[loss=0.2976, simple_loss=0.3548, pruned_loss=0.1202, over 19077.00 frames. ], tot_loss[loss=0.2534, simple_loss=0.3219, pruned_loss=0.09244, over 3837296.43 frames. ], batch size: 69, lr: 1.03e-02, grad_scale: 4.0 +2023-04-01 13:03:36,923 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 13:03:41,012 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 13:03:45,419 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 13:03:50,847 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 13:03:59,150 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=52917.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 13:04:09,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52926.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:04:32,217 INFO [train.py:903] (3/4) Epoch 8, batch 5150, loss[loss=0.2559, simple_loss=0.3221, pruned_loss=0.09488, over 19588.00 frames. ], tot_loss[loss=0.2534, simple_loss=0.3217, pruned_loss=0.09252, over 3822706.26 frames. ], batch size: 52, lr: 1.03e-02, grad_scale: 4.0 +2023-04-01 13:04:41,050 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52951.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:04:45,912 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.391e+02 6.060e+02 7.983e+02 1.041e+03 2.368e+03, threshold=1.597e+03, percent-clipped=6.0 +2023-04-01 13:04:47,183 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 13:04:59,549 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-01 13:05:19,599 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 13:05:27,526 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1808, 1.7850, 1.7624, 2.0287, 1.9493, 1.9434, 1.7123, 1.9859], + device='cuda:3'), covar=tensor([0.0840, 0.1441, 0.1234, 0.0958, 0.1145, 0.0446, 0.1086, 0.0599], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0355, 0.0289, 0.0240, 0.0301, 0.0242, 0.0274, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 13:05:37,043 INFO [train.py:903] (3/4) Epoch 8, batch 5200, loss[loss=0.292, simple_loss=0.3666, pruned_loss=0.1087, over 19531.00 frames. ], tot_loss[loss=0.2541, simple_loss=0.3225, pruned_loss=0.09288, over 3817431.47 frames. ], batch size: 56, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:05:51,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 13:05:54,952 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8076, 4.3059, 2.5349, 3.8109, 1.3180, 4.0526, 4.1039, 4.1886], + device='cuda:3'), covar=tensor([0.0590, 0.1092, 0.2063, 0.0779, 0.3543, 0.0798, 0.0732, 0.0963], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0340, 0.0400, 0.0296, 0.0356, 0.0326, 0.0320, 0.0356], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:06:21,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53032.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 13:06:36,721 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 13:06:38,343 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53045.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:06:39,057 INFO [train.py:903] (3/4) Epoch 8, batch 5250, loss[loss=0.2754, simple_loss=0.3297, pruned_loss=0.1106, over 19851.00 frames. ], tot_loss[loss=0.2531, simple_loss=0.3217, pruned_loss=0.09225, over 3825361.85 frames. ], batch size: 52, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:06:49,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.283e+02 5.973e+02 7.081e+02 8.822e+02 3.028e+03, threshold=1.416e+03, percent-clipped=2.0 +2023-04-01 13:06:51,073 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-01 13:07:08,274 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53070.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:07:13,810 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53074.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:07:39,282 INFO [train.py:903] (3/4) Epoch 8, batch 5300, loss[loss=0.2502, simple_loss=0.3292, pruned_loss=0.08555, over 19106.00 frames. ], tot_loss[loss=0.2536, simple_loss=0.322, pruned_loss=0.09257, over 3810368.28 frames. ], batch size: 69, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:07:57,461 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 13:08:03,088 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53114.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:08:15,992 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7650, 1.4517, 1.3820, 1.5917, 1.7300, 1.3867, 1.3953, 1.5459], + device='cuda:3'), covar=tensor([0.0978, 0.1683, 0.1567, 0.1219, 0.1317, 0.0894, 0.1413, 0.0873], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0354, 0.0286, 0.0238, 0.0298, 0.0241, 0.0273, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 13:08:41,250 INFO [train.py:903] (3/4) Epoch 8, batch 5350, loss[loss=0.2487, simple_loss=0.3225, pruned_loss=0.08743, over 19625.00 frames. ], tot_loss[loss=0.2524, simple_loss=0.321, pruned_loss=0.09186, over 3823535.58 frames. ], batch size: 57, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:08:52,777 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.210e+02 6.176e+02 7.478e+02 9.376e+02 1.338e+03, threshold=1.496e+03, percent-clipped=0.0 +2023-04-01 13:09:18,558 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 13:09:44,017 INFO [train.py:903] (3/4) Epoch 8, batch 5400, loss[loss=0.2258, simple_loss=0.3002, pruned_loss=0.07568, over 19627.00 frames. ], tot_loss[loss=0.252, simple_loss=0.3209, pruned_loss=0.09158, over 3827691.66 frames. ], batch size: 50, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:10:24,915 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53229.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:10:47,203 INFO [train.py:903] (3/4) Epoch 8, batch 5450, loss[loss=0.2384, simple_loss=0.3059, pruned_loss=0.08541, over 19419.00 frames. ], tot_loss[loss=0.253, simple_loss=0.3221, pruned_loss=0.09196, over 3839183.88 frames. ], batch size: 48, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:10:57,345 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.737e+02 5.768e+02 7.150e+02 9.218e+02 2.127e+03, threshold=1.430e+03, percent-clipped=3.0 +2023-04-01 13:11:02,407 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53259.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:11:40,444 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53288.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 13:11:45,052 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-01 13:11:48,914 INFO [train.py:903] (3/4) Epoch 8, batch 5500, loss[loss=0.264, simple_loss=0.3136, pruned_loss=0.1072, over 19799.00 frames. ], tot_loss[loss=0.252, simple_loss=0.3207, pruned_loss=0.09161, over 3828451.07 frames. ], batch size: 48, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:12:11,209 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53313.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 13:12:16,579 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 13:12:23,266 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53322.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:12:35,605 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53333.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:12:50,184 INFO [train.py:903] (3/4) Epoch 8, batch 5550, loss[loss=0.2301, simple_loss=0.2981, pruned_loss=0.08107, over 19581.00 frames. ], tot_loss[loss=0.2533, simple_loss=0.3222, pruned_loss=0.09225, over 3830679.23 frames. ], batch size: 52, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:12:59,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 13:13:03,077 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.436e+02 6.394e+02 7.821e+02 9.803e+02 2.197e+03, threshold=1.564e+03, percent-clipped=2.0 +2023-04-01 13:13:40,394 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53385.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:13:50,361 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 13:13:53,936 INFO [train.py:903] (3/4) Epoch 8, batch 5600, loss[loss=0.2279, simple_loss=0.3101, pruned_loss=0.07285, over 19702.00 frames. ], tot_loss[loss=0.252, simple_loss=0.3207, pruned_loss=0.09165, over 3837611.12 frames. ], batch size: 59, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:14:21,449 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53418.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:14:24,159 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9948, 1.9444, 2.1170, 2.7777, 1.8580, 2.5790, 2.4984, 2.0140], + device='cuda:3'), covar=tensor([0.3022, 0.2634, 0.1285, 0.1436, 0.2907, 0.1188, 0.2771, 0.2237], + device='cuda:3'), in_proj_covar=tensor([0.0729, 0.0741, 0.0611, 0.0863, 0.0735, 0.0643, 0.0763, 0.0660], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:14:28,532 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8439, 1.4689, 1.4859, 1.8704, 3.3333, 1.2670, 2.1784, 3.5957], + device='cuda:3'), covar=tensor([0.0374, 0.2322, 0.2453, 0.1332, 0.0587, 0.1940, 0.1202, 0.0278], + device='cuda:3'), in_proj_covar=tensor([0.0326, 0.0325, 0.0336, 0.0301, 0.0328, 0.0317, 0.0306, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 13:14:57,497 INFO [train.py:903] (3/4) Epoch 8, batch 5650, loss[loss=0.1952, simple_loss=0.2714, pruned_loss=0.0595, over 19381.00 frames. ], tot_loss[loss=0.2513, simple_loss=0.3205, pruned_loss=0.09109, over 3837255.32 frames. ], batch size: 47, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:15:07,837 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.837e+02 5.961e+02 7.306e+02 9.270e+02 2.985e+03, threshold=1.461e+03, percent-clipped=1.0 +2023-04-01 13:15:45,797 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 13:15:46,280 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53485.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:15:58,772 INFO [train.py:903] (3/4) Epoch 8, batch 5700, loss[loss=0.2158, simple_loss=0.2781, pruned_loss=0.07678, over 19748.00 frames. ], tot_loss[loss=0.25, simple_loss=0.3186, pruned_loss=0.09065, over 3833905.16 frames. ], batch size: 47, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:16:15,430 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:16:44,944 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53533.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:16:59,703 INFO [train.py:903] (3/4) Epoch 8, batch 5750, loss[loss=0.339, simple_loss=0.388, pruned_loss=0.145, over 18042.00 frames. ], tot_loss[loss=0.2496, simple_loss=0.3187, pruned_loss=0.09022, over 3847509.51 frames. ], batch size: 83, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:17:00,955 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 13:17:10,453 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 13:17:11,664 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.475e+02 5.699e+02 6.647e+02 8.184e+02 1.829e+03, threshold=1.329e+03, percent-clipped=1.0 +2023-04-01 13:17:15,870 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 13:17:19,443 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6405, 1.4681, 1.4814, 2.0604, 1.5723, 2.0003, 2.0105, 1.7594], + device='cuda:3'), covar=tensor([0.0834, 0.0924, 0.0960, 0.0843, 0.0924, 0.0654, 0.0933, 0.0639], + device='cuda:3'), in_proj_covar=tensor([0.0215, 0.0228, 0.0225, 0.0255, 0.0242, 0.0211, 0.0206, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 13:18:02,116 INFO [train.py:903] (3/4) Epoch 8, batch 5800, loss[loss=0.215, simple_loss=0.2821, pruned_loss=0.07396, over 19789.00 frames. ], tot_loss[loss=0.2495, simple_loss=0.3186, pruned_loss=0.09017, over 3845053.93 frames. ], batch size: 47, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:18:12,762 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53603.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:19:04,521 INFO [train.py:903] (3/4) Epoch 8, batch 5850, loss[loss=0.2608, simple_loss=0.3302, pruned_loss=0.09565, over 18788.00 frames. ], tot_loss[loss=0.2483, simple_loss=0.3179, pruned_loss=0.0894, over 3849993.84 frames. ], batch size: 74, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:19:15,051 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.702e+02 6.371e+02 7.746e+02 9.220e+02 2.993e+03, threshold=1.549e+03, percent-clipped=10.0 +2023-04-01 13:19:27,647 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53666.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:19:42,003 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53677.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:20:05,079 INFO [train.py:903] (3/4) Epoch 8, batch 5900, loss[loss=0.248, simple_loss=0.3268, pruned_loss=0.08454, over 19463.00 frames. ], tot_loss[loss=0.2504, simple_loss=0.3192, pruned_loss=0.09078, over 3831967.07 frames. ], batch size: 64, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:20:09,611 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 13:20:30,262 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 13:20:34,009 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53718.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:20:47,028 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=53729.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:21:06,831 INFO [train.py:903] (3/4) Epoch 8, batch 5950, loss[loss=0.1965, simple_loss=0.2728, pruned_loss=0.06013, over 19363.00 frames. ], tot_loss[loss=0.2509, simple_loss=0.3195, pruned_loss=0.09118, over 3825200.35 frames. ], batch size: 47, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:21:19,042 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.652e+02 5.886e+02 7.195e+02 1.025e+03 2.007e+03, threshold=1.439e+03, percent-clipped=3.0 +2023-04-01 13:21:33,094 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-01 13:21:47,825 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-01 13:21:50,846 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53781.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:22:00,165 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53789.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:22:04,758 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53792.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:22:09,975 INFO [train.py:903] (3/4) Epoch 8, batch 6000, loss[loss=0.2177, simple_loss=0.2881, pruned_loss=0.07368, over 19792.00 frames. ], tot_loss[loss=0.2485, simple_loss=0.3176, pruned_loss=0.08974, over 3822619.10 frames. ], batch size: 49, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:22:09,975 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 13:22:22,635 INFO [train.py:937] (3/4) Epoch 8, validation: loss=0.1864, simple_loss=0.2865, pruned_loss=0.04314, over 944034.00 frames. +2023-04-01 13:22:22,636 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 13:22:29,222 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1310, 1.2932, 1.7926, 1.2673, 2.8403, 3.8433, 3.5670, 4.0374], + device='cuda:3'), covar=tensor([0.1510, 0.3229, 0.2829, 0.1907, 0.0427, 0.0125, 0.0179, 0.0139], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0289, 0.0317, 0.0246, 0.0208, 0.0140, 0.0205, 0.0175], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 13:22:48,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53814.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:23:24,824 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53844.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:23:26,630 INFO [train.py:903] (3/4) Epoch 8, batch 6050, loss[loss=0.3216, simple_loss=0.3643, pruned_loss=0.1395, over 13366.00 frames. ], tot_loss[loss=0.2486, simple_loss=0.3179, pruned_loss=0.08971, over 3823237.17 frames. ], batch size: 136, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:23:39,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.317e+02 5.583e+02 7.013e+02 9.917e+02 2.418e+03, threshold=1.403e+03, percent-clipped=8.0 +2023-04-01 13:24:11,899 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3287, 1.4406, 1.9265, 1.6319, 2.5579, 2.1468, 2.7180, 1.1945], + device='cuda:3'), covar=tensor([0.2173, 0.3646, 0.2091, 0.1733, 0.1448, 0.1889, 0.1538, 0.3437], + device='cuda:3'), in_proj_covar=tensor([0.0465, 0.0539, 0.0546, 0.0416, 0.0574, 0.0472, 0.0635, 0.0472], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:24:18,426 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1537, 5.5053, 3.1110, 4.7702, 0.9160, 5.2537, 5.3997, 5.5641], + device='cuda:3'), covar=tensor([0.0400, 0.0742, 0.1635, 0.0629, 0.4144, 0.0636, 0.0706, 0.0751], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0345, 0.0408, 0.0305, 0.0370, 0.0334, 0.0327, 0.0364], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:24:30,152 INFO [train.py:903] (3/4) Epoch 8, batch 6100, loss[loss=0.2106, simple_loss=0.2761, pruned_loss=0.07256, over 19731.00 frames. ], tot_loss[loss=0.2483, simple_loss=0.3173, pruned_loss=0.08963, over 3830599.98 frames. ], batch size: 45, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:24:48,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1674, 1.1979, 1.0849, 0.9337, 0.9664, 1.0465, 0.0710, 0.3364], + device='cuda:3'), covar=tensor([0.0394, 0.0369, 0.0252, 0.0316, 0.0736, 0.0307, 0.0673, 0.0628], + device='cuda:3'), in_proj_covar=tensor([0.0324, 0.0318, 0.0318, 0.0334, 0.0411, 0.0336, 0.0294, 0.0316], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 13:25:07,691 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-04-01 13:25:22,434 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2205, 2.2564, 2.2046, 3.2341, 2.0977, 3.1890, 2.8419, 2.1028], + device='cuda:3'), covar=tensor([0.3136, 0.2478, 0.1191, 0.1540, 0.3158, 0.1081, 0.2623, 0.2170], + device='cuda:3'), in_proj_covar=tensor([0.0720, 0.0735, 0.0609, 0.0856, 0.0735, 0.0642, 0.0758, 0.0658], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:25:31,558 INFO [train.py:903] (3/4) Epoch 8, batch 6150, loss[loss=0.254, simple_loss=0.3251, pruned_loss=0.09141, over 19300.00 frames. ], tot_loss[loss=0.2486, simple_loss=0.318, pruned_loss=0.08963, over 3826194.99 frames. ], batch size: 66, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:25:42,179 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.923e+02 5.703e+02 7.303e+02 8.849e+02 1.874e+03, threshold=1.461e+03, percent-clipped=4.0 +2023-04-01 13:25:56,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 13:26:07,957 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53974.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:26:12,673 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0148, 1.7000, 2.0558, 1.9462, 4.4432, 0.8502, 2.4044, 4.7661], + device='cuda:3'), covar=tensor([0.0315, 0.2485, 0.2355, 0.1709, 0.0719, 0.2829, 0.1308, 0.0211], + device='cuda:3'), in_proj_covar=tensor([0.0326, 0.0327, 0.0337, 0.0307, 0.0332, 0.0322, 0.0315, 0.0327], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 13:26:33,376 INFO [train.py:903] (3/4) Epoch 8, batch 6200, loss[loss=0.2191, simple_loss=0.2999, pruned_loss=0.06916, over 19680.00 frames. ], tot_loss[loss=0.2491, simple_loss=0.3186, pruned_loss=0.08981, over 3836141.61 frames. ], batch size: 53, lr: 1.02e-02, grad_scale: 8.0 +2023-04-01 13:26:38,487 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53999.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:27:25,521 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=54037.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:27:37,603 INFO [train.py:903] (3/4) Epoch 8, batch 6250, loss[loss=0.2493, simple_loss=0.3252, pruned_loss=0.08667, over 19650.00 frames. ], tot_loss[loss=0.2489, simple_loss=0.3184, pruned_loss=0.08969, over 3831473.77 frames. ], batch size: 53, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:27:41,238 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=54048.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:27:49,336 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.888e+02 5.723e+02 7.068e+02 9.572e+02 2.133e+03, threshold=1.414e+03, percent-clipped=6.0 +2023-04-01 13:27:58,062 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=54062.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:28:04,596 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 13:28:10,774 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=54073.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:28:40,890 INFO [train.py:903] (3/4) Epoch 8, batch 6300, loss[loss=0.2307, simple_loss=0.2951, pruned_loss=0.0831, over 19473.00 frames. ], tot_loss[loss=0.2499, simple_loss=0.3192, pruned_loss=0.09028, over 3827998.46 frames. ], batch size: 49, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:28:45,605 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=54100.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:29:05,554 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.82 vs. limit=5.0 +2023-04-01 13:29:16,191 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=54125.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:29:41,764 INFO [train.py:903] (3/4) Epoch 8, batch 6350, loss[loss=0.219, simple_loss=0.2927, pruned_loss=0.0727, over 19624.00 frames. ], tot_loss[loss=0.2513, simple_loss=0.3207, pruned_loss=0.09097, over 3833282.56 frames. ], batch size: 50, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:29:52,050 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.072e+02 6.055e+02 7.520e+02 8.667e+02 2.456e+03, threshold=1.504e+03, percent-clipped=3.0 +2023-04-01 13:30:03,340 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1643, 1.2413, 1.8944, 1.4043, 2.5485, 2.1356, 2.7422, 1.0494], + device='cuda:3'), covar=tensor([0.2253, 0.3815, 0.2027, 0.1739, 0.1390, 0.1806, 0.1501, 0.3554], + device='cuda:3'), in_proj_covar=tensor([0.0472, 0.0548, 0.0552, 0.0421, 0.0579, 0.0475, 0.0642, 0.0476], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:30:21,619 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1784, 1.1374, 1.4206, 1.1503, 2.4617, 3.1251, 2.9420, 3.3757], + device='cuda:3'), covar=tensor([0.1716, 0.4338, 0.3927, 0.2237, 0.0611, 0.0253, 0.0309, 0.0230], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0287, 0.0317, 0.0247, 0.0209, 0.0140, 0.0206, 0.0175], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 13:30:36,876 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2437, 1.3647, 1.8527, 1.4735, 2.5778, 2.2025, 2.7203, 0.9816], + device='cuda:3'), covar=tensor([0.2050, 0.3520, 0.1952, 0.1637, 0.1365, 0.1705, 0.1385, 0.3487], + device='cuda:3'), in_proj_covar=tensor([0.0469, 0.0546, 0.0552, 0.0421, 0.0578, 0.0475, 0.0640, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:30:43,433 INFO [train.py:903] (3/4) Epoch 8, batch 6400, loss[loss=0.2397, simple_loss=0.3165, pruned_loss=0.08141, over 19628.00 frames. ], tot_loss[loss=0.2511, simple_loss=0.3204, pruned_loss=0.09084, over 3821472.12 frames. ], batch size: 57, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:31:45,992 INFO [train.py:903] (3/4) Epoch 8, batch 6450, loss[loss=0.2251, simple_loss=0.3063, pruned_loss=0.07192, over 19326.00 frames. ], tot_loss[loss=0.249, simple_loss=0.3184, pruned_loss=0.08975, over 3819177.52 frames. ], batch size: 66, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:31:54,381 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-04-01 13:31:54,410 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.92 vs. limit=5.0 +2023-04-01 13:31:55,091 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7148, 4.1916, 2.5916, 3.8231, 1.1170, 3.9706, 4.0046, 4.2017], + device='cuda:3'), covar=tensor([0.0598, 0.1168, 0.2118, 0.0752, 0.4120, 0.0822, 0.0744, 0.1029], + device='cuda:3'), in_proj_covar=tensor([0.0404, 0.0345, 0.0408, 0.0302, 0.0370, 0.0331, 0.0327, 0.0361], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:31:58,337 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.099e+02 5.571e+02 6.761e+02 8.500e+02 1.702e+03, threshold=1.352e+03, percent-clipped=3.0 +2023-04-01 13:32:30,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 13:32:48,752 INFO [train.py:903] (3/4) Epoch 8, batch 6500, loss[loss=0.2308, simple_loss=0.3115, pruned_loss=0.07501, over 19717.00 frames. ], tot_loss[loss=0.251, simple_loss=0.3201, pruned_loss=0.09093, over 3809937.56 frames. ], batch size: 63, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:32:54,456 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 13:33:50,188 INFO [train.py:903] (3/4) Epoch 8, batch 6550, loss[loss=0.2338, simple_loss=0.3046, pruned_loss=0.08153, over 19870.00 frames. ], tot_loss[loss=0.252, simple_loss=0.3208, pruned_loss=0.0916, over 3805462.29 frames. ], batch size: 52, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:34:00,244 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 13:34:00,552 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.237e+02 6.329e+02 8.151e+02 1.084e+03 2.341e+03, threshold=1.630e+03, percent-clipped=12.0 +2023-04-01 13:34:13,780 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-04-01 13:34:51,146 INFO [train.py:903] (3/4) Epoch 8, batch 6600, loss[loss=0.2059, simple_loss=0.2741, pruned_loss=0.06887, over 19726.00 frames. ], tot_loss[loss=0.2514, simple_loss=0.3205, pruned_loss=0.09117, over 3802124.77 frames. ], batch size: 46, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:35:53,846 INFO [train.py:903] (3/4) Epoch 8, batch 6650, loss[loss=0.2191, simple_loss=0.2929, pruned_loss=0.0726, over 19609.00 frames. ], tot_loss[loss=0.2518, simple_loss=0.321, pruned_loss=0.09132, over 3809071.41 frames. ], batch size: 50, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:36:04,916 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.746e+02 5.865e+02 7.808e+02 1.010e+03 1.907e+03, threshold=1.562e+03, percent-clipped=2.0 +2023-04-01 13:36:55,565 INFO [train.py:903] (3/4) Epoch 8, batch 6700, loss[loss=0.2144, simple_loss=0.2855, pruned_loss=0.07164, over 19767.00 frames. ], tot_loss[loss=0.2499, simple_loss=0.3194, pruned_loss=0.09022, over 3816991.53 frames. ], batch size: 48, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:37:14,058 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-01 13:37:52,326 INFO [train.py:903] (3/4) Epoch 8, batch 6750, loss[loss=0.2739, simple_loss=0.3414, pruned_loss=0.1032, over 19505.00 frames. ], tot_loss[loss=0.2495, simple_loss=0.3192, pruned_loss=0.08988, over 3824390.47 frames. ], batch size: 64, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:38:03,620 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.992e+02 6.380e+02 7.224e+02 9.353e+02 2.017e+03, threshold=1.445e+03, percent-clipped=3.0 +2023-04-01 13:38:50,509 INFO [train.py:903] (3/4) Epoch 8, batch 6800, loss[loss=0.2586, simple_loss=0.3334, pruned_loss=0.09184, over 19546.00 frames. ], tot_loss[loss=0.2495, simple_loss=0.3191, pruned_loss=0.08991, over 3818280.63 frames. ], batch size: 56, lr: 1.01e-02, grad_scale: 8.0 +2023-04-01 13:39:34,066 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 13:39:35,158 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 13:39:38,465 INFO [train.py:903] (3/4) Epoch 9, batch 0, loss[loss=0.2179, simple_loss=0.3023, pruned_loss=0.06675, over 19351.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.3023, pruned_loss=0.06675, over 19351.00 frames. ], batch size: 70, lr: 9.56e-03, grad_scale: 8.0 +2023-04-01 13:39:38,465 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 13:39:49,511 INFO [train.py:937] (3/4) Epoch 9, validation: loss=0.1866, simple_loss=0.2872, pruned_loss=0.04294, over 944034.00 frames. +2023-04-01 13:39:49,512 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 13:40:03,819 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 13:40:28,247 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.879e+02 5.500e+02 7.208e+02 8.930e+02 1.459e+03, threshold=1.442e+03, percent-clipped=1.0 +2023-04-01 13:40:51,457 INFO [train.py:903] (3/4) Epoch 9, batch 50, loss[loss=0.2658, simple_loss=0.3314, pruned_loss=0.1001, over 19652.00 frames. ], tot_loss[loss=0.2534, simple_loss=0.3225, pruned_loss=0.09217, over 866084.89 frames. ], batch size: 55, lr: 9.55e-03, grad_scale: 8.0 +2023-04-01 13:41:02,967 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=54682.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 13:41:26,440 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 13:41:53,038 INFO [train.py:903] (3/4) Epoch 9, batch 100, loss[loss=0.299, simple_loss=0.3502, pruned_loss=0.1239, over 13181.00 frames. ], tot_loss[loss=0.2517, simple_loss=0.3201, pruned_loss=0.09164, over 1515338.42 frames. ], batch size: 136, lr: 9.55e-03, grad_scale: 8.0 +2023-04-01 13:42:05,417 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 13:42:18,629 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3235, 1.3545, 1.5647, 1.5103, 2.3134, 2.0674, 2.3253, 0.9221], + device='cuda:3'), covar=tensor([0.2076, 0.3727, 0.2216, 0.1658, 0.1271, 0.1767, 0.1277, 0.3375], + device='cuda:3'), in_proj_covar=tensor([0.0472, 0.0550, 0.0556, 0.0423, 0.0580, 0.0475, 0.0643, 0.0476], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:42:31,166 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 5.564e+02 7.190e+02 9.001e+02 2.877e+03, threshold=1.438e+03, percent-clipped=2.0 +2023-04-01 13:42:53,296 INFO [train.py:903] (3/4) Epoch 9, batch 150, loss[loss=0.2749, simple_loss=0.3468, pruned_loss=0.1015, over 19664.00 frames. ], tot_loss[loss=0.2493, simple_loss=0.3186, pruned_loss=0.08998, over 2026082.66 frames. ], batch size: 58, lr: 9.54e-03, grad_scale: 16.0 +2023-04-01 13:43:05,260 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3086, 1.4650, 1.7964, 1.5444, 2.6520, 2.2365, 2.7598, 1.0530], + device='cuda:3'), covar=tensor([0.2037, 0.3446, 0.2026, 0.1598, 0.1261, 0.1679, 0.1233, 0.3323], + device='cuda:3'), in_proj_covar=tensor([0.0474, 0.0549, 0.0556, 0.0424, 0.0583, 0.0477, 0.0642, 0.0479], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:43:53,897 INFO [train.py:903] (3/4) Epoch 9, batch 200, loss[loss=0.2354, simple_loss=0.3077, pruned_loss=0.08156, over 19515.00 frames. ], tot_loss[loss=0.2496, simple_loss=0.3189, pruned_loss=0.09018, over 2425898.73 frames. ], batch size: 54, lr: 9.54e-03, grad_scale: 8.0 +2023-04-01 13:43:56,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 13:43:57,440 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0279, 5.3346, 3.0235, 4.7528, 1.1185, 5.2389, 5.2675, 5.4763], + device='cuda:3'), covar=tensor([0.0411, 0.0850, 0.1736, 0.0580, 0.3960, 0.0585, 0.0615, 0.0643], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0339, 0.0406, 0.0300, 0.0365, 0.0330, 0.0326, 0.0358], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 13:44:36,458 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.193e+02 5.883e+02 7.738e+02 9.204e+02 1.688e+03, threshold=1.548e+03, percent-clipped=2.0 +2023-04-01 13:44:57,178 INFO [train.py:903] (3/4) Epoch 9, batch 250, loss[loss=0.2927, simple_loss=0.3538, pruned_loss=0.1158, over 19056.00 frames. ], tot_loss[loss=0.2512, simple_loss=0.3198, pruned_loss=0.09127, over 2723575.79 frames. ], batch size: 69, lr: 9.54e-03, grad_scale: 8.0 +2023-04-01 13:45:06,583 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=54880.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:45:39,341 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1565, 3.6554, 2.1370, 1.9374, 3.2217, 1.6868, 1.3878, 2.0973], + device='cuda:3'), covar=tensor([0.1007, 0.0376, 0.0773, 0.0721, 0.0415, 0.0989, 0.0881, 0.0574], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0290, 0.0321, 0.0241, 0.0228, 0.0310, 0.0287, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 13:45:57,840 INFO [train.py:903] (3/4) Epoch 9, batch 300, loss[loss=0.3085, simple_loss=0.3613, pruned_loss=0.1279, over 13085.00 frames. ], tot_loss[loss=0.251, simple_loss=0.3191, pruned_loss=0.09143, over 2963446.27 frames. ], batch size: 135, lr: 9.53e-03, grad_scale: 8.0 +2023-04-01 13:46:39,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.328e+02 5.660e+02 7.032e+02 9.457e+02 2.087e+03, threshold=1.406e+03, percent-clipped=3.0 +2023-04-01 13:47:01,341 INFO [train.py:903] (3/4) Epoch 9, batch 350, loss[loss=0.2666, simple_loss=0.3285, pruned_loss=0.1023, over 18150.00 frames. ], tot_loss[loss=0.2503, simple_loss=0.319, pruned_loss=0.0908, over 3162766.09 frames. ], batch size: 83, lr: 9.53e-03, grad_scale: 8.0 +2023-04-01 13:47:07,268 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 13:48:03,062 INFO [train.py:903] (3/4) Epoch 9, batch 400, loss[loss=0.2248, simple_loss=0.3123, pruned_loss=0.06862, over 19643.00 frames. ], tot_loss[loss=0.2496, simple_loss=0.3188, pruned_loss=0.09025, over 3322451.49 frames. ], batch size: 59, lr: 9.52e-03, grad_scale: 8.0 +2023-04-01 13:48:06,566 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=55026.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 13:48:38,448 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-04-01 13:48:44,520 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.195e+02 5.719e+02 7.898e+02 1.018e+03 2.327e+03, threshold=1.580e+03, percent-clipped=4.0 +2023-04-01 13:49:04,315 INFO [train.py:903] (3/4) Epoch 9, batch 450, loss[loss=0.2397, simple_loss=0.302, pruned_loss=0.08868, over 19404.00 frames. ], tot_loss[loss=0.248, simple_loss=0.3176, pruned_loss=0.08922, over 3439373.35 frames. ], batch size: 48, lr: 9.52e-03, grad_scale: 8.0 +2023-04-01 13:49:42,327 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 13:49:43,518 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 13:50:00,657 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6495, 1.7014, 1.6210, 2.6124, 1.6730, 2.3819, 2.0393, 1.3468], + device='cuda:3'), covar=tensor([0.3889, 0.3249, 0.2060, 0.1836, 0.3314, 0.1483, 0.3968, 0.3726], + device='cuda:3'), in_proj_covar=tensor([0.0741, 0.0750, 0.0622, 0.0870, 0.0750, 0.0657, 0.0771, 0.0675], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:50:07,769 INFO [train.py:903] (3/4) Epoch 9, batch 500, loss[loss=0.2445, simple_loss=0.3144, pruned_loss=0.0873, over 19849.00 frames. ], tot_loss[loss=0.2484, simple_loss=0.318, pruned_loss=0.08941, over 3529250.04 frames. ], batch size: 52, lr: 9.51e-03, grad_scale: 8.0 +2023-04-01 13:50:30,917 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55141.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 13:50:47,935 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.663e+02 6.183e+02 7.261e+02 8.870e+02 1.589e+03, threshold=1.452e+03, percent-clipped=1.0 +2023-04-01 13:51:00,606 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9498, 3.3690, 3.6936, 3.7180, 1.3140, 3.2927, 3.0978, 3.1766], + device='cuda:3'), covar=tensor([0.2209, 0.1590, 0.1244, 0.1283, 0.6580, 0.1339, 0.1136, 0.2319], + device='cuda:3'), in_proj_covar=tensor([0.0617, 0.0545, 0.0733, 0.0613, 0.0667, 0.0475, 0.0462, 0.0671], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 13:51:01,581 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55166.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:51:04,360 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 13:51:10,562 INFO [train.py:903] (3/4) Epoch 9, batch 550, loss[loss=0.2544, simple_loss=0.3231, pruned_loss=0.09287, over 18819.00 frames. ], tot_loss[loss=0.249, simple_loss=0.3186, pruned_loss=0.08967, over 3588112.01 frames. ], batch size: 74, lr: 9.51e-03, grad_scale: 8.0 +2023-04-01 13:52:14,553 INFO [train.py:903] (3/4) Epoch 9, batch 600, loss[loss=0.2276, simple_loss=0.3024, pruned_loss=0.07645, over 19853.00 frames. ], tot_loss[loss=0.2493, simple_loss=0.3187, pruned_loss=0.08991, over 3643681.71 frames. ], batch size: 52, lr: 9.51e-03, grad_scale: 8.0 +2023-04-01 13:52:15,869 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=55224.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:52:21,730 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55229.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:52:47,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1119, 1.1381, 1.5278, 1.0711, 2.5257, 3.4782, 3.2719, 3.6966], + device='cuda:3'), covar=tensor([0.1528, 0.3408, 0.3211, 0.2150, 0.0529, 0.0162, 0.0233, 0.0171], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0290, 0.0319, 0.0250, 0.0211, 0.0142, 0.0208, 0.0177], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 13:52:55,192 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.376e+02 5.838e+02 6.783e+02 8.586e+02 3.812e+03, threshold=1.357e+03, percent-clipped=5.0 +2023-04-01 13:52:57,873 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5354, 4.0564, 4.2206, 4.1538, 1.6259, 3.8558, 3.4393, 3.9060], + device='cuda:3'), covar=tensor([0.1151, 0.0645, 0.0482, 0.0547, 0.4203, 0.0596, 0.0601, 0.0926], + device='cuda:3'), in_proj_covar=tensor([0.0619, 0.0543, 0.0732, 0.0613, 0.0668, 0.0476, 0.0462, 0.0671], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 13:52:58,743 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 13:53:13,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1741, 1.2445, 1.1209, 0.9499, 1.0384, 1.0533, 0.0461, 0.2835], + device='cuda:3'), covar=tensor([0.0381, 0.0372, 0.0231, 0.0327, 0.0734, 0.0316, 0.0696, 0.0644], + device='cuda:3'), in_proj_covar=tensor([0.0319, 0.0313, 0.0313, 0.0333, 0.0408, 0.0334, 0.0296, 0.0315], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 13:53:16,398 INFO [train.py:903] (3/4) Epoch 9, batch 650, loss[loss=0.2259, simple_loss=0.3036, pruned_loss=0.07408, over 19667.00 frames. ], tot_loss[loss=0.2488, simple_loss=0.3186, pruned_loss=0.08953, over 3696042.95 frames. ], batch size: 58, lr: 9.50e-03, grad_scale: 4.0 +2023-04-01 13:54:19,295 INFO [train.py:903] (3/4) Epoch 9, batch 700, loss[loss=0.2661, simple_loss=0.339, pruned_loss=0.09657, over 19686.00 frames. ], tot_loss[loss=0.2502, simple_loss=0.3196, pruned_loss=0.09036, over 3712913.47 frames. ], batch size: 59, lr: 9.50e-03, grad_scale: 4.0 +2023-04-01 13:54:41,861 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55339.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:55:02,468 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.468e+02 5.553e+02 6.808e+02 9.255e+02 2.546e+03, threshold=1.362e+03, percent-clipped=4.0 +2023-04-01 13:55:22,969 INFO [train.py:903] (3/4) Epoch 9, batch 750, loss[loss=0.2415, simple_loss=0.3196, pruned_loss=0.08176, over 19522.00 frames. ], tot_loss[loss=0.2502, simple_loss=0.3198, pruned_loss=0.09031, over 3733874.16 frames. ], batch size: 54, lr: 9.49e-03, grad_scale: 4.0 +2023-04-01 13:55:53,603 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55397.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 13:56:25,994 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55422.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 13:56:26,680 INFO [train.py:903] (3/4) Epoch 9, batch 800, loss[loss=0.1961, simple_loss=0.274, pruned_loss=0.05911, over 19469.00 frames. ], tot_loss[loss=0.2507, simple_loss=0.3201, pruned_loss=0.09064, over 3746006.45 frames. ], batch size: 49, lr: 9.49e-03, grad_scale: 8.0 +2023-04-01 13:56:41,969 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 13:57:07,881 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.102e+02 5.631e+02 7.179e+02 9.586e+02 1.610e+03, threshold=1.436e+03, percent-clipped=4.0 +2023-04-01 13:57:28,778 INFO [train.py:903] (3/4) Epoch 9, batch 850, loss[loss=0.2405, simple_loss=0.3083, pruned_loss=0.08639, over 19687.00 frames. ], tot_loss[loss=0.2486, simple_loss=0.3182, pruned_loss=0.08947, over 3758322.23 frames. ], batch size: 53, lr: 9.48e-03, grad_scale: 8.0 +2023-04-01 13:58:10,388 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1087, 2.0838, 2.2185, 2.8002, 2.2281, 2.6712, 2.5121, 2.0246], + device='cuda:3'), covar=tensor([0.2364, 0.1888, 0.0954, 0.1088, 0.2030, 0.0830, 0.1880, 0.1675], + device='cuda:3'), in_proj_covar=tensor([0.0743, 0.0756, 0.0627, 0.0872, 0.0751, 0.0658, 0.0774, 0.0677], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 13:58:14,371 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=55510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:58:22,708 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 13:58:29,596 INFO [train.py:903] (3/4) Epoch 9, batch 900, loss[loss=0.3169, simple_loss=0.3652, pruned_loss=0.1343, over 13211.00 frames. ], tot_loss[loss=0.2496, simple_loss=0.3183, pruned_loss=0.09041, over 3756602.09 frames. ], batch size: 136, lr: 9.48e-03, grad_scale: 8.0 +2023-04-01 13:59:12,476 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.962e+02 5.974e+02 7.143e+02 8.825e+02 2.413e+03, threshold=1.429e+03, percent-clipped=6.0 +2023-04-01 13:59:32,016 INFO [train.py:903] (3/4) Epoch 9, batch 950, loss[loss=0.2864, simple_loss=0.3357, pruned_loss=0.1186, over 19386.00 frames. ], tot_loss[loss=0.2503, simple_loss=0.3191, pruned_loss=0.09076, over 3761058.52 frames. ], batch size: 48, lr: 9.48e-03, grad_scale: 8.0 +2023-04-01 13:59:32,192 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=55573.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 13:59:37,616 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 14:00:01,053 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55595.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:00:32,390 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55620.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:00:35,415 INFO [train.py:903] (3/4) Epoch 9, batch 1000, loss[loss=0.2572, simple_loss=0.3287, pruned_loss=0.09285, over 19405.00 frames. ], tot_loss[loss=0.2493, simple_loss=0.3187, pruned_loss=0.08993, over 3782837.64 frames. ], batch size: 70, lr: 9.47e-03, grad_scale: 8.0 +2023-04-01 14:00:38,011 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55625.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:01:18,080 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.465e+02 5.596e+02 6.834e+02 8.838e+02 1.578e+03, threshold=1.367e+03, percent-clipped=2.0 +2023-04-01 14:01:29,784 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 14:01:39,291 INFO [train.py:903] (3/4) Epoch 9, batch 1050, loss[loss=0.2344, simple_loss=0.3103, pruned_loss=0.0792, over 19526.00 frames. ], tot_loss[loss=0.2479, simple_loss=0.3178, pruned_loss=0.08906, over 3800994.60 frames. ], batch size: 56, lr: 9.47e-03, grad_scale: 8.0 +2023-04-01 14:01:57,386 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55688.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:02:10,953 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 14:02:42,620 INFO [train.py:903] (3/4) Epoch 9, batch 1100, loss[loss=0.2437, simple_loss=0.3183, pruned_loss=0.08452, over 18705.00 frames. ], tot_loss[loss=0.2485, simple_loss=0.3181, pruned_loss=0.08941, over 3798812.33 frames. ], batch size: 74, lr: 9.46e-03, grad_scale: 8.0 +2023-04-01 14:03:25,905 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.690e+02 6.007e+02 7.412e+02 9.315e+02 2.515e+03, threshold=1.482e+03, percent-clipped=6.0 +2023-04-01 14:03:28,849 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-01 14:03:36,831 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55766.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:03:43,631 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2029, 1.7919, 1.4503, 1.2321, 1.6551, 1.1515, 1.2138, 1.6473], + device='cuda:3'), covar=tensor([0.0589, 0.0648, 0.0856, 0.0552, 0.0413, 0.0976, 0.0461, 0.0350], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0290, 0.0319, 0.0237, 0.0228, 0.0313, 0.0285, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 14:03:45,585 INFO [train.py:903] (3/4) Epoch 9, batch 1150, loss[loss=0.2395, simple_loss=0.3182, pruned_loss=0.08039, over 19788.00 frames. ], tot_loss[loss=0.2492, simple_loss=0.3186, pruned_loss=0.08988, over 3805116.55 frames. ], batch size: 56, lr: 9.46e-03, grad_scale: 8.0 +2023-04-01 14:03:46,128 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6449, 1.7366, 1.8916, 2.1425, 1.3849, 1.8816, 2.1384, 1.7542], + device='cuda:3'), covar=tensor([0.3068, 0.2366, 0.1194, 0.1359, 0.2658, 0.1188, 0.2872, 0.2226], + device='cuda:3'), in_proj_covar=tensor([0.0744, 0.0751, 0.0623, 0.0871, 0.0748, 0.0657, 0.0767, 0.0675], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 14:04:14,222 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 14:04:27,005 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8202, 2.1775, 2.3563, 2.6740, 2.6434, 2.3448, 2.2957, 3.0550], + device='cuda:3'), covar=tensor([0.0727, 0.1652, 0.1218, 0.0939, 0.1094, 0.0486, 0.1037, 0.0476], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0353, 0.0291, 0.0240, 0.0298, 0.0244, 0.0271, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 14:04:50,173 INFO [train.py:903] (3/4) Epoch 9, batch 1200, loss[loss=0.2949, simple_loss=0.3506, pruned_loss=0.1196, over 14199.00 frames. ], tot_loss[loss=0.248, simple_loss=0.3173, pruned_loss=0.08935, over 3795339.82 frames. ], batch size: 137, lr: 9.45e-03, grad_scale: 8.0 +2023-04-01 14:05:13,340 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3696, 1.2102, 1.3362, 1.5457, 2.9046, 1.0449, 2.1310, 3.2004], + device='cuda:3'), covar=tensor([0.0408, 0.2575, 0.2615, 0.1625, 0.0710, 0.2343, 0.1126, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0325, 0.0326, 0.0336, 0.0303, 0.0331, 0.0321, 0.0311, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 14:05:18,980 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 14:05:31,642 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.284e+02 5.926e+02 7.645e+02 1.010e+03 3.329e+03, threshold=1.529e+03, percent-clipped=6.0 +2023-04-01 14:05:44,772 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55866.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:05:53,922 INFO [train.py:903] (3/4) Epoch 9, batch 1250, loss[loss=0.241, simple_loss=0.3097, pruned_loss=0.08613, over 19470.00 frames. ], tot_loss[loss=0.2491, simple_loss=0.3183, pruned_loss=0.08999, over 3797813.26 frames. ], batch size: 49, lr: 9.45e-03, grad_scale: 8.0 +2023-04-01 14:06:03,217 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55881.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:06:35,371 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55906.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:06:55,962 INFO [train.py:903] (3/4) Epoch 9, batch 1300, loss[loss=0.2378, simple_loss=0.3151, pruned_loss=0.08024, over 19789.00 frames. ], tot_loss[loss=0.249, simple_loss=0.3182, pruned_loss=0.08986, over 3805218.71 frames. ], batch size: 56, lr: 9.45e-03, grad_scale: 8.0 +2023-04-01 14:07:00,756 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1768, 3.6473, 3.7699, 3.7733, 1.5423, 3.4875, 3.1566, 3.4608], + device='cuda:3'), covar=tensor([0.1290, 0.0964, 0.0646, 0.0598, 0.4425, 0.0704, 0.0630, 0.1086], + device='cuda:3'), in_proj_covar=tensor([0.0620, 0.0544, 0.0738, 0.0618, 0.0675, 0.0481, 0.0466, 0.0678], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 14:07:20,259 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1670, 5.4125, 3.1473, 4.7941, 1.0148, 5.3398, 5.3332, 5.4583], + device='cuda:3'), covar=tensor([0.0415, 0.0824, 0.1650, 0.0612, 0.3916, 0.0637, 0.0634, 0.0865], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0341, 0.0413, 0.0306, 0.0369, 0.0338, 0.0332, 0.0363], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 14:07:24,033 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55944.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:07:39,615 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.793e+02 5.793e+02 6.909e+02 8.321e+02 2.022e+03, threshold=1.382e+03, percent-clipped=2.0 +2023-04-01 14:07:54,311 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:07:58,594 INFO [train.py:903] (3/4) Epoch 9, batch 1350, loss[loss=0.2323, simple_loss=0.3121, pruned_loss=0.07627, over 19535.00 frames. ], tot_loss[loss=0.2496, simple_loss=0.3191, pruned_loss=0.08999, over 3812010.81 frames. ], batch size: 56, lr: 9.44e-03, grad_scale: 8.0 +2023-04-01 14:08:50,497 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-04-01 14:09:02,600 INFO [train.py:903] (3/4) Epoch 9, batch 1400, loss[loss=0.2399, simple_loss=0.3216, pruned_loss=0.07913, over 19594.00 frames. ], tot_loss[loss=0.2487, simple_loss=0.3183, pruned_loss=0.08956, over 3810600.79 frames. ], batch size: 61, lr: 9.44e-03, grad_scale: 8.0 +2023-04-01 14:09:47,082 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.547e+02 5.536e+02 7.314e+02 8.995e+02 2.483e+03, threshold=1.463e+03, percent-clipped=9.0 +2023-04-01 14:10:07,189 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 14:10:08,160 INFO [train.py:903] (3/4) Epoch 9, batch 1450, loss[loss=0.2983, simple_loss=0.3618, pruned_loss=0.1174, over 18043.00 frames. ], tot_loss[loss=0.2485, simple_loss=0.3176, pruned_loss=0.08971, over 3803888.96 frames. ], batch size: 83, lr: 9.43e-03, grad_scale: 8.0 +2023-04-01 14:10:34,605 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7916, 3.2495, 3.3095, 3.2897, 1.3173, 3.0807, 2.7522, 3.0370], + device='cuda:3'), covar=tensor([0.1381, 0.0803, 0.0744, 0.0768, 0.4503, 0.0750, 0.0788, 0.1227], + device='cuda:3'), in_proj_covar=tensor([0.0615, 0.0543, 0.0739, 0.0621, 0.0680, 0.0486, 0.0467, 0.0675], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 14:10:55,222 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=56110.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:11:11,141 INFO [train.py:903] (3/4) Epoch 9, batch 1500, loss[loss=0.311, simple_loss=0.3611, pruned_loss=0.1305, over 17136.00 frames. ], tot_loss[loss=0.2469, simple_loss=0.3164, pruned_loss=0.08872, over 3803494.13 frames. ], batch size: 101, lr: 9.43e-03, grad_scale: 8.0 +2023-04-01 14:11:52,399 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.776e+02 6.008e+02 7.164e+02 9.066e+02 2.093e+03, threshold=1.433e+03, percent-clipped=3.0 +2023-04-01 14:12:01,130 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56163.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:12:12,299 INFO [train.py:903] (3/4) Epoch 9, batch 1550, loss[loss=0.2501, simple_loss=0.318, pruned_loss=0.09108, over 19593.00 frames. ], tot_loss[loss=0.2471, simple_loss=0.3163, pruned_loss=0.08895, over 3812780.05 frames. ], batch size: 52, lr: 9.43e-03, grad_scale: 8.0 +2023-04-01 14:12:33,677 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-01 14:12:37,078 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.99 vs. limit=5.0 +2023-04-01 14:12:59,836 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=56210.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:13:15,311 INFO [train.py:903] (3/4) Epoch 9, batch 1600, loss[loss=0.2646, simple_loss=0.3284, pruned_loss=0.1004, over 19649.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.3161, pruned_loss=0.08832, over 3814821.17 frames. ], batch size: 60, lr: 9.42e-03, grad_scale: 8.0 +2023-04-01 14:13:18,955 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56225.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:13:41,628 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 14:13:59,883 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.177e+02 5.781e+02 6.769e+02 8.617e+02 2.222e+03, threshold=1.354e+03, percent-clipped=2.0 +2023-04-01 14:14:09,534 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-01 14:14:19,008 INFO [train.py:903] (3/4) Epoch 9, batch 1650, loss[loss=0.2871, simple_loss=0.345, pruned_loss=0.1146, over 17160.00 frames. ], tot_loss[loss=0.246, simple_loss=0.3159, pruned_loss=0.08802, over 3820964.52 frames. ], batch size: 101, lr: 9.42e-03, grad_scale: 4.0 +2023-04-01 14:14:27,609 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7018, 3.1608, 3.2296, 3.2633, 1.1713, 3.0768, 2.7226, 2.9463], + device='cuda:3'), covar=tensor([0.1451, 0.0956, 0.0803, 0.0746, 0.4669, 0.0741, 0.0789, 0.1288], + device='cuda:3'), in_proj_covar=tensor([0.0609, 0.0536, 0.0732, 0.0611, 0.0676, 0.0478, 0.0456, 0.0670], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 14:15:22,544 INFO [train.py:903] (3/4) Epoch 9, batch 1700, loss[loss=0.2596, simple_loss=0.3333, pruned_loss=0.093, over 19564.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3158, pruned_loss=0.08774, over 3818809.91 frames. ], batch size: 61, lr: 9.41e-03, grad_scale: 4.0 +2023-04-01 14:15:25,221 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56325.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:16:02,358 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 14:16:05,919 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.642e+02 5.619e+02 6.768e+02 8.904e+02 2.101e+03, threshold=1.354e+03, percent-clipped=6.0 +2023-04-01 14:16:24,618 INFO [train.py:903] (3/4) Epoch 9, batch 1750, loss[loss=0.2584, simple_loss=0.3282, pruned_loss=0.09426, over 19476.00 frames. ], tot_loss[loss=0.2457, simple_loss=0.3158, pruned_loss=0.08781, over 3833092.24 frames. ], batch size: 64, lr: 9.41e-03, grad_scale: 4.0 +2023-04-01 14:17:26,702 INFO [train.py:903] (3/4) Epoch 9, batch 1800, loss[loss=0.1947, simple_loss=0.2697, pruned_loss=0.05984, over 19751.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.3163, pruned_loss=0.08827, over 3830301.14 frames. ], batch size: 46, lr: 9.40e-03, grad_scale: 4.0 +2023-04-01 14:17:29,849 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-01 14:18:09,954 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.374e+02 5.839e+02 7.002e+02 8.564e+02 1.629e+03, threshold=1.400e+03, percent-clipped=1.0 +2023-04-01 14:18:13,456 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9333, 5.2721, 2.9864, 4.5874, 1.3666, 5.1498, 5.2328, 5.4641], + device='cuda:3'), covar=tensor([0.0453, 0.0987, 0.1929, 0.0683, 0.3809, 0.0587, 0.0582, 0.0800], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0342, 0.0411, 0.0305, 0.0371, 0.0338, 0.0332, 0.0364], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 14:18:25,590 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 14:18:29,984 INFO [train.py:903] (3/4) Epoch 9, batch 1850, loss[loss=0.2149, simple_loss=0.2832, pruned_loss=0.07327, over 19767.00 frames. ], tot_loss[loss=0.2465, simple_loss=0.3168, pruned_loss=0.08808, over 3828442.65 frames. ], batch size: 47, lr: 9.40e-03, grad_scale: 4.0 +2023-04-01 14:18:40,532 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56481.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:19:02,314 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 14:19:11,108 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:19:12,041 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=56507.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:19:32,991 INFO [train.py:903] (3/4) Epoch 9, batch 1900, loss[loss=0.2386, simple_loss=0.3094, pruned_loss=0.08394, over 19741.00 frames. ], tot_loss[loss=0.248, simple_loss=0.318, pruned_loss=0.08901, over 3833325.43 frames. ], batch size: 46, lr: 9.40e-03, grad_scale: 4.0 +2023-04-01 14:19:48,436 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 14:19:54,886 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 14:19:55,175 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56541.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 14:20:16,634 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.412e+02 5.665e+02 6.760e+02 8.403e+02 1.758e+03, threshold=1.352e+03, percent-clipped=2.0 +2023-04-01 14:20:18,933 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 14:20:35,966 INFO [train.py:903] (3/4) Epoch 9, batch 1950, loss[loss=0.2091, simple_loss=0.2818, pruned_loss=0.06817, over 19730.00 frames. ], tot_loss[loss=0.2486, simple_loss=0.3187, pruned_loss=0.08919, over 3820859.99 frames. ], batch size: 46, lr: 9.39e-03, grad_scale: 4.0 +2023-04-01 14:20:46,753 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56581.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:21:19,002 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:21:38,531 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56622.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:21:38,885 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-04-01 14:21:39,347 INFO [train.py:903] (3/4) Epoch 9, batch 2000, loss[loss=0.2454, simple_loss=0.3236, pruned_loss=0.08364, over 19615.00 frames. ], tot_loss[loss=0.2484, simple_loss=0.3186, pruned_loss=0.08912, over 3822914.93 frames. ], batch size: 57, lr: 9.39e-03, grad_scale: 8.0 +2023-04-01 14:22:22,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.621e+02 5.506e+02 7.081e+02 9.090e+02 3.144e+03, threshold=1.416e+03, percent-clipped=7.0 +2023-04-01 14:22:36,096 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 14:22:42,524 INFO [train.py:903] (3/4) Epoch 9, batch 2050, loss[loss=0.2343, simple_loss=0.311, pruned_loss=0.07882, over 19662.00 frames. ], tot_loss[loss=0.2506, simple_loss=0.3203, pruned_loss=0.09047, over 3832823.33 frames. ], batch size: 60, lr: 9.38e-03, grad_scale: 8.0 +2023-04-01 14:22:56,335 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 14:22:57,534 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 14:23:19,228 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 14:23:43,572 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3968, 1.2536, 1.2525, 1.8501, 1.4243, 1.6991, 1.8418, 1.5274], + device='cuda:3'), covar=tensor([0.0827, 0.1060, 0.1077, 0.0657, 0.0842, 0.0729, 0.0722, 0.0697], + device='cuda:3'), in_proj_covar=tensor([0.0216, 0.0232, 0.0228, 0.0256, 0.0243, 0.0216, 0.0205, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 14:23:44,278 INFO [train.py:903] (3/4) Epoch 9, batch 2100, loss[loss=0.2815, simple_loss=0.3401, pruned_loss=0.1115, over 19657.00 frames. ], tot_loss[loss=0.2492, simple_loss=0.3189, pruned_loss=0.08974, over 3833023.93 frames. ], batch size: 60, lr: 9.38e-03, grad_scale: 8.0 +2023-04-01 14:24:12,065 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 14:24:29,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.187e+02 5.573e+02 6.906e+02 8.990e+02 1.566e+03, threshold=1.381e+03, percent-clipped=3.0 +2023-04-01 14:24:35,334 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 14:24:48,310 INFO [train.py:903] (3/4) Epoch 9, batch 2150, loss[loss=0.2302, simple_loss=0.2881, pruned_loss=0.08616, over 19308.00 frames. ], tot_loss[loss=0.248, simple_loss=0.3176, pruned_loss=0.08916, over 3832440.14 frames. ], batch size: 44, lr: 9.38e-03, grad_scale: 8.0 +2023-04-01 14:25:48,770 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56821.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:25:50,905 INFO [train.py:903] (3/4) Epoch 9, batch 2200, loss[loss=0.2603, simple_loss=0.3215, pruned_loss=0.09957, over 19319.00 frames. ], tot_loss[loss=0.2474, simple_loss=0.317, pruned_loss=0.08888, over 3834490.71 frames. ], batch size: 66, lr: 9.37e-03, grad_scale: 8.0 +2023-04-01 14:26:36,144 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.021e+02 5.854e+02 7.300e+02 9.690e+02 2.298e+03, threshold=1.460e+03, percent-clipped=5.0 +2023-04-01 14:26:57,253 INFO [train.py:903] (3/4) Epoch 9, batch 2250, loss[loss=0.3095, simple_loss=0.3679, pruned_loss=0.1256, over 19344.00 frames. ], tot_loss[loss=0.2471, simple_loss=0.3171, pruned_loss=0.08854, over 3828455.56 frames. ], batch size: 66, lr: 9.37e-03, grad_scale: 8.0 +2023-04-01 14:27:03,720 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56878.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:27:11,762 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=56885.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 14:27:34,970 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56903.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:28:00,639 INFO [train.py:903] (3/4) Epoch 9, batch 2300, loss[loss=0.2828, simple_loss=0.3466, pruned_loss=0.1095, over 19445.00 frames. ], tot_loss[loss=0.2471, simple_loss=0.3168, pruned_loss=0.08865, over 3824328.43 frames. ], batch size: 62, lr: 9.36e-03, grad_scale: 8.0 +2023-04-01 14:28:05,853 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56926.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:28:13,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 14:28:46,771 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 5.809e+02 7.207e+02 9.233e+02 1.673e+03, threshold=1.441e+03, percent-clipped=4.0 +2023-04-01 14:28:50,682 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3087, 1.2030, 1.8708, 1.5112, 2.9727, 4.4129, 4.3800, 4.9148], + device='cuda:3'), covar=tensor([0.1598, 0.3465, 0.3005, 0.1938, 0.0489, 0.0176, 0.0157, 0.0104], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0289, 0.0319, 0.0247, 0.0208, 0.0144, 0.0204, 0.0178], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 14:29:04,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3795, 2.1541, 1.9038, 1.7777, 1.6322, 1.8422, 0.4434, 1.1812], + device='cuda:3'), covar=tensor([0.0316, 0.0385, 0.0301, 0.0438, 0.0678, 0.0504, 0.0692, 0.0607], + device='cuda:3'), in_proj_covar=tensor([0.0324, 0.0323, 0.0319, 0.0337, 0.0416, 0.0337, 0.0296, 0.0319], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 14:29:05,081 INFO [train.py:903] (3/4) Epoch 9, batch 2350, loss[loss=0.3056, simple_loss=0.3555, pruned_loss=0.1278, over 19757.00 frames. ], tot_loss[loss=0.2462, simple_loss=0.3163, pruned_loss=0.08808, over 3832702.57 frames. ], batch size: 51, lr: 9.36e-03, grad_scale: 8.0 +2023-04-01 14:29:40,023 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57000.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 14:29:46,434 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 14:29:53,592 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57011.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:29:55,738 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57013.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:30:01,354 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 14:30:07,034 INFO [train.py:903] (3/4) Epoch 9, batch 2400, loss[loss=0.2313, simple_loss=0.2984, pruned_loss=0.08206, over 19608.00 frames. ], tot_loss[loss=0.249, simple_loss=0.3185, pruned_loss=0.0897, over 3828050.97 frames. ], batch size: 50, lr: 9.36e-03, grad_scale: 8.0 +2023-04-01 14:30:51,622 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.434e+02 5.803e+02 7.402e+02 8.988e+02 1.700e+03, threshold=1.480e+03, percent-clipped=2.0 +2023-04-01 14:31:11,536 INFO [train.py:903] (3/4) Epoch 9, batch 2450, loss[loss=0.2008, simple_loss=0.2743, pruned_loss=0.06363, over 19381.00 frames. ], tot_loss[loss=0.2478, simple_loss=0.3179, pruned_loss=0.0888, over 3833083.15 frames. ], batch size: 47, lr: 9.35e-03, grad_scale: 8.0 +2023-04-01 14:32:15,754 INFO [train.py:903] (3/4) Epoch 9, batch 2500, loss[loss=0.2391, simple_loss=0.3182, pruned_loss=0.07998, over 19671.00 frames. ], tot_loss[loss=0.2465, simple_loss=0.3172, pruned_loss=0.08787, over 3835971.93 frames. ], batch size: 59, lr: 9.35e-03, grad_scale: 8.0 +2023-04-01 14:33:00,912 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.384e+02 5.325e+02 6.954e+02 9.918e+02 1.981e+03, threshold=1.391e+03, percent-clipped=3.0 +2023-04-01 14:33:09,583 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57165.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:33:19,749 INFO [train.py:903] (3/4) Epoch 9, batch 2550, loss[loss=0.3383, simple_loss=0.3767, pruned_loss=0.15, over 13700.00 frames. ], tot_loss[loss=0.2472, simple_loss=0.3178, pruned_loss=0.08831, over 3825290.15 frames. ], batch size: 136, lr: 9.34e-03, grad_scale: 8.0 +2023-04-01 14:33:45,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7394, 3.1968, 3.2318, 3.2511, 1.1832, 3.1042, 2.6817, 2.9494], + device='cuda:3'), covar=tensor([0.1417, 0.0903, 0.0797, 0.0730, 0.4541, 0.0702, 0.0808, 0.1315], + device='cuda:3'), in_proj_covar=tensor([0.0625, 0.0544, 0.0742, 0.0618, 0.0687, 0.0488, 0.0468, 0.0689], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 14:33:59,358 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57204.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:34:13,772 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57215.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:34:15,581 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 14:34:22,488 INFO [train.py:903] (3/4) Epoch 9, batch 2600, loss[loss=0.2521, simple_loss=0.3264, pruned_loss=0.08892, over 19496.00 frames. ], tot_loss[loss=0.2458, simple_loss=0.3164, pruned_loss=0.08755, over 3833247.76 frames. ], batch size: 64, lr: 9.34e-03, grad_scale: 8.0 +2023-04-01 14:35:05,293 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57256.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 14:35:07,156 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.285e+02 5.816e+02 6.932e+02 7.774e+02 1.592e+03, threshold=1.386e+03, percent-clipped=3.0 +2023-04-01 14:35:07,957 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.37 vs. limit=5.0 +2023-04-01 14:35:12,155 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57262.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:35:22,436 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57270.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:35:25,691 INFO [train.py:903] (3/4) Epoch 9, batch 2650, loss[loss=0.2755, simple_loss=0.3531, pruned_loss=0.09889, over 19383.00 frames. ], tot_loss[loss=0.2459, simple_loss=0.3166, pruned_loss=0.08755, over 3835955.82 frames. ], batch size: 70, lr: 9.34e-03, grad_scale: 8.0 +2023-04-01 14:35:36,476 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57280.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:35:37,682 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57281.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 14:35:45,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 14:35:51,999 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5935, 4.1585, 2.5741, 3.6554, 0.9988, 3.8127, 3.9519, 3.9914], + device='cuda:3'), covar=tensor([0.0625, 0.1003, 0.2007, 0.0735, 0.3939, 0.0804, 0.0695, 0.1009], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0343, 0.0411, 0.0301, 0.0367, 0.0337, 0.0327, 0.0364], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 14:36:31,184 INFO [train.py:903] (3/4) Epoch 9, batch 2700, loss[loss=0.2741, simple_loss=0.3422, pruned_loss=0.103, over 19698.00 frames. ], tot_loss[loss=0.2467, simple_loss=0.3173, pruned_loss=0.08805, over 3823193.26 frames. ], batch size: 59, lr: 9.33e-03, grad_scale: 8.0 +2023-04-01 14:37:12,087 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57355.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:37:14,489 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57357.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:37:15,518 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.587e+02 5.592e+02 7.209e+02 8.892e+02 3.755e+03, threshold=1.442e+03, percent-clipped=7.0 +2023-04-01 14:37:33,717 INFO [train.py:903] (3/4) Epoch 9, batch 2750, loss[loss=0.189, simple_loss=0.267, pruned_loss=0.05553, over 19476.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.3169, pruned_loss=0.08794, over 3825885.85 frames. ], batch size: 49, lr: 9.33e-03, grad_scale: 8.0 +2023-04-01 14:37:49,327 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57385.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:38:30,635 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57416.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:38:38,519 INFO [train.py:903] (3/4) Epoch 9, batch 2800, loss[loss=0.2268, simple_loss=0.3135, pruned_loss=0.07003, over 19617.00 frames. ], tot_loss[loss=0.2479, simple_loss=0.3183, pruned_loss=0.08873, over 3811140.83 frames. ], batch size: 57, lr: 9.32e-03, grad_scale: 8.0 +2023-04-01 14:39:23,166 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.000e+02 5.525e+02 6.599e+02 8.446e+02 1.316e+03, threshold=1.320e+03, percent-clipped=0.0 +2023-04-01 14:39:38,408 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57470.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:39:41,763 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57472.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:39:42,549 INFO [train.py:903] (3/4) Epoch 9, batch 2850, loss[loss=0.2216, simple_loss=0.3007, pruned_loss=0.07123, over 19668.00 frames. ], tot_loss[loss=0.247, simple_loss=0.3177, pruned_loss=0.08811, over 3820588.88 frames. ], batch size: 53, lr: 9.32e-03, grad_scale: 8.0 +2023-04-01 14:39:51,277 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57480.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:39:55,840 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0623, 1.1603, 1.3386, 1.4365, 2.6206, 0.9912, 1.8661, 2.8203], + device='cuda:3'), covar=tensor([0.0483, 0.2632, 0.2704, 0.1566, 0.0798, 0.2376, 0.1191, 0.0390], + device='cuda:3'), in_proj_covar=tensor([0.0333, 0.0325, 0.0338, 0.0303, 0.0335, 0.0323, 0.0309, 0.0331], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 14:40:33,706 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57513.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:40:41,911 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 14:40:46,555 INFO [train.py:903] (3/4) Epoch 9, batch 2900, loss[loss=0.2543, simple_loss=0.3297, pruned_loss=0.08947, over 18751.00 frames. ], tot_loss[loss=0.2478, simple_loss=0.3181, pruned_loss=0.08876, over 3809908.39 frames. ], batch size: 74, lr: 9.32e-03, grad_scale: 8.0 +2023-04-01 14:41:03,516 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57536.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:41:16,958 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57548.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:41:30,782 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.787e+02 5.829e+02 7.475e+02 8.920e+02 2.516e+03, threshold=1.495e+03, percent-clipped=6.0 +2023-04-01 14:41:32,227 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57559.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:41:35,772 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57561.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:41:49,414 INFO [train.py:903] (3/4) Epoch 9, batch 2950, loss[loss=0.2018, simple_loss=0.2677, pruned_loss=0.06792, over 19391.00 frames. ], tot_loss[loss=0.2461, simple_loss=0.3164, pruned_loss=0.08788, over 3818155.53 frames. ], batch size: 47, lr: 9.31e-03, grad_scale: 8.0 +2023-04-01 14:42:32,126 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:42:53,537 INFO [train.py:903] (3/4) Epoch 9, batch 3000, loss[loss=0.2657, simple_loss=0.3421, pruned_loss=0.09467, over 19759.00 frames. ], tot_loss[loss=0.2454, simple_loss=0.3161, pruned_loss=0.08738, over 3825000.47 frames. ], batch size: 63, lr: 9.31e-03, grad_scale: 8.0 +2023-04-01 14:42:53,537 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 14:43:06,205 INFO [train.py:937] (3/4) Epoch 9, validation: loss=0.1831, simple_loss=0.2838, pruned_loss=0.04122, over 944034.00 frames. +2023-04-01 14:43:06,206 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 14:43:08,504 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 14:43:29,259 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57641.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:43:50,440 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 6.061e+02 7.914e+02 9.800e+02 2.087e+03, threshold=1.583e+03, percent-clipped=4.0 +2023-04-01 14:43:56,590 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57663.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:44:00,184 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57666.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:44:08,758 INFO [train.py:903] (3/4) Epoch 9, batch 3050, loss[loss=0.2974, simple_loss=0.3465, pruned_loss=0.1241, over 19662.00 frames. ], tot_loss[loss=0.2467, simple_loss=0.317, pruned_loss=0.0882, over 3821280.03 frames. ], batch size: 53, lr: 9.30e-03, grad_scale: 8.0 +2023-04-01 14:44:10,184 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57674.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:09,158 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57721.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:11,162 INFO [train.py:903] (3/4) Epoch 9, batch 3100, loss[loss=0.2592, simple_loss=0.3282, pruned_loss=0.09514, over 19284.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3156, pruned_loss=0.08778, over 3834048.23 frames. ], batch size: 66, lr: 9.30e-03, grad_scale: 8.0 +2023-04-01 14:45:16,055 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57726.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:19,266 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57728.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:26,180 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2022, 1.2803, 1.6729, 1.3643, 2.5493, 1.9532, 2.5542, 1.0082], + device='cuda:3'), covar=tensor([0.2161, 0.3693, 0.2097, 0.1682, 0.1188, 0.1878, 0.1303, 0.3428], + device='cuda:3'), in_proj_covar=tensor([0.0467, 0.0542, 0.0552, 0.0418, 0.0574, 0.0469, 0.0633, 0.0470], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 14:45:47,443 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57751.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:47,547 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57751.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:49,760 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57753.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:45:54,924 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.860e+02 5.906e+02 7.408e+02 1.029e+03 2.368e+03, threshold=1.482e+03, percent-clipped=3.0 +2023-04-01 14:45:57,418 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57760.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:46:14,147 INFO [train.py:903] (3/4) Epoch 9, batch 3150, loss[loss=0.2339, simple_loss=0.3092, pruned_loss=0.07934, over 19537.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3157, pruned_loss=0.08776, over 3831422.77 frames. ], batch size: 54, lr: 9.30e-03, grad_scale: 8.0 +2023-04-01 14:46:42,121 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 14:47:14,704 INFO [train.py:903] (3/4) Epoch 9, batch 3200, loss[loss=0.2434, simple_loss=0.3232, pruned_loss=0.08175, over 19658.00 frames. ], tot_loss[loss=0.2455, simple_loss=0.3155, pruned_loss=0.08777, over 3827674.56 frames. ], batch size: 58, lr: 9.29e-03, grad_scale: 8.0 +2023-04-01 14:47:16,010 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57824.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:47:55,932 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=57857.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:47:56,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.598e+02 5.791e+02 7.100e+02 9.261e+02 4.038e+03, threshold=1.420e+03, percent-clipped=7.0 +2023-04-01 14:48:09,922 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 14:48:14,917 INFO [train.py:903] (3/4) Epoch 9, batch 3250, loss[loss=0.2103, simple_loss=0.2815, pruned_loss=0.06956, over 19735.00 frames. ], tot_loss[loss=0.2452, simple_loss=0.3151, pruned_loss=0.08766, over 3828956.99 frames. ], batch size: 46, lr: 9.29e-03, grad_scale: 8.0 +2023-04-01 14:48:18,532 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57875.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:12,207 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57919.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:16,424 INFO [train.py:903] (3/4) Epoch 9, batch 3300, loss[loss=0.1721, simple_loss=0.2455, pruned_loss=0.04937, over 19741.00 frames. ], tot_loss[loss=0.2452, simple_loss=0.3153, pruned_loss=0.08757, over 3827198.42 frames. ], batch size: 46, lr: 9.28e-03, grad_scale: 8.0 +2023-04-01 14:49:23,984 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 14:49:26,382 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57930.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:36,424 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57939.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:43,142 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57944.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:56,596 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57955.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:49:59,470 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.480e+02 5.470e+02 6.783e+02 8.234e+02 1.579e+03, threshold=1.357e+03, percent-clipped=1.0 +2023-04-01 14:50:17,152 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57972.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:50:17,932 INFO [train.py:903] (3/4) Epoch 9, batch 3350, loss[loss=0.2347, simple_loss=0.3042, pruned_loss=0.08263, over 19622.00 frames. ], tot_loss[loss=0.2445, simple_loss=0.3146, pruned_loss=0.08721, over 3838824.11 frames. ], batch size: 50, lr: 9.28e-03, grad_scale: 8.0 +2023-04-01 14:50:22,849 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57977.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:50:54,297 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58002.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:51:20,066 INFO [train.py:903] (3/4) Epoch 9, batch 3400, loss[loss=0.2928, simple_loss=0.3515, pruned_loss=0.117, over 13437.00 frames. ], tot_loss[loss=0.245, simple_loss=0.315, pruned_loss=0.0875, over 3817777.19 frames. ], batch size: 136, lr: 9.28e-03, grad_scale: 8.0 +2023-04-01 14:51:35,099 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-01 14:52:02,956 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.171e+02 5.309e+02 6.592e+02 8.930e+02 1.711e+03, threshold=1.318e+03, percent-clipped=4.0 +2023-04-01 14:52:20,120 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7521, 1.3961, 1.3885, 2.3078, 1.8364, 2.1466, 2.2022, 1.8355], + device='cuda:3'), covar=tensor([0.0825, 0.1049, 0.1109, 0.0806, 0.0877, 0.0705, 0.0889, 0.0683], + device='cuda:3'), in_proj_covar=tensor([0.0217, 0.0232, 0.0229, 0.0256, 0.0243, 0.0216, 0.0204, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 14:52:20,806 INFO [train.py:903] (3/4) Epoch 9, batch 3450, loss[loss=0.2451, simple_loss=0.3002, pruned_loss=0.095, over 19738.00 frames. ], tot_loss[loss=0.2451, simple_loss=0.3149, pruned_loss=0.08762, over 3815596.13 frames. ], batch size: 46, lr: 9.27e-03, grad_scale: 8.0 +2023-04-01 14:52:25,128 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 14:52:48,554 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=58095.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:53:22,372 INFO [train.py:903] (3/4) Epoch 9, batch 3500, loss[loss=0.2765, simple_loss=0.3404, pruned_loss=0.1063, over 19671.00 frames. ], tot_loss[loss=0.2445, simple_loss=0.3149, pruned_loss=0.08707, over 3830729.28 frames. ], batch size: 55, lr: 9.27e-03, grad_scale: 8.0 +2023-04-01 14:53:25,163 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6944, 4.1484, 4.3699, 4.3559, 1.4133, 4.0042, 3.5749, 4.0116], + device='cuda:3'), covar=tensor([0.1225, 0.0787, 0.0560, 0.0525, 0.5327, 0.0622, 0.0595, 0.1043], + device='cuda:3'), in_proj_covar=tensor([0.0624, 0.0550, 0.0750, 0.0620, 0.0686, 0.0491, 0.0467, 0.0685], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 14:53:33,947 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58131.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:53:36,482 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 14:54:03,483 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58156.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:54:05,471 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.510e+02 6.015e+02 7.407e+02 9.414e+02 2.837e+03, threshold=1.481e+03, percent-clipped=3.0 +2023-04-01 14:54:06,988 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58159.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:54:08,258 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3277, 1.3787, 1.7473, 1.5359, 2.5329, 2.0744, 2.5793, 1.0010], + device='cuda:3'), covar=tensor([0.1993, 0.3333, 0.1972, 0.1562, 0.1279, 0.1757, 0.1423, 0.3304], + device='cuda:3'), in_proj_covar=tensor([0.0468, 0.0548, 0.0558, 0.0422, 0.0580, 0.0473, 0.0638, 0.0474], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 14:54:10,345 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58162.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:54:24,270 INFO [train.py:903] (3/4) Epoch 9, batch 3550, loss[loss=0.2096, simple_loss=0.2901, pruned_loss=0.06457, over 19748.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.315, pruned_loss=0.08738, over 3822340.09 frames. ], batch size: 51, lr: 9.26e-03, grad_scale: 8.0 +2023-04-01 14:54:37,431 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1616, 1.2810, 1.1114, 0.9437, 0.9915, 1.0041, 0.0665, 0.2713], + device='cuda:3'), covar=tensor([0.0564, 0.0562, 0.0349, 0.0426, 0.1062, 0.0467, 0.0898, 0.0945], + device='cuda:3'), in_proj_covar=tensor([0.0318, 0.0322, 0.0319, 0.0337, 0.0409, 0.0334, 0.0299, 0.0316], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 14:54:50,516 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58195.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:55:09,229 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58210.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:55:20,399 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58220.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:55:24,265 INFO [train.py:903] (3/4) Epoch 9, batch 3600, loss[loss=0.2571, simple_loss=0.3337, pruned_loss=0.09022, over 18433.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.3149, pruned_loss=0.08742, over 3811541.97 frames. ], batch size: 83, lr: 9.26e-03, grad_scale: 8.0 +2023-04-01 14:55:30,394 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58228.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:56:01,032 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58253.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 14:56:06,162 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.641e+02 5.487e+02 6.500e+02 7.997e+02 2.924e+03, threshold=1.300e+03, percent-clipped=2.0 +2023-04-01 14:56:23,769 INFO [train.py:903] (3/4) Epoch 9, batch 3650, loss[loss=0.2661, simple_loss=0.3384, pruned_loss=0.09692, over 19617.00 frames. ], tot_loss[loss=0.2445, simple_loss=0.3145, pruned_loss=0.08722, over 3823021.60 frames. ], batch size: 57, lr: 9.26e-03, grad_scale: 16.0 +2023-04-01 14:57:24,455 INFO [train.py:903] (3/4) Epoch 9, batch 3700, loss[loss=0.1981, simple_loss=0.2699, pruned_loss=0.06321, over 19744.00 frames. ], tot_loss[loss=0.2439, simple_loss=0.3142, pruned_loss=0.08683, over 3836945.50 frames. ], batch size: 47, lr: 9.25e-03, grad_scale: 8.0 +2023-04-01 14:57:59,836 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-04-01 14:58:07,747 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.190e+02 6.060e+02 7.736e+02 9.843e+02 2.060e+03, threshold=1.547e+03, percent-clipped=9.0 +2023-04-01 14:58:23,949 INFO [train.py:903] (3/4) Epoch 9, batch 3750, loss[loss=0.2798, simple_loss=0.3456, pruned_loss=0.107, over 19636.00 frames. ], tot_loss[loss=0.2446, simple_loss=0.3148, pruned_loss=0.08721, over 3836347.02 frames. ], batch size: 61, lr: 9.25e-03, grad_scale: 8.0 +2023-04-01 14:59:24,634 INFO [train.py:903] (3/4) Epoch 9, batch 3800, loss[loss=0.2289, simple_loss=0.3147, pruned_loss=0.07161, over 19331.00 frames. ], tot_loss[loss=0.244, simple_loss=0.3143, pruned_loss=0.08686, over 3826500.67 frames. ], batch size: 66, lr: 9.24e-03, grad_scale: 8.0 +2023-04-01 14:59:54,363 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 15:00:08,585 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.327e+02 5.230e+02 6.601e+02 8.580e+02 1.875e+03, threshold=1.320e+03, percent-clipped=3.0 +2023-04-01 15:00:11,112 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4257, 1.4102, 1.7816, 1.5665, 2.5911, 2.1689, 2.6847, 1.2873], + device='cuda:3'), covar=tensor([0.1982, 0.3515, 0.2044, 0.1605, 0.1207, 0.1733, 0.1228, 0.3235], + device='cuda:3'), in_proj_covar=tensor([0.0475, 0.0554, 0.0568, 0.0426, 0.0584, 0.0480, 0.0643, 0.0477], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 15:00:17,578 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58466.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:00:24,957 INFO [train.py:903] (3/4) Epoch 9, batch 3850, loss[loss=0.2326, simple_loss=0.3007, pruned_loss=0.08226, over 16882.00 frames. ], tot_loss[loss=0.244, simple_loss=0.3143, pruned_loss=0.08683, over 3826574.89 frames. ], batch size: 37, lr: 9.24e-03, grad_scale: 8.0 +2023-04-01 15:00:46,698 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58491.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:01:02,734 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=58503.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:01:06,100 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=58506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:01:17,420 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2925, 1.3596, 1.8739, 1.3018, 2.6539, 3.4570, 3.2485, 3.6402], + device='cuda:3'), covar=tensor([0.1383, 0.3047, 0.2644, 0.1925, 0.0509, 0.0172, 0.0200, 0.0182], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0286, 0.0317, 0.0247, 0.0209, 0.0144, 0.0204, 0.0182], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:01:25,307 INFO [train.py:903] (3/4) Epoch 9, batch 3900, loss[loss=0.2374, simple_loss=0.3075, pruned_loss=0.08362, over 19601.00 frames. ], tot_loss[loss=0.2438, simple_loss=0.3141, pruned_loss=0.08677, over 3814166.01 frames. ], batch size: 50, lr: 9.24e-03, grad_scale: 8.0 +2023-04-01 15:02:08,184 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.315e+02 6.109e+02 7.308e+02 8.933e+02 2.200e+03, threshold=1.462e+03, percent-clipped=5.0 +2023-04-01 15:02:26,007 INFO [train.py:903] (3/4) Epoch 9, batch 3950, loss[loss=0.2427, simple_loss=0.3163, pruned_loss=0.0846, over 19673.00 frames. ], tot_loss[loss=0.2437, simple_loss=0.3136, pruned_loss=0.0869, over 3809638.13 frames. ], batch size: 59, lr: 9.23e-03, grad_scale: 8.0 +2023-04-01 15:02:28,075 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 15:02:51,967 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8929, 1.5886, 1.4999, 1.9629, 1.7375, 1.5939, 1.6439, 1.7962], + device='cuda:3'), covar=tensor([0.0864, 0.1362, 0.1383, 0.0820, 0.1077, 0.0505, 0.1002, 0.0604], + device='cuda:3'), in_proj_covar=tensor([0.0248, 0.0351, 0.0290, 0.0239, 0.0298, 0.0242, 0.0270, 0.0232], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:03:21,986 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58618.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:03:25,477 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58621.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:03:28,120 INFO [train.py:903] (3/4) Epoch 9, batch 4000, loss[loss=0.2721, simple_loss=0.3343, pruned_loss=0.1049, over 19483.00 frames. ], tot_loss[loss=0.2447, simple_loss=0.3149, pruned_loss=0.08728, over 3811593.61 frames. ], batch size: 64, lr: 9.23e-03, grad_scale: 8.0 +2023-04-01 15:04:10,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.392e+02 5.642e+02 6.845e+02 8.578e+02 2.087e+03, threshold=1.369e+03, percent-clipped=3.0 +2023-04-01 15:04:10,786 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 15:04:27,015 INFO [train.py:903] (3/4) Epoch 9, batch 4050, loss[loss=0.2273, simple_loss=0.2987, pruned_loss=0.07793, over 19533.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3159, pruned_loss=0.08767, over 3822839.51 frames. ], batch size: 56, lr: 9.22e-03, grad_scale: 8.0 +2023-04-01 15:04:29,190 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-01 15:04:59,250 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58699.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:05:28,159 INFO [train.py:903] (3/4) Epoch 9, batch 4100, loss[loss=0.2722, simple_loss=0.3462, pruned_loss=0.09911, over 19769.00 frames. ], tot_loss[loss=0.245, simple_loss=0.3156, pruned_loss=0.08723, over 3815945.64 frames. ], batch size: 56, lr: 9.22e-03, grad_scale: 8.0 +2023-04-01 15:06:03,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 15:06:11,181 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.329e+02 5.572e+02 6.953e+02 8.904e+02 1.888e+03, threshold=1.391e+03, percent-clipped=3.0 +2023-04-01 15:06:28,248 INFO [train.py:903] (3/4) Epoch 9, batch 4150, loss[loss=0.1853, simple_loss=0.2611, pruned_loss=0.05478, over 15282.00 frames. ], tot_loss[loss=0.2437, simple_loss=0.3144, pruned_loss=0.08648, over 3817153.73 frames. ], batch size: 33, lr: 9.22e-03, grad_scale: 8.0 +2023-04-01 15:06:46,514 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-04-01 15:06:54,411 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58793.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:07:30,859 INFO [train.py:903] (3/4) Epoch 9, batch 4200, loss[loss=0.2382, simple_loss=0.3125, pruned_loss=0.08196, over 19622.00 frames. ], tot_loss[loss=0.2439, simple_loss=0.3143, pruned_loss=0.08672, over 3804099.43 frames. ], batch size: 57, lr: 9.21e-03, grad_scale: 8.0 +2023-04-01 15:07:34,199 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 15:07:49,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4563, 2.2945, 1.9435, 1.8677, 1.8006, 1.9034, 0.5633, 1.2340], + device='cuda:3'), covar=tensor([0.0337, 0.0378, 0.0306, 0.0438, 0.0670, 0.0546, 0.0745, 0.0632], + device='cuda:3'), in_proj_covar=tensor([0.0321, 0.0322, 0.0321, 0.0336, 0.0412, 0.0336, 0.0299, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:08:14,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.798e+02 5.991e+02 7.229e+02 9.145e+02 1.621e+03, threshold=1.446e+03, percent-clipped=3.0 +2023-04-01 15:08:32,532 INFO [train.py:903] (3/4) Epoch 9, batch 4250, loss[loss=0.242, simple_loss=0.3178, pruned_loss=0.08308, over 19665.00 frames. ], tot_loss[loss=0.2429, simple_loss=0.3134, pruned_loss=0.08624, over 3803987.89 frames. ], batch size: 55, lr: 9.21e-03, grad_scale: 8.0 +2023-04-01 15:08:34,189 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58874.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:08:37,530 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58877.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:08:48,664 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 15:08:59,999 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 15:09:03,851 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58899.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:09:08,272 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58902.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:09:32,781 INFO [train.py:903] (3/4) Epoch 9, batch 4300, loss[loss=0.223, simple_loss=0.2938, pruned_loss=0.07608, over 19412.00 frames. ], tot_loss[loss=0.2443, simple_loss=0.3146, pruned_loss=0.08702, over 3814048.19 frames. ], batch size: 48, lr: 9.20e-03, grad_scale: 8.0 +2023-04-01 15:09:34,153 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58924.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:10:17,293 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.360e+02 5.584e+02 7.321e+02 9.114e+02 2.155e+03, threshold=1.464e+03, percent-clipped=5.0 +2023-04-01 15:10:17,710 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0432, 2.0305, 1.7602, 1.6077, 1.4592, 1.7048, 0.5010, 1.1283], + device='cuda:3'), covar=tensor([0.0313, 0.0368, 0.0254, 0.0387, 0.0781, 0.0432, 0.0718, 0.0605], + device='cuda:3'), in_proj_covar=tensor([0.0321, 0.0322, 0.0321, 0.0335, 0.0414, 0.0338, 0.0299, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:10:26,449 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 15:10:33,821 INFO [train.py:903] (3/4) Epoch 9, batch 4350, loss[loss=0.2778, simple_loss=0.3448, pruned_loss=0.1054, over 19617.00 frames. ], tot_loss[loss=0.246, simple_loss=0.3165, pruned_loss=0.08778, over 3818935.40 frames. ], batch size: 61, lr: 9.20e-03, grad_scale: 8.0 +2023-04-01 15:10:43,767 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-01 15:11:10,136 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7410, 4.2563, 2.6335, 3.8395, 1.1337, 4.0427, 3.9849, 4.1673], + device='cuda:3'), covar=tensor([0.0613, 0.1032, 0.2124, 0.0757, 0.3943, 0.0783, 0.0777, 0.1044], + device='cuda:3'), in_proj_covar=tensor([0.0410, 0.0350, 0.0415, 0.0311, 0.0369, 0.0344, 0.0335, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 15:11:34,342 INFO [train.py:903] (3/4) Epoch 9, batch 4400, loss[loss=0.2321, simple_loss=0.3088, pruned_loss=0.07769, over 19857.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.3166, pruned_loss=0.08816, over 3803172.94 frames. ], batch size: 52, lr: 9.20e-03, grad_scale: 8.0 +2023-04-01 15:11:58,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 15:11:58,800 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=59043.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:12:08,636 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 15:12:18,356 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.517e+02 5.626e+02 6.682e+02 1.014e+03 2.125e+03, threshold=1.336e+03, percent-clipped=6.0 +2023-04-01 15:12:35,838 INFO [train.py:903] (3/4) Epoch 9, batch 4450, loss[loss=0.2443, simple_loss=0.3102, pruned_loss=0.08918, over 19849.00 frames. ], tot_loss[loss=0.248, simple_loss=0.3182, pruned_loss=0.08893, over 3804498.22 frames. ], batch size: 52, lr: 9.19e-03, grad_scale: 8.0 +2023-04-01 15:13:05,749 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 15:13:36,621 INFO [train.py:903] (3/4) Epoch 9, batch 4500, loss[loss=0.2377, simple_loss=0.3154, pruned_loss=0.08002, over 19650.00 frames. ], tot_loss[loss=0.247, simple_loss=0.3175, pruned_loss=0.08829, over 3804715.96 frames. ], batch size: 58, lr: 9.19e-03, grad_scale: 8.0 +2023-04-01 15:13:54,300 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=59137.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:14:17,431 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-01 15:14:20,650 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59158.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:14:21,456 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.665e+02 5.638e+02 7.209e+02 9.049e+02 2.430e+03, threshold=1.442e+03, percent-clipped=5.0 +2023-04-01 15:14:38,161 INFO [train.py:903] (3/4) Epoch 9, batch 4550, loss[loss=0.2587, simple_loss=0.3376, pruned_loss=0.08996, over 19490.00 frames. ], tot_loss[loss=0.2465, simple_loss=0.3173, pruned_loss=0.08783, over 3804188.39 frames. ], batch size: 64, lr: 9.19e-03, grad_scale: 8.0 +2023-04-01 15:14:46,062 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 15:14:49,767 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59182.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:15:11,262 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 15:15:39,256 INFO [train.py:903] (3/4) Epoch 9, batch 4600, loss[loss=0.2132, simple_loss=0.2895, pruned_loss=0.06843, over 19487.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.316, pruned_loss=0.08692, over 3802651.93 frames. ], batch size: 49, lr: 9.18e-03, grad_scale: 8.0 +2023-04-01 15:16:01,384 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 15:16:15,387 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59252.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:16:24,167 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.716e+02 5.555e+02 6.855e+02 8.371e+02 1.742e+03, threshold=1.371e+03, percent-clipped=2.0 +2023-04-01 15:16:33,268 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8917, 1.1621, 1.4540, 0.5954, 2.1205, 2.4466, 2.1163, 2.5837], + device='cuda:3'), covar=tensor([0.1417, 0.3257, 0.2918, 0.2238, 0.0472, 0.0218, 0.0333, 0.0262], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0290, 0.0321, 0.0248, 0.0214, 0.0147, 0.0206, 0.0185], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:16:35,112 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=59268.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:16:40,777 INFO [train.py:903] (3/4) Epoch 9, batch 4650, loss[loss=0.2666, simple_loss=0.3305, pruned_loss=0.1014, over 19608.00 frames. ], tot_loss[loss=0.245, simple_loss=0.3159, pruned_loss=0.08709, over 3812462.96 frames. ], batch size: 57, lr: 9.18e-03, grad_scale: 8.0 +2023-04-01 15:16:56,380 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 15:17:09,148 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 15:17:41,932 INFO [train.py:903] (3/4) Epoch 9, batch 4700, loss[loss=0.276, simple_loss=0.3437, pruned_loss=0.1042, over 19603.00 frames. ], tot_loss[loss=0.2454, simple_loss=0.3164, pruned_loss=0.08725, over 3817754.96 frames. ], batch size: 61, lr: 9.17e-03, grad_scale: 8.0 +2023-04-01 15:18:03,982 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 15:18:13,129 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7316, 1.2262, 1.3014, 2.1931, 1.7502, 1.9230, 2.0715, 1.6523], + device='cuda:3'), covar=tensor([0.0891, 0.1253, 0.1208, 0.0911, 0.0893, 0.0878, 0.0867, 0.0812], + device='cuda:3'), in_proj_covar=tensor([0.0216, 0.0230, 0.0228, 0.0257, 0.0240, 0.0213, 0.0202, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 15:18:25,865 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.928e+02 5.586e+02 6.716e+02 8.168e+02 1.499e+03, threshold=1.343e+03, percent-clipped=1.0 +2023-04-01 15:18:32,960 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3104, 1.1776, 1.4113, 1.5306, 2.8367, 0.9894, 2.1380, 3.0508], + device='cuda:3'), covar=tensor([0.0486, 0.2711, 0.2602, 0.1627, 0.0820, 0.2443, 0.1155, 0.0370], + device='cuda:3'), in_proj_covar=tensor([0.0338, 0.0329, 0.0338, 0.0311, 0.0343, 0.0326, 0.0315, 0.0338], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:18:41,884 INFO [train.py:903] (3/4) Epoch 9, batch 4750, loss[loss=0.2509, simple_loss=0.325, pruned_loss=0.08841, over 19520.00 frames. ], tot_loss[loss=0.2455, simple_loss=0.3167, pruned_loss=0.08715, over 3820414.40 frames. ], batch size: 56, lr: 9.17e-03, grad_scale: 8.0 +2023-04-01 15:18:54,772 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59383.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:19:32,478 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59414.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:19:43,257 INFO [train.py:903] (3/4) Epoch 9, batch 4800, loss[loss=0.2531, simple_loss=0.3287, pruned_loss=0.08879, over 19446.00 frames. ], tot_loss[loss=0.2453, simple_loss=0.3161, pruned_loss=0.08726, over 3819989.88 frames. ], batch size: 64, lr: 9.17e-03, grad_scale: 8.0 +2023-04-01 15:19:58,761 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0499, 3.4554, 1.9446, 2.1312, 2.9903, 1.5711, 1.3956, 1.9472], + device='cuda:3'), covar=tensor([0.1050, 0.0367, 0.0917, 0.0593, 0.0444, 0.1001, 0.0818, 0.0649], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0293, 0.0319, 0.0242, 0.0230, 0.0315, 0.0287, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:20:03,388 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59439.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:20:27,561 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.770e+02 5.762e+02 6.893e+02 8.818e+02 1.836e+03, threshold=1.379e+03, percent-clipped=7.0 +2023-04-01 15:20:43,944 INFO [train.py:903] (3/4) Epoch 9, batch 4850, loss[loss=0.2208, simple_loss=0.303, pruned_loss=0.0693, over 19537.00 frames. ], tot_loss[loss=0.246, simple_loss=0.3167, pruned_loss=0.08768, over 3836946.98 frames. ], batch size: 54, lr: 9.16e-03, grad_scale: 8.0 +2023-04-01 15:20:45,436 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5955, 1.3475, 1.2824, 2.0714, 1.6346, 1.9034, 1.9221, 1.6464], + device='cuda:3'), covar=tensor([0.0830, 0.0979, 0.1110, 0.0773, 0.0843, 0.0661, 0.0837, 0.0659], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0229, 0.0225, 0.0254, 0.0237, 0.0212, 0.0199, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 15:21:08,912 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 15:21:27,104 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59508.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:21:28,999 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 15:21:35,283 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 15:21:36,464 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 15:21:42,268 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59520.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:21:45,374 INFO [train.py:903] (3/4) Epoch 9, batch 4900, loss[loss=0.2675, simple_loss=0.3374, pruned_loss=0.09879, over 19771.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3164, pruned_loss=0.0874, over 3827576.33 frames. ], batch size: 54, lr: 9.16e-03, grad_scale: 8.0 +2023-04-01 15:21:46,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 15:21:49,055 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=59526.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:21:56,449 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-04-01 15:21:56,973 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59533.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:22:05,288 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 15:22:29,456 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.724e+02 5.274e+02 6.528e+02 8.023e+02 2.606e+03, threshold=1.306e+03, percent-clipped=3.0 +2023-04-01 15:22:45,478 INFO [train.py:903] (3/4) Epoch 9, batch 4950, loss[loss=0.3285, simple_loss=0.3716, pruned_loss=0.1427, over 13668.00 frames. ], tot_loss[loss=0.2466, simple_loss=0.3171, pruned_loss=0.08801, over 3821125.19 frames. ], batch size: 136, lr: 9.15e-03, grad_scale: 8.0 +2023-04-01 15:23:01,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 15:23:24,576 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 15:23:44,261 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9491, 1.3699, 1.0524, 0.9699, 1.1730, 0.9608, 0.9099, 1.2243], + device='cuda:3'), covar=tensor([0.0456, 0.0694, 0.1018, 0.0563, 0.0444, 0.1063, 0.0567, 0.0400], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0291, 0.0318, 0.0241, 0.0230, 0.0314, 0.0286, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:23:46,152 INFO [train.py:903] (3/4) Epoch 9, batch 5000, loss[loss=0.2091, simple_loss=0.2828, pruned_loss=0.06764, over 19685.00 frames. ], tot_loss[loss=0.2457, simple_loss=0.3161, pruned_loss=0.08766, over 3818483.06 frames. ], batch size: 53, lr: 9.15e-03, grad_scale: 4.0 +2023-04-01 15:23:53,579 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 15:24:04,769 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 15:24:06,363 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59639.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:24:08,533 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59641.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:24:30,504 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.264e+02 5.818e+02 6.890e+02 9.185e+02 2.943e+03, threshold=1.378e+03, percent-clipped=3.0 +2023-04-01 15:24:35,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59664.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:24:46,096 INFO [train.py:903] (3/4) Epoch 9, batch 5050, loss[loss=0.2074, simple_loss=0.2736, pruned_loss=0.07057, over 18182.00 frames. ], tot_loss[loss=0.2445, simple_loss=0.3151, pruned_loss=0.08696, over 3813656.98 frames. ], batch size: 40, lr: 9.15e-03, grad_scale: 4.0 +2023-04-01 15:25:21,704 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 15:25:47,328 INFO [train.py:903] (3/4) Epoch 9, batch 5100, loss[loss=0.2622, simple_loss=0.3284, pruned_loss=0.09803, over 19596.00 frames. ], tot_loss[loss=0.2461, simple_loss=0.3165, pruned_loss=0.08788, over 3804835.66 frames. ], batch size: 61, lr: 9.14e-03, grad_scale: 4.0 +2023-04-01 15:25:56,484 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 15:25:59,757 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 15:26:01,393 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9881, 1.4771, 1.1861, 0.9639, 1.3226, 0.9521, 0.9317, 1.3413], + device='cuda:3'), covar=tensor([0.0485, 0.0436, 0.0653, 0.0478, 0.0278, 0.0764, 0.0399, 0.0267], + device='cuda:3'), in_proj_covar=tensor([0.0289, 0.0294, 0.0321, 0.0242, 0.0231, 0.0317, 0.0288, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:26:05,153 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 15:26:33,072 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.640e+02 5.962e+02 7.316e+02 9.170e+02 1.645e+03, threshold=1.463e+03, percent-clipped=5.0 +2023-04-01 15:26:47,506 INFO [train.py:903] (3/4) Epoch 9, batch 5150, loss[loss=0.2864, simple_loss=0.3463, pruned_loss=0.1133, over 19589.00 frames. ], tot_loss[loss=0.2458, simple_loss=0.3161, pruned_loss=0.08775, over 3805336.80 frames. ], batch size: 61, lr: 9.14e-03, grad_scale: 4.0 +2023-04-01 15:26:58,279 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 15:27:32,234 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 15:27:49,839 INFO [train.py:903] (3/4) Epoch 9, batch 5200, loss[loss=0.2262, simple_loss=0.2952, pruned_loss=0.07861, over 17317.00 frames. ], tot_loss[loss=0.2457, simple_loss=0.3159, pruned_loss=0.08781, over 3800040.59 frames. ], batch size: 38, lr: 9.14e-03, grad_scale: 8.0 +2023-04-01 15:27:56,720 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9061, 2.0636, 2.0699, 2.0035, 4.4011, 0.9586, 2.3962, 4.5167], + device='cuda:3'), covar=tensor([0.0327, 0.2183, 0.2164, 0.1490, 0.0640, 0.2607, 0.1277, 0.0217], + device='cuda:3'), in_proj_covar=tensor([0.0338, 0.0328, 0.0338, 0.0310, 0.0338, 0.0324, 0.0316, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:27:59,732 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 15:28:34,608 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.759e+02 5.545e+02 6.726e+02 8.723e+02 1.623e+03, threshold=1.345e+03, percent-clipped=2.0 +2023-04-01 15:28:39,342 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=59864.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:28:41,613 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 15:28:41,851 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59866.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:28:50,360 INFO [train.py:903] (3/4) Epoch 9, batch 5250, loss[loss=0.2607, simple_loss=0.3308, pruned_loss=0.09528, over 18367.00 frames. ], tot_loss[loss=0.2452, simple_loss=0.3153, pruned_loss=0.08759, over 3794856.99 frames. ], batch size: 84, lr: 9.13e-03, grad_scale: 8.0 +2023-04-01 15:29:19,332 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59897.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:29:31,306 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59907.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:29:39,003 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8338, 1.7784, 1.8729, 1.6778, 4.2710, 0.8399, 2.4620, 4.4800], + device='cuda:3'), covar=tensor([0.0309, 0.2248, 0.2379, 0.1625, 0.0669, 0.2617, 0.1198, 0.0232], + device='cuda:3'), in_proj_covar=tensor([0.0335, 0.0327, 0.0338, 0.0310, 0.0336, 0.0323, 0.0314, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:29:50,053 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59922.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:29:50,796 INFO [train.py:903] (3/4) Epoch 9, batch 5300, loss[loss=0.2693, simple_loss=0.3407, pruned_loss=0.09894, over 18275.00 frames. ], tot_loss[loss=0.2443, simple_loss=0.3147, pruned_loss=0.087, over 3798652.13 frames. ], batch size: 83, lr: 9.13e-03, grad_scale: 8.0 +2023-04-01 15:30:04,495 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 15:30:09,128 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7549, 4.2994, 2.7685, 3.7853, 1.1207, 4.0308, 4.0616, 4.2228], + device='cuda:3'), covar=tensor([0.0603, 0.1209, 0.1831, 0.0727, 0.3850, 0.0667, 0.0704, 0.0999], + device='cuda:3'), in_proj_covar=tensor([0.0410, 0.0347, 0.0412, 0.0307, 0.0369, 0.0341, 0.0332, 0.0371], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 15:30:36,750 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.396e+02 5.525e+02 7.204e+02 8.995e+02 2.228e+03, threshold=1.441e+03, percent-clipped=7.0 +2023-04-01 15:30:39,259 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7227, 2.2083, 2.3784, 2.8079, 2.3515, 2.3764, 2.2905, 2.7876], + device='cuda:3'), covar=tensor([0.0747, 0.1611, 0.1114, 0.0853, 0.1185, 0.0414, 0.0944, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0349, 0.0289, 0.0237, 0.0295, 0.0240, 0.0275, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:30:50,971 INFO [train.py:903] (3/4) Epoch 9, batch 5350, loss[loss=0.212, simple_loss=0.2782, pruned_loss=0.07296, over 19113.00 frames. ], tot_loss[loss=0.2438, simple_loss=0.3141, pruned_loss=0.08669, over 3803958.83 frames. ], batch size: 42, lr: 9.12e-03, grad_scale: 8.0 +2023-04-01 15:30:59,192 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59979.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:31:03,381 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0345, 1.1358, 1.4103, 1.5462, 2.6232, 1.0775, 1.9889, 2.7655], + device='cuda:3'), covar=tensor([0.0534, 0.2699, 0.2583, 0.1498, 0.0771, 0.2184, 0.1115, 0.0415], + device='cuda:3'), in_proj_covar=tensor([0.0334, 0.0326, 0.0336, 0.0309, 0.0334, 0.0321, 0.0315, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:31:24,014 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 15:31:52,157 INFO [train.py:903] (3/4) Epoch 9, batch 5400, loss[loss=0.315, simple_loss=0.363, pruned_loss=0.1335, over 13428.00 frames. ], tot_loss[loss=0.2447, simple_loss=0.3147, pruned_loss=0.08732, over 3800219.07 frames. ], batch size: 136, lr: 9.12e-03, grad_scale: 8.0 +2023-04-01 15:32:35,001 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60058.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:32:36,997 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.978e+02 5.587e+02 7.112e+02 9.365e+02 1.948e+03, threshold=1.422e+03, percent-clipped=3.0 +2023-04-01 15:32:50,005 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60070.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:32:53,196 INFO [train.py:903] (3/4) Epoch 9, batch 5450, loss[loss=0.2335, simple_loss=0.3094, pruned_loss=0.07883, over 19580.00 frames. ], tot_loss[loss=0.2444, simple_loss=0.3145, pruned_loss=0.08711, over 3795595.98 frames. ], batch size: 52, lr: 9.12e-03, grad_scale: 8.0 +2023-04-01 15:33:06,639 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60083.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:33:21,789 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60096.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:33:34,846 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6594, 1.4574, 1.3726, 2.0457, 1.6969, 2.1077, 2.0927, 1.8558], + device='cuda:3'), covar=tensor([0.0775, 0.0953, 0.1089, 0.0887, 0.0887, 0.0646, 0.0760, 0.0622], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0229, 0.0228, 0.0253, 0.0241, 0.0213, 0.0199, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 15:33:55,434 INFO [train.py:903] (3/4) Epoch 9, batch 5500, loss[loss=0.2013, simple_loss=0.2831, pruned_loss=0.05973, over 19730.00 frames. ], tot_loss[loss=0.2458, simple_loss=0.3163, pruned_loss=0.08772, over 3788786.66 frames. ], batch size: 46, lr: 9.11e-03, grad_scale: 8.0 +2023-04-01 15:34:17,037 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 15:34:40,128 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.116e+02 5.485e+02 6.858e+02 8.862e+02 1.983e+03, threshold=1.372e+03, percent-clipped=4.0 +2023-04-01 15:34:46,364 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-01 15:34:56,042 INFO [train.py:903] (3/4) Epoch 9, batch 5550, loss[loss=0.2211, simple_loss=0.3002, pruned_loss=0.07101, over 19851.00 frames. ], tot_loss[loss=0.247, simple_loss=0.3171, pruned_loss=0.08847, over 3791051.31 frames. ], batch size: 52, lr: 9.11e-03, grad_scale: 8.0 +2023-04-01 15:35:01,728 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 15:35:28,633 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8690, 1.8431, 1.8456, 1.8221, 4.3609, 0.9092, 2.3236, 4.5296], + device='cuda:3'), covar=tensor([0.0360, 0.2296, 0.2400, 0.1586, 0.0651, 0.2638, 0.1331, 0.0228], + device='cuda:3'), in_proj_covar=tensor([0.0336, 0.0327, 0.0336, 0.0309, 0.0336, 0.0322, 0.0314, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:35:42,135 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60210.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:35:48,914 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60216.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:35:49,879 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 15:35:57,518 INFO [train.py:903] (3/4) Epoch 9, batch 5600, loss[loss=0.2086, simple_loss=0.2825, pruned_loss=0.06737, over 14633.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3157, pruned_loss=0.08775, over 3797079.40 frames. ], batch size: 32, lr: 9.11e-03, grad_scale: 8.0 +2023-04-01 15:36:13,020 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60235.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:36:28,798 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-01 15:36:31,694 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60251.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:36:41,625 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.287e+02 5.903e+02 7.092e+02 9.595e+02 1.671e+03, threshold=1.418e+03, percent-clipped=4.0 +2023-04-01 15:36:42,028 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60260.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:36:58,108 INFO [train.py:903] (3/4) Epoch 9, batch 5650, loss[loss=0.2622, simple_loss=0.3332, pruned_loss=0.09558, over 19085.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.3166, pruned_loss=0.08814, over 3800930.78 frames. ], batch size: 69, lr: 9.10e-03, grad_scale: 8.0 +2023-04-01 15:37:45,065 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 15:38:00,339 INFO [train.py:903] (3/4) Epoch 9, batch 5700, loss[loss=0.2393, simple_loss=0.316, pruned_loss=0.08135, over 19669.00 frames. ], tot_loss[loss=0.2465, simple_loss=0.3167, pruned_loss=0.08813, over 3804760.16 frames. ], batch size: 60, lr: 9.10e-03, grad_scale: 8.0 +2023-04-01 15:38:02,961 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60325.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:38:45,030 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.384e+02 5.631e+02 6.670e+02 7.834e+02 2.342e+03, threshold=1.334e+03, percent-clipped=2.0 +2023-04-01 15:38:52,992 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60366.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:38:59,662 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 15:39:00,756 INFO [train.py:903] (3/4) Epoch 9, batch 5750, loss[loss=0.2091, simple_loss=0.2788, pruned_loss=0.06967, over 19793.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.3153, pruned_loss=0.08721, over 3813482.47 frames. ], batch size: 48, lr: 9.09e-03, grad_scale: 8.0 +2023-04-01 15:39:07,415 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 15:39:12,820 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 15:39:36,979 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60402.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:39:50,842 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60414.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:40:00,906 INFO [train.py:903] (3/4) Epoch 9, batch 5800, loss[loss=0.205, simple_loss=0.2735, pruned_loss=0.06824, over 19766.00 frames. ], tot_loss[loss=0.2443, simple_loss=0.3148, pruned_loss=0.0869, over 3818899.18 frames. ], batch size: 47, lr: 9.09e-03, grad_scale: 8.0 +2023-04-01 15:40:06,307 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60427.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:40:12,306 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 15:40:22,428 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60440.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:40:45,425 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.333e+02 6.303e+02 7.771e+02 9.979e+02 2.257e+03, threshold=1.554e+03, percent-clipped=12.0 +2023-04-01 15:41:01,368 INFO [train.py:903] (3/4) Epoch 9, batch 5850, loss[loss=0.2327, simple_loss=0.3097, pruned_loss=0.07786, over 19782.00 frames. ], tot_loss[loss=0.2445, simple_loss=0.315, pruned_loss=0.08704, over 3826535.53 frames. ], batch size: 56, lr: 9.09e-03, grad_scale: 8.0 +2023-04-01 15:41:56,054 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60517.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:42:02,429 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 15:42:03,584 INFO [train.py:903] (3/4) Epoch 9, batch 5900, loss[loss=0.2149, simple_loss=0.2974, pruned_loss=0.06622, over 19539.00 frames. ], tot_loss[loss=0.2441, simple_loss=0.3147, pruned_loss=0.08671, over 3804852.22 frames. ], batch size: 54, lr: 9.08e-03, grad_scale: 8.0 +2023-04-01 15:42:10,690 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60529.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:42:22,838 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 15:42:25,393 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60542.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:42:42,841 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60555.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:42:46,125 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3236, 0.8752, 1.0097, 2.2426, 1.3880, 1.2295, 1.7751, 1.2499], + device='cuda:3'), covar=tensor([0.1291, 0.2095, 0.1603, 0.0968, 0.1363, 0.1649, 0.1304, 0.1261], + device='cuda:3'), in_proj_covar=tensor([0.0218, 0.0232, 0.0234, 0.0259, 0.0247, 0.0218, 0.0204, 0.0210], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 15:42:47,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.985e+02 5.460e+02 7.799e+02 9.723e+02 2.713e+03, threshold=1.560e+03, percent-clipped=1.0 +2023-04-01 15:42:48,019 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=60560.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:43:03,513 INFO [train.py:903] (3/4) Epoch 9, batch 5950, loss[loss=0.2836, simple_loss=0.342, pruned_loss=0.1126, over 18405.00 frames. ], tot_loss[loss=0.2455, simple_loss=0.3157, pruned_loss=0.08772, over 3805114.56 frames. ], batch size: 84, lr: 9.08e-03, grad_scale: 8.0 +2023-04-01 15:43:10,774 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4639, 2.2950, 1.7593, 1.3138, 2.1494, 1.1803, 1.2248, 1.9574], + device='cuda:3'), covar=tensor([0.0751, 0.0510, 0.0791, 0.0638, 0.0375, 0.1041, 0.0666, 0.0363], + device='cuda:3'), in_proj_covar=tensor([0.0290, 0.0298, 0.0326, 0.0245, 0.0233, 0.0324, 0.0292, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:43:13,154 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60581.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:43:20,764 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3404, 1.4634, 2.0082, 1.8065, 3.0181, 4.8227, 4.5913, 5.1202], + device='cuda:3'), covar=tensor([0.1513, 0.3081, 0.2721, 0.1672, 0.0438, 0.0110, 0.0135, 0.0070], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0288, 0.0317, 0.0245, 0.0209, 0.0145, 0.0202, 0.0183], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:43:45,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:44:03,653 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60622.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:44:04,412 INFO [train.py:903] (3/4) Epoch 9, batch 6000, loss[loss=0.2809, simple_loss=0.3493, pruned_loss=0.1063, over 19731.00 frames. ], tot_loss[loss=0.2462, simple_loss=0.3163, pruned_loss=0.08809, over 3810197.81 frames. ], batch size: 63, lr: 9.08e-03, grad_scale: 8.0 +2023-04-01 15:44:04,412 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 15:44:16,868 INFO [train.py:937] (3/4) Epoch 9, validation: loss=0.1828, simple_loss=0.2835, pruned_loss=0.04105, over 944034.00 frames. +2023-04-01 15:44:16,869 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 15:44:47,016 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60647.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:45:02,309 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.543e+02 5.785e+02 7.329e+02 9.590e+02 1.552e+03, threshold=1.466e+03, percent-clipped=0.0 +2023-04-01 15:45:17,221 INFO [train.py:903] (3/4) Epoch 9, batch 6050, loss[loss=0.2396, simple_loss=0.31, pruned_loss=0.08462, over 19590.00 frames. ], tot_loss[loss=0.2441, simple_loss=0.3148, pruned_loss=0.08674, over 3818582.03 frames. ], batch size: 52, lr: 9.07e-03, grad_scale: 8.0 +2023-04-01 15:45:20,651 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60675.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:45:41,565 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-04-01 15:46:18,269 INFO [train.py:903] (3/4) Epoch 9, batch 6100, loss[loss=0.2761, simple_loss=0.3464, pruned_loss=0.1029, over 19689.00 frames. ], tot_loss[loss=0.243, simple_loss=0.3142, pruned_loss=0.08595, over 3829532.43 frames. ], batch size: 59, lr: 9.07e-03, grad_scale: 8.0 +2023-04-01 15:46:34,606 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60736.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:47:03,006 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.337e+02 5.339e+02 6.367e+02 8.531e+02 1.806e+03, threshold=1.273e+03, percent-clipped=4.0 +2023-04-01 15:47:18,986 INFO [train.py:903] (3/4) Epoch 9, batch 6150, loss[loss=0.2212, simple_loss=0.289, pruned_loss=0.07666, over 19812.00 frames. ], tot_loss[loss=0.2432, simple_loss=0.3138, pruned_loss=0.0863, over 3828019.05 frames. ], batch size: 49, lr: 9.06e-03, grad_scale: 8.0 +2023-04-01 15:47:19,397 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60773.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:47:33,287 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60785.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:47:44,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 15:47:48,639 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60798.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:47:48,670 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60798.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:48:03,522 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60810.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:48:04,585 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60811.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:48:18,465 INFO [train.py:903] (3/4) Epoch 9, batch 6200, loss[loss=0.2605, simple_loss=0.3262, pruned_loss=0.09735, over 19542.00 frames. ], tot_loss[loss=0.2438, simple_loss=0.3142, pruned_loss=0.08665, over 3816282.40 frames. ], batch size: 56, lr: 9.06e-03, grad_scale: 8.0 +2023-04-01 15:48:19,773 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60823.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:48:34,270 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60836.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:49:03,643 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.706e+02 5.965e+02 7.614e+02 9.330e+02 2.107e+03, threshold=1.523e+03, percent-clipped=6.0 +2023-04-01 15:49:19,598 INFO [train.py:903] (3/4) Epoch 9, batch 6250, loss[loss=0.2712, simple_loss=0.3406, pruned_loss=0.1009, over 19603.00 frames. ], tot_loss[loss=0.2439, simple_loss=0.3147, pruned_loss=0.08657, over 3823123.04 frames. ], batch size: 61, lr: 9.06e-03, grad_scale: 8.0 +2023-04-01 15:49:49,664 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 15:50:20,302 INFO [train.py:903] (3/4) Epoch 9, batch 6300, loss[loss=0.2437, simple_loss=0.3235, pruned_loss=0.08196, over 19748.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.3151, pruned_loss=0.08738, over 3794219.71 frames. ], batch size: 63, lr: 9.05e-03, grad_scale: 8.0 +2023-04-01 15:50:31,290 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60931.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:51:01,601 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60956.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:51:05,845 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.132e+02 5.456e+02 6.999e+02 8.896e+02 1.665e+03, threshold=1.400e+03, percent-clipped=1.0 +2023-04-01 15:51:21,558 INFO [train.py:903] (3/4) Epoch 9, batch 6350, loss[loss=0.2485, simple_loss=0.3309, pruned_loss=0.08309, over 19294.00 frames. ], tot_loss[loss=0.2451, simple_loss=0.3154, pruned_loss=0.08739, over 3795087.61 frames. ], batch size: 66, lr: 9.05e-03, grad_scale: 8.0 +2023-04-01 15:51:59,711 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-04-01 15:52:08,288 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9474, 2.0225, 2.1162, 2.7454, 1.8769, 2.5804, 2.5881, 2.1405], + device='cuda:3'), covar=tensor([0.3273, 0.2774, 0.1337, 0.1557, 0.3167, 0.1273, 0.2903, 0.2328], + device='cuda:3'), in_proj_covar=tensor([0.0755, 0.0767, 0.0629, 0.0873, 0.0754, 0.0674, 0.0783, 0.0690], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 15:52:22,395 INFO [train.py:903] (3/4) Epoch 9, batch 6400, loss[loss=0.2388, simple_loss=0.3126, pruned_loss=0.0825, over 19451.00 frames. ], tot_loss[loss=0.2447, simple_loss=0.3149, pruned_loss=0.08724, over 3801226.76 frames. ], batch size: 64, lr: 9.05e-03, grad_scale: 8.0 +2023-04-01 15:52:25,007 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3042, 2.2921, 2.3771, 3.2815, 2.1631, 3.2505, 2.9309, 2.3710], + device='cuda:3'), covar=tensor([0.3471, 0.2948, 0.1368, 0.1776, 0.3703, 0.1318, 0.2979, 0.2360], + device='cuda:3'), in_proj_covar=tensor([0.0754, 0.0766, 0.0628, 0.0872, 0.0752, 0.0673, 0.0782, 0.0688], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 15:53:07,168 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.915e+02 6.196e+02 7.384e+02 8.672e+02 1.804e+03, threshold=1.477e+03, percent-clipped=4.0 +2023-04-01 15:53:22,487 INFO [train.py:903] (3/4) Epoch 9, batch 6450, loss[loss=0.288, simple_loss=0.3478, pruned_loss=0.1141, over 18706.00 frames. ], tot_loss[loss=0.2451, simple_loss=0.3154, pruned_loss=0.08739, over 3800220.43 frames. ], batch size: 74, lr: 9.04e-03, grad_scale: 8.0 +2023-04-01 15:53:30,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61079.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:53:31,426 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=61080.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:53:46,184 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-01 15:54:05,090 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 15:54:22,077 INFO [train.py:903] (3/4) Epoch 9, batch 6500, loss[loss=0.229, simple_loss=0.2989, pruned_loss=0.07958, over 19848.00 frames. ], tot_loss[loss=0.2462, simple_loss=0.3164, pruned_loss=0.08807, over 3814004.36 frames. ], batch size: 52, lr: 9.04e-03, grad_scale: 8.0 +2023-04-01 15:54:27,554 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 15:54:44,277 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1730, 2.0977, 1.7204, 1.5555, 1.5457, 1.6249, 0.2950, 0.9693], + device='cuda:3'), covar=tensor([0.0317, 0.0366, 0.0293, 0.0499, 0.0788, 0.0505, 0.0803, 0.0656], + device='cuda:3'), in_proj_covar=tensor([0.0320, 0.0316, 0.0318, 0.0334, 0.0411, 0.0334, 0.0295, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 15:55:06,066 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 6.706e+02 8.302e+02 1.012e+03 2.679e+03, threshold=1.660e+03, percent-clipped=7.0 +2023-04-01 15:55:21,994 INFO [train.py:903] (3/4) Epoch 9, batch 6550, loss[loss=0.2314, simple_loss=0.3025, pruned_loss=0.08016, over 19604.00 frames. ], tot_loss[loss=0.2448, simple_loss=0.3154, pruned_loss=0.08714, over 3832625.48 frames. ], batch size: 50, lr: 9.03e-03, grad_scale: 8.0 +2023-04-01 15:55:45,494 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2909, 3.7810, 3.8487, 3.8712, 1.4603, 3.6410, 3.1900, 3.5425], + device='cuda:3'), covar=tensor([0.1317, 0.0781, 0.0649, 0.0627, 0.5047, 0.0678, 0.0673, 0.1129], + device='cuda:3'), in_proj_covar=tensor([0.0624, 0.0558, 0.0746, 0.0627, 0.0689, 0.0496, 0.0459, 0.0687], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 15:55:50,205 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=61195.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 15:56:24,271 INFO [train.py:903] (3/4) Epoch 9, batch 6600, loss[loss=0.1934, simple_loss=0.2667, pruned_loss=0.06, over 19785.00 frames. ], tot_loss[loss=0.2439, simple_loss=0.3149, pruned_loss=0.08646, over 3831973.45 frames. ], batch size: 48, lr: 9.03e-03, grad_scale: 8.0 +2023-04-01 15:56:57,860 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5406, 1.3992, 1.3791, 1.7523, 1.4150, 1.7657, 1.7770, 1.6047], + device='cuda:3'), covar=tensor([0.0800, 0.0974, 0.1030, 0.0682, 0.0838, 0.0741, 0.0783, 0.0709], + device='cuda:3'), in_proj_covar=tensor([0.0216, 0.0230, 0.0231, 0.0257, 0.0246, 0.0217, 0.0202, 0.0210], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 15:57:09,254 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.208e+02 5.654e+02 6.784e+02 8.392e+02 1.741e+03, threshold=1.357e+03, percent-clipped=1.0 +2023-04-01 15:57:24,975 INFO [train.py:903] (3/4) Epoch 9, batch 6650, loss[loss=0.2269, simple_loss=0.2987, pruned_loss=0.07756, over 19750.00 frames. ], tot_loss[loss=0.2438, simple_loss=0.3147, pruned_loss=0.08648, over 3838898.76 frames. ], batch size: 48, lr: 9.03e-03, grad_scale: 8.0 +2023-04-01 15:57:50,974 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2646, 1.3850, 1.8871, 1.5806, 2.6158, 2.3487, 2.8486, 1.1128], + device='cuda:3'), covar=tensor([0.2083, 0.3524, 0.1982, 0.1565, 0.1321, 0.1667, 0.1312, 0.3402], + device='cuda:3'), in_proj_covar=tensor([0.0476, 0.0554, 0.0570, 0.0427, 0.0586, 0.0479, 0.0639, 0.0479], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 15:58:06,963 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9136, 1.6050, 1.4627, 2.0984, 1.8689, 1.6311, 1.5143, 1.7758], + device='cuda:3'), covar=tensor([0.0917, 0.1581, 0.1366, 0.0816, 0.1104, 0.0537, 0.1190, 0.0680], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0354, 0.0293, 0.0239, 0.0296, 0.0244, 0.0275, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:58:11,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0871, 1.6581, 1.5855, 2.0911, 1.8744, 1.7828, 1.5770, 1.8984], + device='cuda:3'), covar=tensor([0.0821, 0.1479, 0.1365, 0.0859, 0.1040, 0.0483, 0.1121, 0.0614], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0354, 0.0292, 0.0239, 0.0296, 0.0244, 0.0275, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 15:58:25,585 INFO [train.py:903] (3/4) Epoch 9, batch 6700, loss[loss=0.3094, simple_loss=0.3601, pruned_loss=0.1294, over 19616.00 frames. ], tot_loss[loss=0.2449, simple_loss=0.3156, pruned_loss=0.08707, over 3839877.02 frames. ], batch size: 57, lr: 9.02e-03, grad_scale: 8.0 +2023-04-01 15:59:08,506 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.478e+02 5.683e+02 7.501e+02 9.618e+02 2.603e+03, threshold=1.500e+03, percent-clipped=7.0 +2023-04-01 15:59:23,017 INFO [train.py:903] (3/4) Epoch 9, batch 6750, loss[loss=0.2082, simple_loss=0.2917, pruned_loss=0.06235, over 19758.00 frames. ], tot_loss[loss=0.2442, simple_loss=0.3147, pruned_loss=0.08682, over 3841418.06 frames. ], batch size: 54, lr: 9.02e-03, grad_scale: 8.0 +2023-04-01 15:59:44,572 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8464, 1.9636, 2.0552, 2.7154, 1.8560, 2.6382, 2.4936, 1.9924], + device='cuda:3'), covar=tensor([0.3333, 0.2625, 0.1278, 0.1631, 0.3180, 0.1236, 0.2858, 0.2328], + device='cuda:3'), in_proj_covar=tensor([0.0762, 0.0774, 0.0632, 0.0878, 0.0760, 0.0676, 0.0784, 0.0690], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 15:59:47,297 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-04-01 16:00:18,968 INFO [train.py:903] (3/4) Epoch 9, batch 6800, loss[loss=0.2131, simple_loss=0.2966, pruned_loss=0.0648, over 19766.00 frames. ], tot_loss[loss=0.2436, simple_loss=0.3144, pruned_loss=0.08641, over 3849120.43 frames. ], batch size: 54, lr: 9.02e-03, grad_scale: 8.0 +2023-04-01 16:00:19,093 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=61423.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:01:03,176 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 16:01:04,245 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 16:01:06,621 INFO [train.py:903] (3/4) Epoch 10, batch 0, loss[loss=0.2698, simple_loss=0.3386, pruned_loss=0.1005, over 19599.00 frames. ], tot_loss[loss=0.2698, simple_loss=0.3386, pruned_loss=0.1005, over 19599.00 frames. ], batch size: 61, lr: 8.57e-03, grad_scale: 8.0 +2023-04-01 16:01:06,621 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 16:01:17,503 INFO [train.py:937] (3/4) Epoch 10, validation: loss=0.1825, simple_loss=0.2836, pruned_loss=0.04072, over 944034.00 frames. +2023-04-01 16:01:17,504 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 16:01:17,992 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61451.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:01:21,304 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3740, 2.2503, 2.3590, 3.3039, 2.1378, 3.3326, 2.9202, 2.2290], + device='cuda:3'), covar=tensor([0.3500, 0.3022, 0.1434, 0.1723, 0.3667, 0.1279, 0.2919, 0.2533], + device='cuda:3'), in_proj_covar=tensor([0.0759, 0.0770, 0.0630, 0.0873, 0.0757, 0.0673, 0.0781, 0.0688], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 16:01:27,607 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.284e+02 5.689e+02 6.760e+02 8.116e+02 1.440e+03, threshold=1.352e+03, percent-clipped=0.0 +2023-04-01 16:01:29,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 16:01:48,088 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61476.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:02:17,430 INFO [train.py:903] (3/4) Epoch 10, batch 50, loss[loss=0.2356, simple_loss=0.3092, pruned_loss=0.08105, over 19721.00 frames. ], tot_loss[loss=0.2413, simple_loss=0.3135, pruned_loss=0.08455, over 857879.49 frames. ], batch size: 63, lr: 8.57e-03, grad_scale: 8.0 +2023-04-01 16:02:50,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 16:02:56,286 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9232, 3.4118, 2.0216, 2.1550, 3.0368, 1.7119, 1.3131, 1.8622], + device='cuda:3'), covar=tensor([0.1107, 0.0566, 0.0923, 0.0646, 0.0422, 0.1078, 0.0926, 0.0738], + device='cuda:3'), in_proj_covar=tensor([0.0290, 0.0300, 0.0324, 0.0245, 0.0234, 0.0326, 0.0289, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:03:03,254 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=61538.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:03:18,485 INFO [train.py:903] (3/4) Epoch 10, batch 100, loss[loss=0.2231, simple_loss=0.3073, pruned_loss=0.06939, over 19672.00 frames. ], tot_loss[loss=0.2447, simple_loss=0.3157, pruned_loss=0.08681, over 1516446.05 frames. ], batch size: 58, lr: 8.57e-03, grad_scale: 8.0 +2023-04-01 16:03:24,171 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 16:03:29,311 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.995e+02 6.266e+02 7.755e+02 9.384e+02 2.029e+03, threshold=1.551e+03, percent-clipped=6.0 +2023-04-01 16:04:19,612 INFO [train.py:903] (3/4) Epoch 10, batch 150, loss[loss=0.201, simple_loss=0.2699, pruned_loss=0.06604, over 19797.00 frames. ], tot_loss[loss=0.2431, simple_loss=0.3139, pruned_loss=0.08617, over 2020961.37 frames. ], batch size: 48, lr: 8.56e-03, grad_scale: 16.0 +2023-04-01 16:05:12,386 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 16:05:20,102 INFO [train.py:903] (3/4) Epoch 10, batch 200, loss[loss=0.2379, simple_loss=0.3123, pruned_loss=0.08177, over 19582.00 frames. ], tot_loss[loss=0.2434, simple_loss=0.3145, pruned_loss=0.08608, over 2431353.49 frames. ], batch size: 52, lr: 8.56e-03, grad_scale: 8.0 +2023-04-01 16:05:32,341 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.478e+02 5.325e+02 6.934e+02 9.117e+02 1.602e+03, threshold=1.387e+03, percent-clipped=3.0 +2023-04-01 16:05:34,707 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6744, 2.2888, 2.1005, 2.7570, 2.6422, 2.4025, 2.2524, 2.5053], + device='cuda:3'), covar=tensor([0.0754, 0.1533, 0.1223, 0.0878, 0.1060, 0.0396, 0.0932, 0.0547], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0354, 0.0289, 0.0238, 0.0297, 0.0243, 0.0274, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:05:35,095 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-01 16:05:45,453 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9289, 4.3768, 4.6541, 4.6378, 2.0119, 4.3084, 3.7773, 4.2960], + device='cuda:3'), covar=tensor([0.1313, 0.0794, 0.0512, 0.0514, 0.4621, 0.0667, 0.0608, 0.1020], + device='cuda:3'), in_proj_covar=tensor([0.0632, 0.0563, 0.0751, 0.0633, 0.0694, 0.0506, 0.0465, 0.0701], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 16:06:20,989 INFO [train.py:903] (3/4) Epoch 10, batch 250, loss[loss=0.2281, simple_loss=0.3067, pruned_loss=0.07478, over 17390.00 frames. ], tot_loss[loss=0.2433, simple_loss=0.3146, pruned_loss=0.08604, over 2743325.86 frames. ], batch size: 101, lr: 8.56e-03, grad_scale: 8.0 +2023-04-01 16:07:19,589 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:07:20,496 INFO [train.py:903] (3/4) Epoch 10, batch 300, loss[loss=0.3128, simple_loss=0.3622, pruned_loss=0.1317, over 13685.00 frames. ], tot_loss[loss=0.2444, simple_loss=0.3152, pruned_loss=0.0868, over 2975757.47 frames. ], batch size: 136, lr: 8.55e-03, grad_scale: 8.0 +2023-04-01 16:07:32,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.696e+02 5.563e+02 6.785e+02 8.281e+02 1.821e+03, threshold=1.357e+03, percent-clipped=1.0 +2023-04-01 16:07:33,036 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61761.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:08:12,970 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61794.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:08:20,599 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 16:08:21,871 INFO [train.py:903] (3/4) Epoch 10, batch 350, loss[loss=0.2555, simple_loss=0.3257, pruned_loss=0.09267, over 19541.00 frames. ], tot_loss[loss=0.2439, simple_loss=0.3147, pruned_loss=0.08651, over 3156292.03 frames. ], batch size: 56, lr: 8.55e-03, grad_scale: 8.0 +2023-04-01 16:08:44,880 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61819.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:09:23,123 INFO [train.py:903] (3/4) Epoch 10, batch 400, loss[loss=0.2706, simple_loss=0.3425, pruned_loss=0.09935, over 19679.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.3158, pruned_loss=0.08769, over 3301553.93 frames. ], batch size: 53, lr: 8.54e-03, grad_scale: 8.0 +2023-04-01 16:09:36,144 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.162e+02 5.438e+02 6.846e+02 8.745e+02 2.106e+03, threshold=1.369e+03, percent-clipped=7.0 +2023-04-01 16:10:26,713 INFO [train.py:903] (3/4) Epoch 10, batch 450, loss[loss=0.2435, simple_loss=0.321, pruned_loss=0.08302, over 19083.00 frames. ], tot_loss[loss=0.245, simple_loss=0.3156, pruned_loss=0.08719, over 3400091.60 frames. ], batch size: 69, lr: 8.54e-03, grad_scale: 8.0 +2023-04-01 16:10:29,711 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-01 16:10:50,935 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 16:10:50,968 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 16:11:05,207 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61932.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:11:19,197 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2315, 4.2913, 4.7788, 4.7651, 2.5158, 4.4742, 4.0703, 4.4897], + device='cuda:3'), covar=tensor([0.1030, 0.2606, 0.0483, 0.0481, 0.3890, 0.0581, 0.0475, 0.0858], + device='cuda:3'), in_proj_covar=tensor([0.0624, 0.0559, 0.0743, 0.0631, 0.0687, 0.0500, 0.0460, 0.0690], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 16:11:27,991 INFO [train.py:903] (3/4) Epoch 10, batch 500, loss[loss=0.2707, simple_loss=0.3347, pruned_loss=0.1033, over 19675.00 frames. ], tot_loss[loss=0.2447, simple_loss=0.3157, pruned_loss=0.08691, over 3504379.24 frames. ], batch size: 53, lr: 8.54e-03, grad_scale: 8.0 +2023-04-01 16:11:30,098 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-04-01 16:11:39,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.771e+02 5.473e+02 6.472e+02 7.882e+02 1.512e+03, threshold=1.294e+03, percent-clipped=2.0 +2023-04-01 16:12:30,820 INFO [train.py:903] (3/4) Epoch 10, batch 550, loss[loss=0.2442, simple_loss=0.3182, pruned_loss=0.08513, over 19469.00 frames. ], tot_loss[loss=0.2443, simple_loss=0.3153, pruned_loss=0.08664, over 3583186.40 frames. ], batch size: 49, lr: 8.53e-03, grad_scale: 8.0 +2023-04-01 16:13:32,118 INFO [train.py:903] (3/4) Epoch 10, batch 600, loss[loss=0.2078, simple_loss=0.287, pruned_loss=0.06432, over 19766.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3128, pruned_loss=0.08498, over 3632954.71 frames. ], batch size: 54, lr: 8.53e-03, grad_scale: 8.0 +2023-04-01 16:13:42,211 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-01 16:13:46,037 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.499e+02 5.366e+02 7.096e+02 8.541e+02 1.663e+03, threshold=1.419e+03, percent-clipped=2.0 +2023-04-01 16:14:11,915 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 16:14:26,439 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62094.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:14:35,311 INFO [train.py:903] (3/4) Epoch 10, batch 650, loss[loss=0.2609, simple_loss=0.3284, pruned_loss=0.09674, over 17382.00 frames. ], tot_loss[loss=0.2409, simple_loss=0.3121, pruned_loss=0.08484, over 3674913.58 frames. ], batch size: 101, lr: 8.53e-03, grad_scale: 8.0 +2023-04-01 16:14:40,071 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62105.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:15:24,628 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62140.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:15:39,370 INFO [train.py:903] (3/4) Epoch 10, batch 700, loss[loss=0.2402, simple_loss=0.314, pruned_loss=0.08314, over 19299.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.3123, pruned_loss=0.08456, over 3713313.03 frames. ], batch size: 66, lr: 8.52e-03, grad_scale: 8.0 +2023-04-01 16:15:51,155 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.410e+02 5.986e+02 7.007e+02 9.230e+02 2.462e+03, threshold=1.401e+03, percent-clipped=6.0 +2023-04-01 16:15:51,679 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2237, 1.2805, 1.6081, 1.5030, 2.3376, 2.0330, 2.3750, 0.8573], + device='cuda:3'), covar=tensor([0.2348, 0.4098, 0.2367, 0.1851, 0.1435, 0.2060, 0.1401, 0.3871], + device='cuda:3'), in_proj_covar=tensor([0.0476, 0.0557, 0.0575, 0.0426, 0.0585, 0.0480, 0.0639, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 16:16:41,947 INFO [train.py:903] (3/4) Epoch 10, batch 750, loss[loss=0.2198, simple_loss=0.284, pruned_loss=0.07777, over 18705.00 frames. ], tot_loss[loss=0.2412, simple_loss=0.3122, pruned_loss=0.08505, over 3732674.37 frames. ], batch size: 41, lr: 8.52e-03, grad_scale: 8.0 +2023-04-01 16:16:51,086 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62209.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:17:04,546 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62220.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:17:11,246 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3906, 1.3244, 1.4573, 1.5775, 2.8948, 1.0610, 2.1195, 3.2365], + device='cuda:3'), covar=tensor([0.0466, 0.2607, 0.2583, 0.1587, 0.0783, 0.2373, 0.1143, 0.0309], + device='cuda:3'), in_proj_covar=tensor([0.0338, 0.0326, 0.0337, 0.0311, 0.0338, 0.0322, 0.0317, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:17:35,223 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.67 vs. limit=5.0 +2023-04-01 16:17:42,578 INFO [train.py:903] (3/4) Epoch 10, batch 800, loss[loss=0.2409, simple_loss=0.3241, pruned_loss=0.07887, over 19617.00 frames. ], tot_loss[loss=0.2422, simple_loss=0.3129, pruned_loss=0.08573, over 3749362.37 frames. ], batch size: 50, lr: 8.52e-03, grad_scale: 8.0 +2023-04-01 16:17:54,856 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.040e+02 5.237e+02 6.930e+02 8.487e+02 1.526e+03, threshold=1.386e+03, percent-clipped=2.0 +2023-04-01 16:17:58,879 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 16:18:15,001 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62276.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:18:32,358 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62291.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:18:43,719 INFO [train.py:903] (3/4) Epoch 10, batch 850, loss[loss=0.2372, simple_loss=0.3127, pruned_loss=0.08084, over 19735.00 frames. ], tot_loss[loss=0.2415, simple_loss=0.3125, pruned_loss=0.08525, over 3777182.60 frames. ], batch size: 51, lr: 8.51e-03, grad_scale: 8.0 +2023-04-01 16:19:37,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 16:19:45,727 INFO [train.py:903] (3/4) Epoch 10, batch 900, loss[loss=0.2269, simple_loss=0.3063, pruned_loss=0.07379, over 17579.00 frames. ], tot_loss[loss=0.2424, simple_loss=0.3133, pruned_loss=0.08572, over 3780581.46 frames. ], batch size: 101, lr: 8.51e-03, grad_scale: 8.0 +2023-04-01 16:19:59,162 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.268e+02 6.023e+02 7.076e+02 9.770e+02 2.916e+03, threshold=1.415e+03, percent-clipped=7.0 +2023-04-01 16:20:35,903 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62391.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:20:48,986 INFO [train.py:903] (3/4) Epoch 10, batch 950, loss[loss=0.2601, simple_loss=0.3363, pruned_loss=0.092, over 19701.00 frames. ], tot_loss[loss=0.2417, simple_loss=0.313, pruned_loss=0.08526, over 3787537.37 frames. ], batch size: 59, lr: 8.51e-03, grad_scale: 8.0 +2023-04-01 16:20:50,159 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 16:20:58,685 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62409.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:21:50,240 INFO [train.py:903] (3/4) Epoch 10, batch 1000, loss[loss=0.2581, simple_loss=0.3324, pruned_loss=0.09194, over 19678.00 frames. ], tot_loss[loss=0.242, simple_loss=0.313, pruned_loss=0.08552, over 3799239.78 frames. ], batch size: 60, lr: 8.50e-03, grad_scale: 8.0 +2023-04-01 16:22:01,598 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.402e+02 5.469e+02 6.659e+02 8.311e+02 1.987e+03, threshold=1.332e+03, percent-clipped=4.0 +2023-04-01 16:22:07,629 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62465.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:22:21,508 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62476.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:22:31,382 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62484.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:22:38,713 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62490.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:22:43,037 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 16:22:51,163 INFO [train.py:903] (3/4) Epoch 10, batch 1050, loss[loss=0.2544, simple_loss=0.3322, pruned_loss=0.08828, over 17959.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3118, pruned_loss=0.08439, over 3818017.16 frames. ], batch size: 83, lr: 8.50e-03, grad_scale: 8.0 +2023-04-01 16:22:51,565 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62501.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:23:23,743 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 16:23:54,503 INFO [train.py:903] (3/4) Epoch 10, batch 1100, loss[loss=0.293, simple_loss=0.3625, pruned_loss=0.1117, over 19072.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3119, pruned_loss=0.0846, over 3814520.95 frames. ], batch size: 69, lr: 8.50e-03, grad_scale: 8.0 +2023-04-01 16:24:07,765 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.238e+02 5.569e+02 6.927e+02 9.101e+02 1.941e+03, threshold=1.385e+03, percent-clipped=3.0 +2023-04-01 16:24:10,211 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62563.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:24:54,321 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62599.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:24:56,396 INFO [train.py:903] (3/4) Epoch 10, batch 1150, loss[loss=0.2702, simple_loss=0.332, pruned_loss=0.1042, over 19553.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.3111, pruned_loss=0.08429, over 3828967.01 frames. ], batch size: 56, lr: 8.49e-03, grad_scale: 8.0 +2023-04-01 16:24:59,699 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62603.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:25:19,247 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2253, 2.0780, 1.8388, 1.6401, 1.4609, 1.7023, 0.3764, 1.0397], + device='cuda:3'), covar=tensor([0.0327, 0.0364, 0.0264, 0.0424, 0.0836, 0.0499, 0.0772, 0.0646], + device='cuda:3'), in_proj_covar=tensor([0.0325, 0.0321, 0.0325, 0.0338, 0.0411, 0.0335, 0.0302, 0.0315], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 16:25:21,134 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7927, 4.3628, 2.7331, 3.8964, 1.2346, 4.0244, 4.0901, 4.1642], + device='cuda:3'), covar=tensor([0.0536, 0.0933, 0.1764, 0.0684, 0.3444, 0.0690, 0.0689, 0.0794], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0352, 0.0422, 0.0309, 0.0371, 0.0346, 0.0342, 0.0373], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 16:25:37,691 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62635.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:25:53,855 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62647.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:25:56,072 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3060, 3.0263, 2.3000, 2.8334, 1.1258, 2.8818, 2.8229, 2.8900], + device='cuda:3'), covar=tensor([0.1083, 0.1334, 0.1796, 0.0855, 0.3088, 0.0936, 0.0945, 0.1270], + device='cuda:3'), in_proj_covar=tensor([0.0411, 0.0350, 0.0420, 0.0308, 0.0369, 0.0344, 0.0339, 0.0370], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 16:25:58,136 INFO [train.py:903] (3/4) Epoch 10, batch 1200, loss[loss=0.2121, simple_loss=0.2784, pruned_loss=0.07292, over 19757.00 frames. ], tot_loss[loss=0.2394, simple_loss=0.3106, pruned_loss=0.08414, over 3837074.43 frames. ], batch size: 45, lr: 8.49e-03, grad_scale: 8.0 +2023-04-01 16:26:09,504 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.384e+02 5.581e+02 6.902e+02 9.588e+02 2.703e+03, threshold=1.380e+03, percent-clipped=8.0 +2023-04-01 16:26:24,966 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62672.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:26:32,529 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 16:26:40,542 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62684.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:26:59,720 INFO [train.py:903] (3/4) Epoch 10, batch 1250, loss[loss=0.2853, simple_loss=0.3459, pruned_loss=0.1123, over 19359.00 frames. ], tot_loss[loss=0.2395, simple_loss=0.3107, pruned_loss=0.08409, over 3836221.01 frames. ], batch size: 66, lr: 8.49e-03, grad_scale: 4.0 +2023-04-01 16:28:00,865 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:28:01,682 INFO [train.py:903] (3/4) Epoch 10, batch 1300, loss[loss=0.2504, simple_loss=0.3218, pruned_loss=0.08945, over 19378.00 frames. ], tot_loss[loss=0.239, simple_loss=0.3099, pruned_loss=0.08401, over 3831405.74 frames. ], batch size: 70, lr: 8.48e-03, grad_scale: 4.0 +2023-04-01 16:28:05,058 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62753.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:28:16,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.133e+02 5.064e+02 6.596e+02 8.862e+02 1.920e+03, threshold=1.319e+03, percent-clipped=1.0 +2023-04-01 16:28:27,818 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4345, 2.2106, 1.6138, 1.5717, 2.1379, 1.1748, 1.3348, 1.6714], + device='cuda:3'), covar=tensor([0.0821, 0.0682, 0.0899, 0.0571, 0.0446, 0.1090, 0.0593, 0.0424], + device='cuda:3'), in_proj_covar=tensor([0.0286, 0.0298, 0.0323, 0.0241, 0.0233, 0.0324, 0.0286, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:28:51,840 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62791.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:29:05,051 INFO [train.py:903] (3/4) Epoch 10, batch 1350, loss[loss=0.2471, simple_loss=0.3263, pruned_loss=0.08398, over 19042.00 frames. ], tot_loss[loss=0.2401, simple_loss=0.3112, pruned_loss=0.08456, over 3825639.39 frames. ], batch size: 69, lr: 8.48e-03, grad_scale: 4.0 +2023-04-01 16:29:16,161 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.68 vs. limit=2.0 +2023-04-01 16:29:34,259 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8289, 1.4809, 1.6123, 1.5387, 3.2875, 0.9591, 2.2438, 3.7256], + device='cuda:3'), covar=tensor([0.0390, 0.2301, 0.2357, 0.1666, 0.0736, 0.2444, 0.1212, 0.0246], + device='cuda:3'), in_proj_covar=tensor([0.0343, 0.0326, 0.0342, 0.0313, 0.0339, 0.0324, 0.0320, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:30:07,946 INFO [train.py:903] (3/4) Epoch 10, batch 1400, loss[loss=0.2494, simple_loss=0.3305, pruned_loss=0.08412, over 19542.00 frames. ], tot_loss[loss=0.2391, simple_loss=0.3105, pruned_loss=0.08385, over 3834147.68 frames. ], batch size: 56, lr: 8.48e-03, grad_scale: 4.0 +2023-04-01 16:30:13,130 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62855.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:30:20,949 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.159e+02 5.591e+02 6.749e+02 8.217e+02 1.554e+03, threshold=1.350e+03, percent-clipped=4.0 +2023-04-01 16:30:27,972 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62868.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:30:44,902 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62880.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:31:07,150 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 16:31:09,396 INFO [train.py:903] (3/4) Epoch 10, batch 1450, loss[loss=0.2481, simple_loss=0.3226, pruned_loss=0.08681, over 19396.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3122, pruned_loss=0.08525, over 3798792.86 frames. ], batch size: 70, lr: 8.47e-03, grad_scale: 4.0 +2023-04-01 16:31:16,534 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62907.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:32:07,386 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=62947.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:32:11,884 INFO [train.py:903] (3/4) Epoch 10, batch 1500, loss[loss=0.2353, simple_loss=0.3166, pruned_loss=0.077, over 19625.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3116, pruned_loss=0.08471, over 3815329.73 frames. ], batch size: 57, lr: 8.47e-03, grad_scale: 4.0 +2023-04-01 16:32:27,777 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.889e+02 5.675e+02 6.883e+02 8.252e+02 2.690e+03, threshold=1.377e+03, percent-clipped=4.0 +2023-04-01 16:33:08,486 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.84 vs. limit=5.0 +2023-04-01 16:33:17,112 INFO [train.py:903] (3/4) Epoch 10, batch 1550, loss[loss=0.209, simple_loss=0.2894, pruned_loss=0.06427, over 18611.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3118, pruned_loss=0.08469, over 3817905.34 frames. ], batch size: 41, lr: 8.47e-03, grad_scale: 4.0 +2023-04-01 16:33:23,579 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63006.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:33:33,747 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5449, 1.1160, 1.3158, 1.2176, 2.1712, 0.8748, 1.8809, 2.3274], + device='cuda:3'), covar=tensor([0.0653, 0.2574, 0.2593, 0.1473, 0.0860, 0.2046, 0.0957, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0324, 0.0338, 0.0311, 0.0338, 0.0322, 0.0318, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:33:42,922 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63022.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:33:49,634 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63028.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:33:53,315 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63031.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:34:20,243 INFO [train.py:903] (3/4) Epoch 10, batch 1600, loss[loss=0.2567, simple_loss=0.3284, pruned_loss=0.09252, over 19675.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.3122, pruned_loss=0.08457, over 3818638.27 frames. ], batch size: 58, lr: 8.46e-03, grad_scale: 8.0 +2023-04-01 16:34:33,008 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.609e+02 5.367e+02 6.713e+02 8.573e+02 1.582e+03, threshold=1.343e+03, percent-clipped=2.0 +2023-04-01 16:34:33,419 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63062.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:34:42,301 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 16:35:21,340 INFO [train.py:903] (3/4) Epoch 10, batch 1650, loss[loss=0.206, simple_loss=0.2828, pruned_loss=0.06459, over 19488.00 frames. ], tot_loss[loss=0.2408, simple_loss=0.312, pruned_loss=0.0848, over 3814964.79 frames. ], batch size: 49, lr: 8.46e-03, grad_scale: 8.0 +2023-04-01 16:35:51,917 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63124.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:36:05,479 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63135.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:36:14,719 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63143.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:36:21,878 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63149.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:36:23,854 INFO [train.py:903] (3/4) Epoch 10, batch 1700, loss[loss=0.2441, simple_loss=0.3161, pruned_loss=0.08608, over 19530.00 frames. ], tot_loss[loss=0.2409, simple_loss=0.3122, pruned_loss=0.08476, over 3823285.25 frames. ], batch size: 54, lr: 8.46e-03, grad_scale: 8.0 +2023-04-01 16:36:38,454 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.334e+02 5.819e+02 7.178e+02 9.040e+02 2.117e+03, threshold=1.436e+03, percent-clipped=7.0 +2023-04-01 16:36:54,037 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5322, 1.9401, 2.1202, 2.1655, 3.2763, 1.7277, 2.7922, 3.3496], + device='cuda:3'), covar=tensor([0.0379, 0.1825, 0.1830, 0.1243, 0.0498, 0.1876, 0.1169, 0.0281], + device='cuda:3'), in_proj_covar=tensor([0.0340, 0.0325, 0.0339, 0.0311, 0.0338, 0.0324, 0.0319, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:37:04,089 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 16:37:06,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6694, 1.7822, 1.5610, 2.5949, 1.7189, 2.3885, 2.0102, 1.3407], + device='cuda:3'), covar=tensor([0.4166, 0.3489, 0.2362, 0.2109, 0.3660, 0.1723, 0.4577, 0.4241], + device='cuda:3'), in_proj_covar=tensor([0.0771, 0.0780, 0.0634, 0.0880, 0.0758, 0.0678, 0.0779, 0.0691], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 16:37:28,659 INFO [train.py:903] (3/4) Epoch 10, batch 1750, loss[loss=0.2184, simple_loss=0.3045, pruned_loss=0.06618, over 19475.00 frames. ], tot_loss[loss=0.2415, simple_loss=0.3128, pruned_loss=0.08512, over 3822243.79 frames. ], batch size: 64, lr: 8.45e-03, grad_scale: 4.0 +2023-04-01 16:37:36,150 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4228, 2.3272, 1.9006, 1.9149, 1.7524, 1.9405, 0.7885, 1.4450], + device='cuda:3'), covar=tensor([0.0372, 0.0375, 0.0319, 0.0462, 0.0709, 0.0529, 0.0756, 0.0631], + device='cuda:3'), in_proj_covar=tensor([0.0326, 0.0322, 0.0326, 0.0338, 0.0413, 0.0338, 0.0303, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 16:38:14,973 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63239.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:38:30,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:38:30,877 INFO [train.py:903] (3/4) Epoch 10, batch 1800, loss[loss=0.278, simple_loss=0.333, pruned_loss=0.1115, over 19773.00 frames. ], tot_loss[loss=0.242, simple_loss=0.3132, pruned_loss=0.0854, over 3818613.61 frames. ], batch size: 54, lr: 8.45e-03, grad_scale: 4.0 +2023-04-01 16:38:44,560 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.600e+02 5.870e+02 7.659e+02 9.293e+02 2.596e+03, threshold=1.532e+03, percent-clipped=8.0 +2023-04-01 16:39:03,325 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63278.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:39:16,014 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4807, 1.8204, 2.0266, 1.9776, 3.1932, 1.5029, 2.6553, 3.3124], + device='cuda:3'), covar=tensor([0.0472, 0.2107, 0.2041, 0.1449, 0.0598, 0.2123, 0.1573, 0.0352], + device='cuda:3'), in_proj_covar=tensor([0.0343, 0.0327, 0.0341, 0.0315, 0.0340, 0.0325, 0.0322, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:39:30,790 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 16:39:32,967 INFO [train.py:903] (3/4) Epoch 10, batch 1850, loss[loss=0.2308, simple_loss=0.3029, pruned_loss=0.07937, over 19629.00 frames. ], tot_loss[loss=0.2421, simple_loss=0.3132, pruned_loss=0.08548, over 3825945.69 frames. ], batch size: 50, lr: 8.45e-03, grad_scale: 4.0 +2023-04-01 16:39:35,776 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63303.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:39:54,136 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63318.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:40:10,162 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 16:40:26,771 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63343.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:40:35,821 INFO [train.py:903] (3/4) Epoch 10, batch 1900, loss[loss=0.2523, simple_loss=0.3301, pruned_loss=0.08721, over 19490.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3122, pruned_loss=0.08447, over 3831889.70 frames. ], batch size: 64, lr: 8.44e-03, grad_scale: 4.0 +2023-04-01 16:40:45,541 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-01 16:40:52,656 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.343e+02 5.562e+02 7.038e+02 8.618e+02 1.834e+03, threshold=1.408e+03, percent-clipped=3.0 +2023-04-01 16:40:55,973 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 16:41:02,768 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 16:41:24,106 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3792, 1.1632, 1.3499, 1.3487, 2.9701, 0.8081, 2.0507, 3.1716], + device='cuda:3'), covar=tensor([0.0459, 0.2534, 0.2669, 0.1691, 0.0715, 0.2427, 0.1228, 0.0327], + device='cuda:3'), in_proj_covar=tensor([0.0344, 0.0328, 0.0345, 0.0314, 0.0343, 0.0325, 0.0323, 0.0346], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:41:24,920 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 16:41:37,235 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63399.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:41:40,410 INFO [train.py:903] (3/4) Epoch 10, batch 1950, loss[loss=0.2221, simple_loss=0.2972, pruned_loss=0.07349, over 19495.00 frames. ], tot_loss[loss=0.2423, simple_loss=0.3133, pruned_loss=0.0856, over 3819263.92 frames. ], batch size: 49, lr: 8.44e-03, grad_scale: 4.0 +2023-04-01 16:42:10,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63424.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:42:27,987 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63438.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:42:31,462 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63440.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:42:44,764 INFO [train.py:903] (3/4) Epoch 10, batch 2000, loss[loss=0.2101, simple_loss=0.2866, pruned_loss=0.06681, over 19710.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3131, pruned_loss=0.08486, over 3834479.03 frames. ], batch size: 51, lr: 8.44e-03, grad_scale: 8.0 +2023-04-01 16:42:58,757 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.14 vs. limit=5.0 +2023-04-01 16:43:00,248 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.731e+02 5.221e+02 6.641e+02 8.776e+02 2.044e+03, threshold=1.328e+03, percent-clipped=3.0 +2023-04-01 16:43:36,551 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4478, 1.4604, 1.8918, 1.7063, 3.0932, 2.8430, 3.4540, 1.4338], + device='cuda:3'), covar=tensor([0.2065, 0.3607, 0.2163, 0.1562, 0.1403, 0.1510, 0.1469, 0.3344], + device='cuda:3'), in_proj_covar=tensor([0.0475, 0.0555, 0.0577, 0.0425, 0.0582, 0.0475, 0.0637, 0.0481], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 16:43:43,958 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 16:43:47,388 INFO [train.py:903] (3/4) Epoch 10, batch 2050, loss[loss=0.2701, simple_loss=0.3385, pruned_loss=0.1009, over 19524.00 frames. ], tot_loss[loss=0.24, simple_loss=0.3119, pruned_loss=0.08406, over 3833319.37 frames. ], batch size: 54, lr: 8.43e-03, grad_scale: 8.0 +2023-04-01 16:43:53,589 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:44:03,926 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 16:44:05,967 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 16:44:26,408 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63531.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:44:28,410 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 16:44:35,229 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-01 16:44:35,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5951, 4.1027, 2.6248, 3.7253, 1.1946, 3.8849, 3.9055, 4.0582], + device='cuda:3'), covar=tensor([0.0599, 0.1037, 0.1833, 0.0752, 0.3533, 0.0734, 0.0648, 0.0920], + device='cuda:3'), in_proj_covar=tensor([0.0416, 0.0353, 0.0421, 0.0311, 0.0372, 0.0351, 0.0344, 0.0374], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 16:44:50,578 INFO [train.py:903] (3/4) Epoch 10, batch 2100, loss[loss=0.2688, simple_loss=0.3412, pruned_loss=0.09818, over 19523.00 frames. ], tot_loss[loss=0.2394, simple_loss=0.311, pruned_loss=0.0839, over 3835015.41 frames. ], batch size: 54, lr: 8.43e-03, grad_scale: 8.0 +2023-04-01 16:45:06,403 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.339e+02 5.645e+02 7.348e+02 9.688e+02 2.351e+03, threshold=1.470e+03, percent-clipped=4.0 +2023-04-01 16:45:22,180 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1183, 1.1815, 1.6384, 1.0178, 2.6679, 3.3222, 3.0818, 3.5312], + device='cuda:3'), covar=tensor([0.1539, 0.3304, 0.2922, 0.2120, 0.0436, 0.0146, 0.0210, 0.0171], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0292, 0.0321, 0.0249, 0.0212, 0.0151, 0.0205, 0.0188], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 16:45:26,358 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 16:45:32,506 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63583.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:45:33,940 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3674, 2.1601, 1.9445, 1.7758, 1.5471, 1.8423, 0.3647, 1.1600], + device='cuda:3'), covar=tensor([0.0338, 0.0388, 0.0296, 0.0456, 0.0807, 0.0533, 0.0860, 0.0698], + device='cuda:3'), in_proj_covar=tensor([0.0330, 0.0328, 0.0329, 0.0343, 0.0420, 0.0342, 0.0306, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 16:45:37,342 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63587.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:45:46,572 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 16:45:54,834 INFO [train.py:903] (3/4) Epoch 10, batch 2150, loss[loss=0.239, simple_loss=0.3159, pruned_loss=0.08108, over 19418.00 frames. ], tot_loss[loss=0.2392, simple_loss=0.3107, pruned_loss=0.0838, over 3843914.24 frames. ], batch size: 70, lr: 8.43e-03, grad_scale: 8.0 +2023-04-01 16:46:00,093 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63604.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:46:58,774 INFO [train.py:903] (3/4) Epoch 10, batch 2200, loss[loss=0.2037, simple_loss=0.2937, pruned_loss=0.05688, over 19696.00 frames. ], tot_loss[loss=0.2394, simple_loss=0.3109, pruned_loss=0.08396, over 3834025.58 frames. ], batch size: 59, lr: 8.42e-03, grad_scale: 8.0 +2023-04-01 16:47:13,909 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.200e+02 5.616e+02 6.875e+02 8.680e+02 1.983e+03, threshold=1.375e+03, percent-clipped=4.0 +2023-04-01 16:48:00,213 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63698.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:48:03,398 INFO [train.py:903] (3/4) Epoch 10, batch 2250, loss[loss=0.2211, simple_loss=0.2926, pruned_loss=0.07478, over 19485.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.3113, pruned_loss=0.08423, over 3818382.43 frames. ], batch size: 49, lr: 8.42e-03, grad_scale: 8.0 +2023-04-01 16:49:04,922 INFO [train.py:903] (3/4) Epoch 10, batch 2300, loss[loss=0.2126, simple_loss=0.2912, pruned_loss=0.06693, over 19824.00 frames. ], tot_loss[loss=0.2432, simple_loss=0.3138, pruned_loss=0.08627, over 3790429.29 frames. ], batch size: 52, lr: 8.42e-03, grad_scale: 4.0 +2023-04-01 16:49:18,654 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 16:49:23,048 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.591e+02 5.837e+02 6.965e+02 8.642e+02 2.205e+03, threshold=1.393e+03, percent-clipped=3.0 +2023-04-01 16:49:45,522 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63782.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:49:47,898 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63784.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:50:09,266 INFO [train.py:903] (3/4) Epoch 10, batch 2350, loss[loss=0.2352, simple_loss=0.311, pruned_loss=0.07964, over 19669.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3124, pruned_loss=0.08521, over 3802289.30 frames. ], batch size: 58, lr: 8.41e-03, grad_scale: 4.0 +2023-04-01 16:50:44,333 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63828.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:50:50,884 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 16:51:10,027 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 16:51:11,957 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-01 16:51:13,627 INFO [train.py:903] (3/4) Epoch 10, batch 2400, loss[loss=0.2432, simple_loss=0.3092, pruned_loss=0.08857, over 19405.00 frames. ], tot_loss[loss=0.2423, simple_loss=0.313, pruned_loss=0.08581, over 3797621.68 frames. ], batch size: 48, lr: 8.41e-03, grad_scale: 8.0 +2023-04-01 16:51:29,410 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63863.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:51:30,163 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.763e+02 5.915e+02 7.022e+02 9.244e+02 2.907e+03, threshold=1.404e+03, percent-clipped=10.0 +2023-04-01 16:51:55,363 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9366, 2.0079, 2.1239, 2.7547, 2.0452, 2.6806, 2.4216, 2.0566], + device='cuda:3'), covar=tensor([0.3160, 0.2569, 0.1281, 0.1495, 0.2838, 0.1185, 0.2841, 0.2241], + device='cuda:3'), in_proj_covar=tensor([0.0772, 0.0780, 0.0638, 0.0882, 0.0763, 0.0683, 0.0781, 0.0695], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 16:52:03,547 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.70 vs. limit=2.0 +2023-04-01 16:52:12,812 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63897.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:52:15,964 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63899.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:52:17,969 INFO [train.py:903] (3/4) Epoch 10, batch 2450, loss[loss=0.2785, simple_loss=0.335, pruned_loss=0.111, over 19583.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3127, pruned_loss=0.08506, over 3805940.64 frames. ], batch size: 52, lr: 8.41e-03, grad_scale: 8.0 +2023-04-01 16:52:55,569 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63931.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:52:59,305 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63934.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:53:16,602 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=63948.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:53:19,964 INFO [train.py:903] (3/4) Epoch 10, batch 2500, loss[loss=0.2509, simple_loss=0.3208, pruned_loss=0.09055, over 19702.00 frames. ], tot_loss[loss=0.2417, simple_loss=0.3129, pruned_loss=0.08525, over 3796536.47 frames. ], batch size: 59, lr: 8.40e-03, grad_scale: 8.0 +2023-04-01 16:53:24,612 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63954.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:53:35,919 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.566e+02 5.261e+02 6.826e+02 9.233e+02 1.687e+03, threshold=1.365e+03, percent-clipped=6.0 +2023-04-01 16:53:56,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63979.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:54:23,059 INFO [train.py:903] (3/4) Epoch 10, batch 2550, loss[loss=0.2785, simple_loss=0.3373, pruned_loss=0.1099, over 19757.00 frames. ], tot_loss[loss=0.2404, simple_loss=0.3117, pruned_loss=0.08458, over 3803813.37 frames. ], batch size: 63, lr: 8.40e-03, grad_scale: 8.0 +2023-04-01 16:55:13,554 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64041.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:55:15,684 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 16:55:20,883 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64046.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:55:26,452 INFO [train.py:903] (3/4) Epoch 10, batch 2600, loss[loss=0.216, simple_loss=0.2824, pruned_loss=0.07485, over 19723.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.312, pruned_loss=0.08469, over 3802134.81 frames. ], batch size: 46, lr: 8.40e-03, grad_scale: 8.0 +2023-04-01 16:55:41,464 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64063.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:55:42,246 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.625e+02 5.188e+02 6.411e+02 7.864e+02 1.888e+03, threshold=1.282e+03, percent-clipped=1.0 +2023-04-01 16:56:27,597 INFO [train.py:903] (3/4) Epoch 10, batch 2650, loss[loss=0.2355, simple_loss=0.2959, pruned_loss=0.08754, over 19474.00 frames. ], tot_loss[loss=0.241, simple_loss=0.312, pruned_loss=0.085, over 3816391.81 frames. ], batch size: 49, lr: 8.39e-03, grad_scale: 8.0 +2023-04-01 16:56:45,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 16:57:18,005 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3602, 1.7280, 1.6936, 2.6656, 2.1726, 2.4076, 2.5517, 2.4555], + device='cuda:3'), covar=tensor([0.0749, 0.1017, 0.1047, 0.0988, 0.0975, 0.0725, 0.0888, 0.0634], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0226, 0.0225, 0.0254, 0.0240, 0.0213, 0.0201, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 16:57:29,272 INFO [train.py:903] (3/4) Epoch 10, batch 2700, loss[loss=0.2668, simple_loss=0.3396, pruned_loss=0.09701, over 19687.00 frames. ], tot_loss[loss=0.2398, simple_loss=0.3113, pruned_loss=0.08417, over 3814869.46 frames. ], batch size: 60, lr: 8.39e-03, grad_scale: 8.0 +2023-04-01 16:57:29,582 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9555, 4.8986, 5.7826, 5.7797, 1.5667, 5.5118, 4.7050, 5.3594], + device='cuda:3'), covar=tensor([0.1285, 0.0861, 0.0512, 0.0428, 0.5830, 0.0469, 0.0514, 0.1070], + device='cuda:3'), in_proj_covar=tensor([0.0642, 0.0581, 0.0769, 0.0641, 0.0715, 0.0519, 0.0477, 0.0709], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 16:57:32,849 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64153.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:57:34,005 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0718, 1.7531, 2.1352, 1.7114, 4.6179, 0.8370, 2.4970, 4.8500], + device='cuda:3'), covar=tensor([0.0282, 0.2355, 0.2167, 0.1628, 0.0603, 0.2557, 0.1217, 0.0175], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0328, 0.0342, 0.0311, 0.0336, 0.0323, 0.0319, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:57:35,356 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64155.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:57:45,495 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.698e+02 5.891e+02 6.747e+02 8.779e+02 3.257e+03, threshold=1.349e+03, percent-clipped=11.0 +2023-04-01 16:57:56,914 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64172.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:58:05,001 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64178.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:58:07,365 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64180.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:58:33,083 INFO [train.py:903] (3/4) Epoch 10, batch 2750, loss[loss=0.1735, simple_loss=0.2533, pruned_loss=0.04685, over 19376.00 frames. ], tot_loss[loss=0.239, simple_loss=0.311, pruned_loss=0.08353, over 3806671.12 frames. ], batch size: 47, lr: 8.39e-03, grad_scale: 8.0 +2023-04-01 16:58:41,443 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64207.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:59:36,117 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 16:59:36,901 INFO [train.py:903] (3/4) Epoch 10, batch 2800, loss[loss=0.2516, simple_loss=0.3075, pruned_loss=0.09786, over 19395.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3119, pruned_loss=0.08433, over 3809854.39 frames. ], batch size: 48, lr: 8.38e-03, grad_scale: 8.0 +2023-04-01 16:59:43,002 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9745, 1.8285, 2.0362, 1.7819, 4.3799, 0.8504, 2.2267, 4.6298], + device='cuda:3'), covar=tensor([0.0285, 0.2350, 0.2280, 0.1675, 0.0694, 0.2669, 0.1486, 0.0211], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0327, 0.0342, 0.0313, 0.0339, 0.0326, 0.0319, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 16:59:45,217 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9215, 3.5072, 2.4794, 3.2200, 0.9534, 3.3242, 3.2838, 3.3899], + device='cuda:3'), covar=tensor([0.0773, 0.1194, 0.2018, 0.0808, 0.3866, 0.0894, 0.0848, 0.1103], + device='cuda:3'), in_proj_covar=tensor([0.0409, 0.0353, 0.0415, 0.0309, 0.0368, 0.0343, 0.0341, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 16:59:52,865 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.388e+02 5.243e+02 6.695e+02 7.923e+02 2.040e+03, threshold=1.339e+03, percent-clipped=2.0 +2023-04-01 17:00:10,660 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64278.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:00:21,841 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2186, 1.2500, 1.7317, 1.2213, 2.6763, 3.5516, 3.2817, 3.7764], + device='cuda:3'), covar=tensor([0.1481, 0.3256, 0.2804, 0.1977, 0.0453, 0.0162, 0.0192, 0.0159], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0291, 0.0320, 0.0249, 0.0211, 0.0149, 0.0203, 0.0190], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:00:21,853 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64287.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:00:40,327 INFO [train.py:903] (3/4) Epoch 10, batch 2850, loss[loss=0.2149, simple_loss=0.2836, pruned_loss=0.07306, over 19760.00 frames. ], tot_loss[loss=0.2386, simple_loss=0.3106, pruned_loss=0.08334, over 3816456.93 frames. ], batch size: 47, lr: 8.38e-03, grad_scale: 8.0 +2023-04-01 17:00:41,921 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:03,219 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64319.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:06,431 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64322.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:12,428 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64327.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:35,607 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64344.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:39,746 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 17:01:43,205 INFO [train.py:903] (3/4) Epoch 10, batch 2900, loss[loss=0.2345, simple_loss=0.3061, pruned_loss=0.08144, over 19663.00 frames. ], tot_loss[loss=0.2391, simple_loss=0.3109, pruned_loss=0.08362, over 3816084.06 frames. ], batch size: 53, lr: 8.38e-03, grad_scale: 8.0 +2023-04-01 17:01:58,188 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64363.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:01:58,958 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.152e+02 6.015e+02 7.349e+02 1.031e+03 2.008e+03, threshold=1.470e+03, percent-clipped=12.0 +2023-04-01 17:02:26,637 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64385.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:02:28,051 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2091, 1.9755, 1.4344, 1.3032, 1.8594, 1.1354, 1.1455, 1.6142], + device='cuda:3'), covar=tensor([0.0915, 0.0623, 0.0979, 0.0655, 0.0436, 0.1161, 0.0650, 0.0422], + device='cuda:3'), in_proj_covar=tensor([0.0288, 0.0296, 0.0323, 0.0242, 0.0231, 0.0322, 0.0288, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 17:02:36,423 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64393.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:02:44,649 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8439, 1.6913, 1.4812, 1.9644, 1.9470, 1.6699, 1.4989, 1.7856], + device='cuda:3'), covar=tensor([0.0876, 0.1434, 0.1363, 0.0910, 0.1064, 0.0529, 0.1188, 0.0630], + device='cuda:3'), in_proj_covar=tensor([0.0249, 0.0349, 0.0288, 0.0237, 0.0296, 0.0243, 0.0273, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 17:02:46,685 INFO [train.py:903] (3/4) Epoch 10, batch 2950, loss[loss=0.2402, simple_loss=0.3161, pruned_loss=0.08217, over 19852.00 frames. ], tot_loss[loss=0.2394, simple_loss=0.3116, pruned_loss=0.08361, over 3825141.12 frames. ], batch size: 52, lr: 8.37e-03, grad_scale: 8.0 +2023-04-01 17:03:48,250 INFO [train.py:903] (3/4) Epoch 10, batch 3000, loss[loss=0.2037, simple_loss=0.2882, pruned_loss=0.05955, over 19778.00 frames. ], tot_loss[loss=0.2377, simple_loss=0.3099, pruned_loss=0.08278, over 3834571.47 frames. ], batch size: 54, lr: 8.37e-03, grad_scale: 8.0 +2023-04-01 17:03:48,250 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 17:04:00,864 INFO [train.py:937] (3/4) Epoch 10, validation: loss=0.1811, simple_loss=0.2816, pruned_loss=0.04036, over 944034.00 frames. +2023-04-01 17:04:00,865 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 17:04:04,327 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 17:04:18,120 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.319e+02 5.533e+02 6.661e+02 8.174e+02 1.809e+03, threshold=1.332e+03, percent-clipped=2.0 +2023-04-01 17:04:30,052 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5630, 1.3565, 1.3523, 1.9602, 1.5246, 1.7385, 1.8318, 1.6509], + device='cuda:3'), covar=tensor([0.0848, 0.1078, 0.1052, 0.0776, 0.0821, 0.0834, 0.0922, 0.0738], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0226, 0.0224, 0.0252, 0.0237, 0.0215, 0.0201, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 17:04:41,339 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64483.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:05:03,239 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64500.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:05:04,026 INFO [train.py:903] (3/4) Epoch 10, batch 3050, loss[loss=0.1981, simple_loss=0.277, pruned_loss=0.05961, over 19736.00 frames. ], tot_loss[loss=0.2402, simple_loss=0.3122, pruned_loss=0.08411, over 3832564.59 frames. ], batch size: 51, lr: 8.37e-03, grad_scale: 8.0 +2023-04-01 17:05:57,723 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64543.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:06:06,594 INFO [train.py:903] (3/4) Epoch 10, batch 3100, loss[loss=0.2119, simple_loss=0.2772, pruned_loss=0.07334, over 19750.00 frames. ], tot_loss[loss=0.2411, simple_loss=0.313, pruned_loss=0.08459, over 3825125.67 frames. ], batch size: 45, lr: 8.37e-03, grad_scale: 8.0 +2023-04-01 17:06:22,864 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.367e+02 5.961e+02 7.216e+02 8.690e+02 2.208e+03, threshold=1.443e+03, percent-clipped=3.0 +2023-04-01 17:06:28,012 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64568.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:06:40,873 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64578.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:07:00,334 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5216, 1.2776, 1.7563, 1.5973, 2.9244, 4.8145, 4.6078, 5.0960], + device='cuda:3'), covar=tensor([0.1573, 0.4204, 0.3796, 0.1998, 0.0570, 0.0157, 0.0164, 0.0116], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0292, 0.0319, 0.0248, 0.0211, 0.0150, 0.0202, 0.0190], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:07:01,272 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64594.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:07:10,477 INFO [train.py:903] (3/4) Epoch 10, batch 3150, loss[loss=0.2546, simple_loss=0.3262, pruned_loss=0.09151, over 19699.00 frames. ], tot_loss[loss=0.2419, simple_loss=0.3138, pruned_loss=0.08498, over 3831302.20 frames. ], batch size: 60, lr: 8.36e-03, grad_scale: 8.0 +2023-04-01 17:07:13,023 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64603.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:07:17,543 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64607.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:07:39,761 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 17:07:57,640 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3009, 1.6695, 1.7239, 2.7245, 2.3070, 2.3263, 2.5686, 2.2778], + device='cuda:3'), covar=tensor([0.0758, 0.0983, 0.1021, 0.0831, 0.0767, 0.0768, 0.0801, 0.0659], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0226, 0.0225, 0.0253, 0.0239, 0.0216, 0.0200, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 17:08:10,679 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64649.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:08:12,598 INFO [train.py:903] (3/4) Epoch 10, batch 3200, loss[loss=0.28, simple_loss=0.34, pruned_loss=0.11, over 13333.00 frames. ], tot_loss[loss=0.2396, simple_loss=0.312, pruned_loss=0.08361, over 3830012.22 frames. ], batch size: 136, lr: 8.36e-03, grad_scale: 8.0 +2023-04-01 17:08:30,175 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.006e+02 5.483e+02 6.754e+02 8.204e+02 1.644e+03, threshold=1.351e+03, percent-clipped=2.0 +2023-04-01 17:08:32,822 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64666.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:08:43,788 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64674.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:09:02,728 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4714, 2.1050, 2.1364, 2.9784, 2.1429, 2.6723, 2.6343, 2.6815], + device='cuda:3'), covar=tensor([0.0683, 0.0807, 0.0823, 0.0694, 0.0875, 0.0642, 0.0776, 0.0538], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0224, 0.0223, 0.0251, 0.0236, 0.0213, 0.0199, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 17:09:16,024 INFO [train.py:903] (3/4) Epoch 10, batch 3250, loss[loss=0.2485, simple_loss=0.3217, pruned_loss=0.08766, over 18226.00 frames. ], tot_loss[loss=0.2386, simple_loss=0.3109, pruned_loss=0.08319, over 3841110.90 frames. ], batch size: 83, lr: 8.36e-03, grad_scale: 8.0 +2023-04-01 17:09:24,060 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64707.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:09:26,628 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64709.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:09:42,833 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64721.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:10:20,328 INFO [train.py:903] (3/4) Epoch 10, batch 3300, loss[loss=0.225, simple_loss=0.3028, pruned_loss=0.07363, over 19605.00 frames. ], tot_loss[loss=0.2389, simple_loss=0.3109, pruned_loss=0.08349, over 3839650.83 frames. ], batch size: 57, lr: 8.35e-03, grad_scale: 8.0 +2023-04-01 17:10:26,741 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64756.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:10:28,761 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 17:10:35,540 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64763.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:10:37,431 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.278e+02 5.384e+02 6.605e+02 8.281e+02 2.311e+03, threshold=1.321e+03, percent-clipped=9.0 +2023-04-01 17:10:43,939 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64770.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:10:57,711 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64781.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:11:23,421 INFO [train.py:903] (3/4) Epoch 10, batch 3350, loss[loss=0.2975, simple_loss=0.3579, pruned_loss=0.1186, over 19601.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.3115, pruned_loss=0.08416, over 3820532.13 frames. ], batch size: 57, lr: 8.35e-03, grad_scale: 4.0 +2023-04-01 17:11:49,545 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64822.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:11:55,096 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64827.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:12:24,011 INFO [train.py:903] (3/4) Epoch 10, batch 3400, loss[loss=0.2882, simple_loss=0.3476, pruned_loss=0.1144, over 19612.00 frames. ], tot_loss[loss=0.2404, simple_loss=0.3116, pruned_loss=0.08461, over 3829377.16 frames. ], batch size: 57, lr: 8.35e-03, grad_scale: 4.0 +2023-04-01 17:12:42,253 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.893e+02 5.724e+02 7.342e+02 8.665e+02 1.913e+03, threshold=1.468e+03, percent-clipped=6.0 +2023-04-01 17:13:27,816 INFO [train.py:903] (3/4) Epoch 10, batch 3450, loss[loss=0.2627, simple_loss=0.3339, pruned_loss=0.09579, over 19847.00 frames. ], tot_loss[loss=0.2397, simple_loss=0.3109, pruned_loss=0.08419, over 3833070.96 frames. ], batch size: 52, lr: 8.34e-03, grad_scale: 4.0 +2023-04-01 17:13:35,084 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 17:14:05,176 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2129, 2.2379, 2.2996, 3.3378, 2.3125, 3.2303, 2.8001, 2.2158], + device='cuda:3'), covar=tensor([0.3426, 0.2982, 0.1380, 0.1712, 0.3371, 0.1325, 0.3038, 0.2408], + device='cuda:3'), in_proj_covar=tensor([0.0769, 0.0779, 0.0633, 0.0879, 0.0762, 0.0682, 0.0776, 0.0691], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:14:20,591 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64942.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:14:30,655 INFO [train.py:903] (3/4) Epoch 10, batch 3500, loss[loss=0.2122, simple_loss=0.28, pruned_loss=0.07222, over 19297.00 frames. ], tot_loss[loss=0.2389, simple_loss=0.31, pruned_loss=0.08389, over 3825663.61 frames. ], batch size: 44, lr: 8.34e-03, grad_scale: 4.0 +2023-04-01 17:14:31,770 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=64951.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:14:48,653 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.495e+02 5.616e+02 6.822e+02 8.996e+02 1.764e+03, threshold=1.364e+03, percent-clipped=1.0 +2023-04-01 17:14:49,108 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64965.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:15:19,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64990.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:15:25,305 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64994.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:15:27,106 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-01 17:15:34,048 INFO [train.py:903] (3/4) Epoch 10, batch 3550, loss[loss=0.286, simple_loss=0.3441, pruned_loss=0.114, over 18248.00 frames. ], tot_loss[loss=0.2387, simple_loss=0.3099, pruned_loss=0.08376, over 3818583.03 frames. ], batch size: 83, lr: 8.34e-03, grad_scale: 4.0 +2023-04-01 17:15:44,363 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65010.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:16:16,119 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3335, 1.3606, 1.7454, 1.4293, 2.9897, 4.3788, 4.2895, 4.8623], + device='cuda:3'), covar=tensor([0.1521, 0.3327, 0.3104, 0.2034, 0.0511, 0.0154, 0.0172, 0.0121], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0290, 0.0318, 0.0246, 0.0211, 0.0149, 0.0202, 0.0190], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:16:34,667 INFO [train.py:903] (3/4) Epoch 10, batch 3600, loss[loss=0.3005, simple_loss=0.3591, pruned_loss=0.121, over 19780.00 frames. ], tot_loss[loss=0.2396, simple_loss=0.311, pruned_loss=0.08413, over 3821218.32 frames. ], batch size: 56, lr: 8.33e-03, grad_scale: 8.0 +2023-04-01 17:16:37,753 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-01 17:16:52,000 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.218e+02 5.791e+02 6.813e+02 8.568e+02 1.743e+03, threshold=1.363e+03, percent-clipped=4.0 +2023-04-01 17:16:52,169 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65065.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:16:53,539 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65066.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:17:02,295 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65073.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:17:08,952 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65078.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:17:20,310 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.72 vs. limit=5.0 +2023-04-01 17:17:35,955 INFO [train.py:903] (3/4) Epoch 10, batch 3650, loss[loss=0.286, simple_loss=0.3477, pruned_loss=0.1121, over 19664.00 frames. ], tot_loss[loss=0.24, simple_loss=0.3113, pruned_loss=0.08436, over 3819014.19 frames. ], batch size: 60, lr: 8.33e-03, grad_scale: 8.0 +2023-04-01 17:17:38,648 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65103.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:17:42,761 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65107.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:17:45,377 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2012, 3.6721, 2.0367, 2.2224, 3.2129, 1.8253, 1.4193, 2.2201], + device='cuda:3'), covar=tensor([0.1038, 0.0368, 0.1009, 0.0641, 0.0464, 0.1038, 0.0919, 0.0608], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0295, 0.0321, 0.0243, 0.0232, 0.0319, 0.0287, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 17:17:51,937 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65114.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:18:05,475 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65125.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:18:36,264 INFO [train.py:903] (3/4) Epoch 10, batch 3700, loss[loss=0.2146, simple_loss=0.2962, pruned_loss=0.0665, over 18073.00 frames. ], tot_loss[loss=0.2401, simple_loss=0.3117, pruned_loss=0.08431, over 3817807.05 frames. ], batch size: 83, lr: 8.33e-03, grad_scale: 8.0 +2023-04-01 17:18:53,968 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.395e+02 5.629e+02 7.088e+02 8.754e+02 1.818e+03, threshold=1.418e+03, percent-clipped=5.0 +2023-04-01 17:19:12,228 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65180.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:19:34,116 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65198.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:19:37,048 INFO [train.py:903] (3/4) Epoch 10, batch 3750, loss[loss=0.2463, simple_loss=0.3187, pruned_loss=0.0869, over 19781.00 frames. ], tot_loss[loss=0.2417, simple_loss=0.3128, pruned_loss=0.08531, over 3828983.62 frames. ], batch size: 56, lr: 8.32e-03, grad_scale: 8.0 +2023-04-01 17:20:02,515 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65222.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:20:03,636 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65223.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:20:10,553 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65229.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:20:36,852 INFO [train.py:903] (3/4) Epoch 10, batch 3800, loss[loss=0.2584, simple_loss=0.3321, pruned_loss=0.09235, over 19543.00 frames. ], tot_loss[loss=0.2423, simple_loss=0.3135, pruned_loss=0.08557, over 3819133.46 frames. ], batch size: 56, lr: 8.32e-03, grad_scale: 8.0 +2023-04-01 17:20:54,005 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.166e+02 5.620e+02 6.620e+02 8.005e+02 1.692e+03, threshold=1.324e+03, percent-clipped=4.0 +2023-04-01 17:21:06,343 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 17:21:37,286 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-01 17:21:37,759 INFO [train.py:903] (3/4) Epoch 10, batch 3850, loss[loss=0.2599, simple_loss=0.3253, pruned_loss=0.09725, over 19523.00 frames. ], tot_loss[loss=0.2418, simple_loss=0.3132, pruned_loss=0.08517, over 3831168.86 frames. ], batch size: 54, lr: 8.32e-03, grad_scale: 8.0 +2023-04-01 17:21:40,425 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65303.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:22:03,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65322.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:22:21,622 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65338.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:22:33,979 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65347.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:22:37,971 INFO [train.py:903] (3/4) Epoch 10, batch 3900, loss[loss=0.2556, simple_loss=0.3293, pruned_loss=0.09093, over 18749.00 frames. ], tot_loss[loss=0.24, simple_loss=0.3118, pruned_loss=0.0841, over 3839025.74 frames. ], batch size: 74, lr: 8.31e-03, grad_scale: 8.0 +2023-04-01 17:22:44,184 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 17:22:55,831 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.027e+02 5.653e+02 7.180e+02 9.093e+02 1.633e+03, threshold=1.436e+03, percent-clipped=2.0 +2023-04-01 17:23:15,720 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65381.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:23:40,379 INFO [train.py:903] (3/4) Epoch 10, batch 3950, loss[loss=0.2323, simple_loss=0.2903, pruned_loss=0.08718, over 19384.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3122, pruned_loss=0.08453, over 3833621.99 frames. ], batch size: 48, lr: 8.31e-03, grad_scale: 8.0 +2023-04-01 17:23:40,411 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 17:23:46,504 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65406.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:23:53,339 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4411, 1.2451, 1.3265, 1.8288, 1.4571, 1.6323, 1.7647, 1.4764], + device='cuda:3'), covar=tensor([0.0871, 0.0998, 0.1035, 0.0694, 0.0813, 0.0705, 0.0750, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0227, 0.0226, 0.0252, 0.0242, 0.0216, 0.0202, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 17:23:58,955 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65417.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:24:23,321 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65436.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:24:41,330 INFO [train.py:903] (3/4) Epoch 10, batch 4000, loss[loss=0.2179, simple_loss=0.2761, pruned_loss=0.07987, over 19354.00 frames. ], tot_loss[loss=0.2398, simple_loss=0.3112, pruned_loss=0.08423, over 3832929.77 frames. ], batch size: 47, lr: 8.31e-03, grad_scale: 8.0 +2023-04-01 17:24:44,060 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65453.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:24:54,595 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65461.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:24:58,711 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.016e+02 5.448e+02 7.265e+02 9.508e+02 1.942e+03, threshold=1.453e+03, percent-clipped=2.0 +2023-04-01 17:25:14,524 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65478.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:25:23,062 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 17:25:23,371 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65485.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:25:41,796 INFO [train.py:903] (3/4) Epoch 10, batch 4050, loss[loss=0.2443, simple_loss=0.3143, pruned_loss=0.08714, over 19581.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3116, pruned_loss=0.08446, over 3837434.52 frames. ], batch size: 52, lr: 8.30e-03, grad_scale: 8.0 +2023-04-01 17:25:45,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65503.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:25:53,336 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65510.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:26:19,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65532.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:26:40,723 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65550.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:26:41,427 INFO [train.py:903] (3/4) Epoch 10, batch 4100, loss[loss=0.2386, simple_loss=0.2978, pruned_loss=0.08967, over 19730.00 frames. ], tot_loss[loss=0.2402, simple_loss=0.3113, pruned_loss=0.08452, over 3823243.99 frames. ], batch size: 46, lr: 8.30e-03, grad_scale: 8.0 +2023-04-01 17:26:59,039 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 5.925e+02 7.032e+02 8.463e+02 2.911e+03, threshold=1.406e+03, percent-clipped=6.0 +2023-04-01 17:27:04,283 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3300, 2.3000, 2.4472, 3.3900, 2.2849, 3.2965, 2.8919, 2.2784], + device='cuda:3'), covar=tensor([0.3661, 0.3178, 0.1444, 0.1878, 0.3776, 0.1390, 0.3175, 0.2664], + device='cuda:3'), in_proj_covar=tensor([0.0776, 0.0789, 0.0641, 0.0884, 0.0771, 0.0691, 0.0787, 0.0700], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:27:11,840 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 17:27:23,829 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2483, 1.2333, 1.7463, 1.1457, 2.5720, 3.3382, 3.0649, 3.5837], + device='cuda:3'), covar=tensor([0.1408, 0.3316, 0.2803, 0.2018, 0.0426, 0.0171, 0.0213, 0.0167], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0291, 0.0317, 0.0245, 0.0212, 0.0149, 0.0203, 0.0191], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:27:42,542 INFO [train.py:903] (3/4) Epoch 10, batch 4150, loss[loss=0.271, simple_loss=0.3439, pruned_loss=0.09903, over 17980.00 frames. ], tot_loss[loss=0.2395, simple_loss=0.311, pruned_loss=0.08394, over 3830747.82 frames. ], batch size: 83, lr: 8.30e-03, grad_scale: 8.0 +2023-04-01 17:28:39,193 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65647.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:28:43,359 INFO [train.py:903] (3/4) Epoch 10, batch 4200, loss[loss=0.2829, simple_loss=0.3515, pruned_loss=0.1071, over 19583.00 frames. ], tot_loss[loss=0.2397, simple_loss=0.3113, pruned_loss=0.08405, over 3815583.34 frames. ], batch size: 61, lr: 8.30e-03, grad_scale: 8.0 +2023-04-01 17:28:43,399 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 17:28:59,406 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.075e+02 5.851e+02 6.725e+02 9.221e+02 2.199e+03, threshold=1.345e+03, percent-clipped=6.0 +2023-04-01 17:29:42,451 INFO [train.py:903] (3/4) Epoch 10, batch 4250, loss[loss=0.2481, simple_loss=0.3286, pruned_loss=0.08381, over 18663.00 frames. ], tot_loss[loss=0.2412, simple_loss=0.3126, pruned_loss=0.08487, over 3813502.09 frames. ], batch size: 74, lr: 8.29e-03, grad_scale: 8.0 +2023-04-01 17:29:54,057 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65709.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:29:55,743 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 17:30:06,598 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 17:30:22,866 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65734.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:30:40,307 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7154, 1.7966, 1.9860, 2.4168, 1.5719, 2.0764, 2.2477, 1.8812], + device='cuda:3'), covar=tensor([0.3208, 0.2709, 0.1367, 0.1561, 0.2965, 0.1479, 0.3136, 0.2440], + device='cuda:3'), in_proj_covar=tensor([0.0770, 0.0788, 0.0637, 0.0882, 0.0766, 0.0691, 0.0779, 0.0698], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:30:43,321 INFO [train.py:903] (3/4) Epoch 10, batch 4300, loss[loss=0.2573, simple_loss=0.3309, pruned_loss=0.09186, over 19217.00 frames. ], tot_loss[loss=0.2396, simple_loss=0.3111, pruned_loss=0.08408, over 3826286.32 frames. ], batch size: 69, lr: 8.29e-03, grad_scale: 8.0 +2023-04-01 17:30:56,174 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65762.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:31:00,064 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.946e+02 5.358e+02 7.223e+02 8.854e+02 2.636e+03, threshold=1.445e+03, percent-clipped=3.0 +2023-04-01 17:31:17,207 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-04-01 17:31:27,984 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65788.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:31:35,313 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 17:31:43,144 INFO [train.py:903] (3/4) Epoch 10, batch 4350, loss[loss=0.2692, simple_loss=0.3418, pruned_loss=0.09831, over 18750.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.3114, pruned_loss=0.08423, over 3829341.40 frames. ], batch size: 74, lr: 8.29e-03, grad_scale: 8.0 +2023-04-01 17:31:58,661 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65813.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:32:33,598 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.65 vs. limit=2.0 +2023-04-01 17:32:44,687 INFO [train.py:903] (3/4) Epoch 10, batch 4400, loss[loss=0.2553, simple_loss=0.322, pruned_loss=0.09426, over 19842.00 frames. ], tot_loss[loss=0.2401, simple_loss=0.3119, pruned_loss=0.08422, over 3834043.56 frames. ], batch size: 52, lr: 8.28e-03, grad_scale: 8.0 +2023-04-01 17:32:50,612 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65856.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:33:00,116 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.366e+02 5.748e+02 7.554e+02 9.760e+02 1.805e+03, threshold=1.511e+03, percent-clipped=4.0 +2023-04-01 17:33:10,942 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 17:33:19,854 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 17:33:36,075 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=65894.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:33:44,057 INFO [train.py:903] (3/4) Epoch 10, batch 4450, loss[loss=0.1955, simple_loss=0.2689, pruned_loss=0.06106, over 19743.00 frames. ], tot_loss[loss=0.241, simple_loss=0.3127, pruned_loss=0.08467, over 3837722.93 frames. ], batch size: 45, lr: 8.28e-03, grad_scale: 8.0 +2023-04-01 17:34:45,020 INFO [train.py:903] (3/4) Epoch 10, batch 4500, loss[loss=0.2002, simple_loss=0.2713, pruned_loss=0.06453, over 19759.00 frames. ], tot_loss[loss=0.2411, simple_loss=0.3127, pruned_loss=0.08473, over 3823113.18 frames. ], batch size: 45, lr: 8.28e-03, grad_scale: 8.0 +2023-04-01 17:35:01,306 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.423e+02 5.376e+02 6.626e+02 8.325e+02 1.832e+03, threshold=1.325e+03, percent-clipped=3.0 +2023-04-01 17:35:47,213 INFO [train.py:903] (3/4) Epoch 10, batch 4550, loss[loss=0.1899, simple_loss=0.2699, pruned_loss=0.05492, over 19747.00 frames. ], tot_loss[loss=0.2392, simple_loss=0.311, pruned_loss=0.08373, over 3811128.24 frames. ], batch size: 45, lr: 8.27e-03, grad_scale: 8.0 +2023-04-01 17:35:56,025 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 17:35:56,362 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66009.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:36:03,141 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-01 17:36:07,021 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66018.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:36:17,581 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 17:36:36,238 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66043.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:36:45,870 INFO [train.py:903] (3/4) Epoch 10, batch 4600, loss[loss=0.3028, simple_loss=0.3699, pruned_loss=0.1178, over 19603.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.3124, pruned_loss=0.08447, over 3809744.13 frames. ], batch size: 57, lr: 8.27e-03, grad_scale: 8.0 +2023-04-01 17:36:55,845 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8562, 1.6503, 1.6282, 2.0356, 1.5730, 2.2896, 2.0683, 2.0015], + device='cuda:3'), covar=tensor([0.0704, 0.0847, 0.0923, 0.0820, 0.0880, 0.0592, 0.0821, 0.0562], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0227, 0.0224, 0.0253, 0.0242, 0.0216, 0.0201, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 17:37:00,875 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.957e+02 5.689e+02 6.858e+02 9.462e+02 1.667e+03, threshold=1.372e+03, percent-clipped=8.0 +2023-04-01 17:37:43,790 INFO [train.py:903] (3/4) Epoch 10, batch 4650, loss[loss=0.2648, simple_loss=0.3315, pruned_loss=0.09912, over 17441.00 frames. ], tot_loss[loss=0.2414, simple_loss=0.3132, pruned_loss=0.08483, over 3806607.60 frames. ], batch size: 101, lr: 8.27e-03, grad_scale: 8.0 +2023-04-01 17:38:00,794 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 17:38:10,785 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 17:38:44,405 INFO [train.py:903] (3/4) Epoch 10, batch 4700, loss[loss=0.2415, simple_loss=0.3142, pruned_loss=0.08441, over 19749.00 frames. ], tot_loss[loss=0.2409, simple_loss=0.3126, pruned_loss=0.08463, over 3804424.27 frames. ], batch size: 51, lr: 8.26e-03, grad_scale: 8.0 +2023-04-01 17:38:57,803 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66162.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:39:00,970 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.007e+02 6.063e+02 7.892e+02 1.039e+03 2.104e+03, threshold=1.578e+03, percent-clipped=6.0 +2023-04-01 17:39:05,610 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 17:39:17,525 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66178.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:39:43,296 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=66200.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:39:44,102 INFO [train.py:903] (3/4) Epoch 10, batch 4750, loss[loss=0.2194, simple_loss=0.2865, pruned_loss=0.07614, over 19389.00 frames. ], tot_loss[loss=0.2421, simple_loss=0.3133, pruned_loss=0.08542, over 3805098.16 frames. ], batch size: 47, lr: 8.26e-03, grad_scale: 8.0 +2023-04-01 17:39:50,415 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2834, 1.4537, 1.8007, 1.5342, 2.6915, 2.2498, 2.8980, 1.1117], + device='cuda:3'), covar=tensor([0.2046, 0.3500, 0.2108, 0.1623, 0.1310, 0.1705, 0.1365, 0.3471], + device='cuda:3'), in_proj_covar=tensor([0.0478, 0.0563, 0.0586, 0.0430, 0.0587, 0.0486, 0.0646, 0.0483], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:39:53,516 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0995, 5.3824, 2.9395, 4.7142, 1.1171, 5.3476, 5.4278, 5.5264], + device='cuda:3'), covar=tensor([0.0358, 0.0836, 0.1727, 0.0609, 0.3713, 0.0571, 0.0555, 0.0818], + device='cuda:3'), in_proj_covar=tensor([0.0421, 0.0358, 0.0422, 0.0314, 0.0372, 0.0352, 0.0347, 0.0380], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 17:40:43,461 INFO [train.py:903] (3/4) Epoch 10, batch 4800, loss[loss=0.2871, simple_loss=0.3539, pruned_loss=0.1102, over 18855.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3125, pruned_loss=0.08422, over 3822311.41 frames. ], batch size: 74, lr: 8.26e-03, grad_scale: 8.0 +2023-04-01 17:41:01,217 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.573e+02 5.914e+02 7.080e+02 8.206e+02 1.527e+03, threshold=1.416e+03, percent-clipped=0.0 +2023-04-01 17:41:01,675 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66265.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:41:15,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66276.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:41:32,185 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66290.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:41:43,187 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-01 17:41:44,778 INFO [train.py:903] (3/4) Epoch 10, batch 4850, loss[loss=0.2294, simple_loss=0.3065, pruned_loss=0.07614, over 19529.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.312, pruned_loss=0.08395, over 3823173.54 frames. ], batch size: 56, lr: 8.25e-03, grad_scale: 8.0 +2023-04-01 17:42:02,638 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66315.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:42:10,354 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 17:42:30,026 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 17:42:35,766 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 17:42:35,792 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 17:42:45,285 INFO [train.py:903] (3/4) Epoch 10, batch 4900, loss[loss=0.2079, simple_loss=0.2798, pruned_loss=0.06795, over 19751.00 frames. ], tot_loss[loss=0.239, simple_loss=0.3111, pruned_loss=0.08344, over 3830335.53 frames. ], batch size: 46, lr: 8.25e-03, grad_scale: 8.0 +2023-04-01 17:42:45,302 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 17:43:01,824 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.050e+02 5.967e+02 7.269e+02 9.008e+02 2.184e+03, threshold=1.454e+03, percent-clipped=11.0 +2023-04-01 17:43:03,041 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 17:43:22,101 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5230, 2.1640, 1.6632, 1.5951, 2.0820, 1.2269, 1.4749, 1.8260], + device='cuda:3'), covar=tensor([0.0751, 0.0573, 0.0752, 0.0498, 0.0424, 0.0942, 0.0517, 0.0348], + device='cuda:3'), in_proj_covar=tensor([0.0284, 0.0294, 0.0321, 0.0240, 0.0233, 0.0314, 0.0285, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 17:43:45,943 INFO [train.py:903] (3/4) Epoch 10, batch 4950, loss[loss=0.2522, simple_loss=0.3251, pruned_loss=0.08963, over 19671.00 frames. ], tot_loss[loss=0.2392, simple_loss=0.3116, pruned_loss=0.08337, over 3815594.14 frames. ], batch size: 58, lr: 8.25e-03, grad_scale: 8.0 +2023-04-01 17:44:02,373 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 17:44:05,931 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9088, 2.0002, 2.1652, 2.7383, 1.8038, 2.6271, 2.4727, 2.0205], + device='cuda:3'), covar=tensor([0.3440, 0.2957, 0.1408, 0.1659, 0.3396, 0.1387, 0.3218, 0.2649], + device='cuda:3'), in_proj_covar=tensor([0.0771, 0.0786, 0.0637, 0.0884, 0.0763, 0.0688, 0.0777, 0.0698], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:44:26,188 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 17:44:44,954 INFO [train.py:903] (3/4) Epoch 10, batch 5000, loss[loss=0.2901, simple_loss=0.3571, pruned_loss=0.1115, over 19466.00 frames. ], tot_loss[loss=0.2404, simple_loss=0.3125, pruned_loss=0.08418, over 3816389.25 frames. ], batch size: 64, lr: 8.25e-03, grad_scale: 8.0 +2023-04-01 17:44:56,400 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 17:45:01,937 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.596e+02 5.419e+02 6.464e+02 8.305e+02 1.628e+03, threshold=1.293e+03, percent-clipped=3.0 +2023-04-01 17:45:06,368 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 17:45:45,182 INFO [train.py:903] (3/4) Epoch 10, batch 5050, loss[loss=0.2215, simple_loss=0.2888, pruned_loss=0.07712, over 19405.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3125, pruned_loss=0.08431, over 3820985.12 frames. ], batch size: 47, lr: 8.24e-03, grad_scale: 8.0 +2023-04-01 17:45:51,100 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=66506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:46:10,887 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=66522.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:46:19,336 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 17:46:31,393 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66539.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:46:44,533 INFO [train.py:903] (3/4) Epoch 10, batch 5100, loss[loss=0.2, simple_loss=0.2662, pruned_loss=0.06691, over 19730.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3125, pruned_loss=0.08437, over 3827858.37 frames. ], batch size: 46, lr: 8.24e-03, grad_scale: 8.0 +2023-04-01 17:46:57,232 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 17:46:59,522 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 17:47:01,862 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.567e+02 5.864e+02 7.142e+02 9.030e+02 2.803e+03, threshold=1.428e+03, percent-clipped=6.0 +2023-04-01 17:47:04,021 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 17:47:09,004 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66571.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:47:21,909 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9278, 1.9607, 2.0463, 2.7565, 1.8399, 2.5227, 2.4813, 2.0201], + device='cuda:3'), covar=tensor([0.3135, 0.2481, 0.1345, 0.1616, 0.2933, 0.1258, 0.2912, 0.2395], + device='cuda:3'), in_proj_covar=tensor([0.0776, 0.0787, 0.0641, 0.0887, 0.0767, 0.0691, 0.0778, 0.0702], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:47:26,367 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-01 17:47:39,566 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66596.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:47:45,735 INFO [train.py:903] (3/4) Epoch 10, batch 5150, loss[loss=0.2489, simple_loss=0.3217, pruned_loss=0.08807, over 19515.00 frames. ], tot_loss[loss=0.2409, simple_loss=0.313, pruned_loss=0.08441, over 3827942.63 frames. ], batch size: 64, lr: 8.24e-03, grad_scale: 8.0 +2023-04-01 17:47:56,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 17:48:09,060 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=66620.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:48:10,434 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66621.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:48:29,350 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66637.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:48:33,113 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 17:48:45,642 INFO [train.py:903] (3/4) Epoch 10, batch 5200, loss[loss=0.2659, simple_loss=0.3337, pruned_loss=0.09901, over 18451.00 frames. ], tot_loss[loss=0.2401, simple_loss=0.3121, pruned_loss=0.0841, over 3824336.81 frames. ], batch size: 84, lr: 8.23e-03, grad_scale: 8.0 +2023-04-01 17:49:00,693 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 17:49:02,887 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.107e+02 5.688e+02 7.032e+02 8.430e+02 1.656e+03, threshold=1.406e+03, percent-clipped=2.0 +2023-04-01 17:49:44,615 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 17:49:46,898 INFO [train.py:903] (3/4) Epoch 10, batch 5250, loss[loss=0.2056, simple_loss=0.2833, pruned_loss=0.06393, over 19473.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.3122, pruned_loss=0.08456, over 3827508.41 frames. ], batch size: 49, lr: 8.23e-03, grad_scale: 8.0 +2023-04-01 17:50:28,115 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66735.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:50:45,010 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66749.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:50:47,102 INFO [train.py:903] (3/4) Epoch 10, batch 5300, loss[loss=0.2452, simple_loss=0.3248, pruned_loss=0.08274, over 19526.00 frames. ], tot_loss[loss=0.242, simple_loss=0.3134, pruned_loss=0.08528, over 3830063.35 frames. ], batch size: 54, lr: 8.23e-03, grad_scale: 8.0 +2023-04-01 17:51:03,224 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 17:51:04,359 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.476e+02 5.985e+02 7.858e+02 1.069e+03 1.957e+03, threshold=1.572e+03, percent-clipped=7.0 +2023-04-01 17:51:37,422 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4763, 1.1560, 1.4880, 1.2830, 2.6778, 3.6430, 3.3479, 3.8593], + device='cuda:3'), covar=tensor([0.1543, 0.4532, 0.4043, 0.2204, 0.0562, 0.0188, 0.0264, 0.0201], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0290, 0.0317, 0.0246, 0.0209, 0.0150, 0.0202, 0.0193], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:51:47,695 INFO [train.py:903] (3/4) Epoch 10, batch 5350, loss[loss=0.2683, simple_loss=0.3351, pruned_loss=0.1008, over 19361.00 frames. ], tot_loss[loss=0.2421, simple_loss=0.3136, pruned_loss=0.08533, over 3822725.62 frames. ], batch size: 70, lr: 8.22e-03, grad_scale: 8.0 +2023-04-01 17:52:19,370 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 17:52:46,938 INFO [train.py:903] (3/4) Epoch 10, batch 5400, loss[loss=0.226, simple_loss=0.2927, pruned_loss=0.07965, over 19415.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.3123, pruned_loss=0.08444, over 3821649.92 frames. ], batch size: 48, lr: 8.22e-03, grad_scale: 8.0 +2023-04-01 17:53:05,681 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.674e+02 5.633e+02 6.871e+02 8.512e+02 1.525e+03, threshold=1.374e+03, percent-clipped=0.0 +2023-04-01 17:53:18,042 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66877.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:53:26,402 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=66883.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:53:37,747 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66893.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:53:47,838 INFO [train.py:903] (3/4) Epoch 10, batch 5450, loss[loss=0.2667, simple_loss=0.3328, pruned_loss=0.1003, over 13266.00 frames. ], tot_loss[loss=0.242, simple_loss=0.3133, pruned_loss=0.08537, over 3815621.90 frames. ], batch size: 135, lr: 8.22e-03, grad_scale: 8.0 +2023-04-01 17:53:49,487 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66902.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:54:08,173 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66918.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:54:47,612 INFO [train.py:903] (3/4) Epoch 10, batch 5500, loss[loss=0.2758, simple_loss=0.3396, pruned_loss=0.106, over 19544.00 frames. ], tot_loss[loss=0.2417, simple_loss=0.3128, pruned_loss=0.08527, over 3814100.40 frames. ], batch size: 56, lr: 8.21e-03, grad_scale: 8.0 +2023-04-01 17:55:06,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.718e+02 5.851e+02 7.428e+02 9.674e+02 2.066e+03, threshold=1.486e+03, percent-clipped=6.0 +2023-04-01 17:55:10,681 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 17:55:16,769 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66975.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:55:36,728 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66991.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:55:44,724 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66998.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 17:55:48,016 INFO [train.py:903] (3/4) Epoch 10, batch 5550, loss[loss=0.2108, simple_loss=0.294, pruned_loss=0.0638, over 19782.00 frames. ], tot_loss[loss=0.2408, simple_loss=0.3121, pruned_loss=0.08475, over 3812744.08 frames. ], batch size: 54, lr: 8.21e-03, grad_scale: 8.0 +2023-04-01 17:55:54,311 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 17:56:06,556 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67016.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:56:20,719 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0973, 2.0569, 1.7437, 1.6076, 1.2653, 1.5274, 0.4906, 1.0253], + device='cuda:3'), covar=tensor([0.0680, 0.0594, 0.0455, 0.0672, 0.1298, 0.0845, 0.1064, 0.1008], + device='cuda:3'), in_proj_covar=tensor([0.0330, 0.0328, 0.0329, 0.0349, 0.0418, 0.0348, 0.0305, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:56:42,313 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 17:56:47,796 INFO [train.py:903] (3/4) Epoch 10, batch 5600, loss[loss=0.2417, simple_loss=0.317, pruned_loss=0.08317, over 19790.00 frames. ], tot_loss[loss=0.2404, simple_loss=0.3116, pruned_loss=0.08458, over 3816505.80 frames. ], batch size: 56, lr: 8.21e-03, grad_scale: 8.0 +2023-04-01 17:57:06,482 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.953e+02 5.918e+02 7.804e+02 1.015e+03 2.269e+03, threshold=1.561e+03, percent-clipped=7.0 +2023-04-01 17:57:38,372 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=67093.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:57:47,808 INFO [train.py:903] (3/4) Epoch 10, batch 5650, loss[loss=0.2452, simple_loss=0.3191, pruned_loss=0.08567, over 19787.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3114, pruned_loss=0.0846, over 3808301.53 frames. ], batch size: 56, lr: 8.21e-03, grad_scale: 8.0 +2023-04-01 17:57:48,108 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67101.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 17:58:33,180 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 17:58:47,423 INFO [train.py:903] (3/4) Epoch 10, batch 5700, loss[loss=0.2632, simple_loss=0.3343, pruned_loss=0.09603, over 19299.00 frames. ], tot_loss[loss=0.2407, simple_loss=0.3117, pruned_loss=0.08483, over 3810396.53 frames. ], batch size: 66, lr: 8.20e-03, grad_scale: 8.0 +2023-04-01 17:59:05,253 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.236e+02 6.145e+02 7.592e+02 1.064e+03 2.520e+03, threshold=1.518e+03, percent-clipped=7.0 +2023-04-01 17:59:06,732 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3487, 1.2143, 1.6414, 1.3064, 2.8882, 3.8086, 3.5137, 3.9370], + device='cuda:3'), covar=tensor([0.1393, 0.3278, 0.2828, 0.1915, 0.0413, 0.0143, 0.0177, 0.0180], + device='cuda:3'), in_proj_covar=tensor([0.0251, 0.0290, 0.0313, 0.0245, 0.0206, 0.0150, 0.0202, 0.0193], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 17:59:47,166 INFO [train.py:903] (3/4) Epoch 10, batch 5750, loss[loss=0.3161, simple_loss=0.3649, pruned_loss=0.1337, over 13235.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3116, pruned_loss=0.0847, over 3809264.69 frames. ], batch size: 136, lr: 8.20e-03, grad_scale: 8.0 +2023-04-01 17:59:48,358 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 17:59:48,859 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8118, 1.8682, 2.0120, 2.6075, 1.8493, 2.4575, 2.3441, 1.9208], + device='cuda:3'), covar=tensor([0.3471, 0.2872, 0.1382, 0.1603, 0.3007, 0.1338, 0.3283, 0.2555], + device='cuda:3'), in_proj_covar=tensor([0.0773, 0.0785, 0.0639, 0.0885, 0.0762, 0.0690, 0.0773, 0.0697], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 17:59:57,164 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8004, 1.2965, 1.3812, 1.7232, 3.3825, 1.0213, 2.3102, 3.6022], + device='cuda:3'), covar=tensor([0.0406, 0.2680, 0.2926, 0.1656, 0.0700, 0.2628, 0.1265, 0.0302], + device='cuda:3'), in_proj_covar=tensor([0.0348, 0.0334, 0.0348, 0.0315, 0.0344, 0.0329, 0.0320, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 17:59:57,209 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67208.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 17:59:57,978 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 18:00:02,613 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 18:00:47,882 INFO [train.py:903] (3/4) Epoch 10, batch 5800, loss[loss=0.1823, simple_loss=0.2615, pruned_loss=0.05159, over 19602.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3114, pruned_loss=0.08457, over 3812113.53 frames. ], batch size: 50, lr: 8.20e-03, grad_scale: 8.0 +2023-04-01 18:00:51,541 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67254.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:01:06,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.662e+02 5.628e+02 6.923e+02 8.923e+02 2.275e+03, threshold=1.385e+03, percent-clipped=6.0 +2023-04-01 18:01:20,739 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67279.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:01:47,314 INFO [train.py:903] (3/4) Epoch 10, batch 5850, loss[loss=0.2581, simple_loss=0.3246, pruned_loss=0.09582, over 19773.00 frames. ], tot_loss[loss=0.2409, simple_loss=0.3115, pruned_loss=0.08516, over 3805738.74 frames. ], batch size: 56, lr: 8.19e-03, grad_scale: 8.0 +2023-04-01 18:01:50,841 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6346, 1.3038, 1.3702, 2.0140, 1.6071, 2.0246, 2.0009, 1.7870], + device='cuda:3'), covar=tensor([0.0762, 0.0981, 0.1045, 0.0839, 0.0805, 0.0647, 0.0808, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0226, 0.0225, 0.0251, 0.0238, 0.0214, 0.0198, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 18:02:09,370 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=67319.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:02:10,106 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.92 vs. limit=5.0 +2023-04-01 18:02:48,429 INFO [train.py:903] (3/4) Epoch 10, batch 5900, loss[loss=0.2272, simple_loss=0.3073, pruned_loss=0.07359, over 19531.00 frames. ], tot_loss[loss=0.2411, simple_loss=0.3121, pruned_loss=0.08502, over 3807952.61 frames. ], batch size: 54, lr: 8.19e-03, grad_scale: 8.0 +2023-04-01 18:02:52,959 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 18:03:05,151 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.548e+02 5.688e+02 6.588e+02 8.588e+02 1.646e+03, threshold=1.318e+03, percent-clipped=2.0 +2023-04-01 18:03:11,567 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 18:03:47,174 INFO [train.py:903] (3/4) Epoch 10, batch 5950, loss[loss=0.2424, simple_loss=0.3211, pruned_loss=0.08179, over 18640.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3118, pruned_loss=0.0846, over 3814740.35 frames. ], batch size: 74, lr: 8.19e-03, grad_scale: 8.0 +2023-04-01 18:04:02,524 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2402, 1.2706, 1.4386, 1.3231, 2.7647, 0.9943, 2.1188, 3.0752], + device='cuda:3'), covar=tensor([0.0581, 0.2616, 0.2671, 0.1954, 0.0874, 0.2542, 0.1272, 0.0377], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0333, 0.0350, 0.0316, 0.0345, 0.0331, 0.0324, 0.0345], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:04:26,278 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67434.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:04:38,311 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=67445.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:04:45,818 INFO [train.py:903] (3/4) Epoch 10, batch 6000, loss[loss=0.2106, simple_loss=0.2794, pruned_loss=0.07093, over 19772.00 frames. ], tot_loss[loss=0.2389, simple_loss=0.3102, pruned_loss=0.08379, over 3828286.99 frames. ], batch size: 47, lr: 8.18e-03, grad_scale: 8.0 +2023-04-01 18:04:45,818 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 18:04:58,247 INFO [train.py:937] (3/4) Epoch 10, validation: loss=0.1798, simple_loss=0.2805, pruned_loss=0.03952, over 944034.00 frames. +2023-04-01 18:04:58,248 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 18:05:15,107 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67464.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:05:17,959 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.621e+02 5.414e+02 6.867e+02 8.657e+02 1.897e+03, threshold=1.373e+03, percent-clipped=4.0 +2023-04-01 18:05:44,654 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67489.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:05:59,558 INFO [train.py:903] (3/4) Epoch 10, batch 6050, loss[loss=0.2328, simple_loss=0.315, pruned_loss=0.0753, over 19545.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.3109, pruned_loss=0.08445, over 3832445.15 frames. ], batch size: 56, lr: 8.18e-03, grad_scale: 8.0 +2023-04-01 18:06:35,116 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67531.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:06:59,164 INFO [train.py:903] (3/4) Epoch 10, batch 6100, loss[loss=0.2497, simple_loss=0.3248, pruned_loss=0.08731, over 19785.00 frames. ], tot_loss[loss=0.241, simple_loss=0.3119, pruned_loss=0.08508, over 3804006.84 frames. ], batch size: 56, lr: 8.18e-03, grad_scale: 4.0 +2023-04-01 18:07:10,573 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67560.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:07:19,955 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.598e+02 5.925e+02 6.739e+02 8.410e+02 2.337e+03, threshold=1.348e+03, percent-clipped=5.0 +2023-04-01 18:07:59,366 INFO [train.py:903] (3/4) Epoch 10, batch 6150, loss[loss=0.2166, simple_loss=0.2811, pruned_loss=0.07602, over 19762.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.311, pruned_loss=0.08481, over 3803528.95 frames. ], batch size: 46, lr: 8.18e-03, grad_scale: 4.0 +2023-04-01 18:08:28,207 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 18:08:33,638 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-01 18:08:36,627 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0293, 4.4478, 4.7645, 4.7276, 1.6000, 4.3901, 3.8424, 4.3737], + device='cuda:3'), covar=tensor([0.1222, 0.0687, 0.0496, 0.0485, 0.5297, 0.0635, 0.0564, 0.0983], + device='cuda:3'), in_proj_covar=tensor([0.0650, 0.0583, 0.0767, 0.0649, 0.0708, 0.0531, 0.0474, 0.0708], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 18:08:54,458 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5556, 3.0818, 2.0566, 2.0589, 2.1759, 2.5027, 1.0102, 2.2084], + device='cuda:3'), covar=tensor([0.0378, 0.0440, 0.0535, 0.0829, 0.0739, 0.0805, 0.0947, 0.0771], + device='cuda:3'), in_proj_covar=tensor([0.0331, 0.0329, 0.0326, 0.0348, 0.0420, 0.0346, 0.0306, 0.0321], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 18:08:59,752 INFO [train.py:903] (3/4) Epoch 10, batch 6200, loss[loss=0.2004, simple_loss=0.2803, pruned_loss=0.0603, over 19682.00 frames. ], tot_loss[loss=0.2384, simple_loss=0.3094, pruned_loss=0.08375, over 3806256.14 frames. ], batch size: 53, lr: 8.17e-03, grad_scale: 4.0 +2023-04-01 18:09:20,096 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.127e+02 5.059e+02 6.776e+02 9.553e+02 2.024e+03, threshold=1.355e+03, percent-clipped=7.0 +2023-04-01 18:09:46,814 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67690.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:09:50,146 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3033, 1.5874, 2.0437, 1.5673, 3.1113, 2.5257, 3.4884, 1.4576], + device='cuda:3'), covar=tensor([0.2256, 0.3750, 0.2208, 0.1738, 0.1435, 0.1845, 0.1477, 0.3604], + device='cuda:3'), in_proj_covar=tensor([0.0480, 0.0566, 0.0588, 0.0431, 0.0589, 0.0486, 0.0645, 0.0487], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:09:56,468 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67698.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:09:59,568 INFO [train.py:903] (3/4) Epoch 10, batch 6250, loss[loss=0.2457, simple_loss=0.3129, pruned_loss=0.08922, over 19470.00 frames. ], tot_loss[loss=0.2392, simple_loss=0.3103, pruned_loss=0.08403, over 3810383.29 frames. ], batch size: 64, lr: 8.17e-03, grad_scale: 4.0 +2023-04-01 18:10:09,918 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67710.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:10:15,785 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67715.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:10:30,842 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 18:10:59,326 INFO [train.py:903] (3/4) Epoch 10, batch 6300, loss[loss=0.2972, simple_loss=0.3502, pruned_loss=0.1221, over 13386.00 frames. ], tot_loss[loss=0.2386, simple_loss=0.3101, pruned_loss=0.08358, over 3816907.00 frames. ], batch size: 135, lr: 8.17e-03, grad_scale: 4.0 +2023-04-01 18:11:19,321 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.616e+02 5.371e+02 6.798e+02 8.698e+02 1.915e+03, threshold=1.360e+03, percent-clipped=7.0 +2023-04-01 18:11:58,678 INFO [train.py:903] (3/4) Epoch 10, batch 6350, loss[loss=0.2733, simple_loss=0.346, pruned_loss=0.1003, over 19763.00 frames. ], tot_loss[loss=0.2392, simple_loss=0.3108, pruned_loss=0.08379, over 3824066.18 frames. ], batch size: 54, lr: 8.16e-03, grad_scale: 4.0 +2023-04-01 18:12:15,097 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.59 vs. limit=5.0 +2023-04-01 18:12:16,884 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67816.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:12:47,264 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:12:58,576 INFO [train.py:903] (3/4) Epoch 10, batch 6400, loss[loss=0.2284, simple_loss=0.2866, pruned_loss=0.08508, over 19760.00 frames. ], tot_loss[loss=0.2397, simple_loss=0.3117, pruned_loss=0.08384, over 3829463.17 frames. ], batch size: 47, lr: 8.16e-03, grad_scale: 8.0 +2023-04-01 18:13:07,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5707, 1.6570, 1.7803, 2.0858, 1.4221, 1.8029, 2.0049, 1.7001], + device='cuda:3'), covar=tensor([0.3206, 0.2621, 0.1405, 0.1564, 0.2860, 0.1385, 0.3301, 0.2411], + device='cuda:3'), in_proj_covar=tensor([0.0775, 0.0790, 0.0641, 0.0884, 0.0768, 0.0690, 0.0779, 0.0700], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:13:08,903 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.46 vs. limit=5.0 +2023-04-01 18:13:18,757 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.354e+02 5.283e+02 6.601e+02 9.298e+02 1.582e+03, threshold=1.320e+03, percent-clipped=4.0 +2023-04-01 18:13:26,981 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=67875.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:13:59,328 INFO [train.py:903] (3/4) Epoch 10, batch 6450, loss[loss=0.2831, simple_loss=0.3435, pruned_loss=0.1114, over 19573.00 frames. ], tot_loss[loss=0.2387, simple_loss=0.3111, pruned_loss=0.08318, over 3827391.54 frames. ], batch size: 61, lr: 8.16e-03, grad_scale: 8.0 +2023-04-01 18:14:12,165 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8547, 1.4506, 1.7985, 1.7020, 4.2429, 1.0818, 2.2173, 4.4128], + device='cuda:3'), covar=tensor([0.0364, 0.2879, 0.2565, 0.1826, 0.0774, 0.2757, 0.1486, 0.0249], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0333, 0.0346, 0.0313, 0.0342, 0.0329, 0.0323, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:14:42,675 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 18:14:59,310 INFO [train.py:903] (3/4) Epoch 10, batch 6500, loss[loss=0.229, simple_loss=0.3111, pruned_loss=0.07347, over 19673.00 frames. ], tot_loss[loss=0.2378, simple_loss=0.3103, pruned_loss=0.08267, over 3826539.54 frames. ], batch size: 59, lr: 8.15e-03, grad_scale: 8.0 +2023-04-01 18:15:05,079 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 18:15:19,446 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.752e+02 5.758e+02 7.165e+02 9.309e+02 2.290e+03, threshold=1.433e+03, percent-clipped=7.0 +2023-04-01 18:15:29,997 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7104, 1.5259, 1.4633, 2.1770, 1.5540, 1.9822, 2.0767, 1.9050], + device='cuda:3'), covar=tensor([0.0762, 0.0913, 0.1018, 0.0717, 0.0887, 0.0674, 0.0783, 0.0613], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0228, 0.0226, 0.0253, 0.0238, 0.0214, 0.0198, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 18:15:47,574 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67990.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:15:58,447 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67999.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:16:01,748 INFO [train.py:903] (3/4) Epoch 10, batch 6550, loss[loss=0.2568, simple_loss=0.3284, pruned_loss=0.09259, over 18277.00 frames. ], tot_loss[loss=0.2387, simple_loss=0.311, pruned_loss=0.08315, over 3826542.71 frames. ], batch size: 84, lr: 8.15e-03, grad_scale: 8.0 +2023-04-01 18:16:03,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7422, 1.5831, 1.3578, 1.7636, 1.5787, 1.3322, 1.3025, 1.6456], + device='cuda:3'), covar=tensor([0.1017, 0.1413, 0.1557, 0.0989, 0.1258, 0.0773, 0.1490, 0.0773], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0355, 0.0293, 0.0242, 0.0297, 0.0245, 0.0279, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:16:51,313 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68042.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:16:51,812 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-01 18:17:02,074 INFO [train.py:903] (3/4) Epoch 10, batch 6600, loss[loss=0.2319, simple_loss=0.3127, pruned_loss=0.07554, over 19541.00 frames. ], tot_loss[loss=0.2382, simple_loss=0.3107, pruned_loss=0.0829, over 3818701.78 frames. ], batch size: 54, lr: 8.15e-03, grad_scale: 8.0 +2023-04-01 18:17:05,670 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68054.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:17:23,324 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.670e+02 5.913e+02 7.324e+02 8.709e+02 1.479e+03, threshold=1.465e+03, percent-clipped=1.0 +2023-04-01 18:17:46,987 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2845, 1.2666, 1.4144, 1.3710, 1.7812, 1.8619, 1.8961, 0.6165], + device='cuda:3'), covar=tensor([0.2076, 0.3700, 0.2179, 0.1684, 0.1330, 0.1900, 0.1114, 0.3624], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0569, 0.0590, 0.0432, 0.0590, 0.0488, 0.0646, 0.0488], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:18:02,823 INFO [train.py:903] (3/4) Epoch 10, batch 6650, loss[loss=0.2434, simple_loss=0.3224, pruned_loss=0.08219, over 19684.00 frames. ], tot_loss[loss=0.2384, simple_loss=0.3105, pruned_loss=0.08308, over 3813939.42 frames. ], batch size: 59, lr: 8.15e-03, grad_scale: 8.0 +2023-04-01 18:18:07,759 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68105.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:19:02,660 INFO [train.py:903] (3/4) Epoch 10, batch 6700, loss[loss=0.2263, simple_loss=0.3082, pruned_loss=0.07218, over 19660.00 frames. ], tot_loss[loss=0.2382, simple_loss=0.3102, pruned_loss=0.08308, over 3820506.68 frames. ], batch size: 58, lr: 8.14e-03, grad_scale: 8.0 +2023-04-01 18:19:09,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68157.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:19:22,651 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.762e+02 5.919e+02 7.224e+02 9.287e+02 2.274e+03, threshold=1.445e+03, percent-clipped=4.0 +2023-04-01 18:19:24,180 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68169.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:20:00,034 INFO [train.py:903] (3/4) Epoch 10, batch 6750, loss[loss=0.2008, simple_loss=0.2733, pruned_loss=0.06415, over 19062.00 frames. ], tot_loss[loss=0.2377, simple_loss=0.3096, pruned_loss=0.08285, over 3820499.66 frames. ], batch size: 42, lr: 8.14e-03, grad_scale: 8.0 +2023-04-01 18:20:07,079 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.62 vs. limit=2.0 +2023-04-01 18:20:17,458 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3328, 3.9335, 2.5292, 3.4760, 1.1964, 3.6293, 3.7451, 3.7729], + device='cuda:3'), covar=tensor([0.0658, 0.0997, 0.1857, 0.0815, 0.3415, 0.0855, 0.0775, 0.0963], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0361, 0.0425, 0.0316, 0.0376, 0.0357, 0.0348, 0.0379], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 18:20:51,854 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68246.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:20:56,748 INFO [train.py:903] (3/4) Epoch 10, batch 6800, loss[loss=0.1819, simple_loss=0.2567, pruned_loss=0.05355, over 19765.00 frames. ], tot_loss[loss=0.2379, simple_loss=0.3098, pruned_loss=0.08301, over 3822664.87 frames. ], batch size: 47, lr: 8.14e-03, grad_scale: 8.0 +2023-04-01 18:21:15,043 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.210e+02 6.150e+02 7.610e+02 9.082e+02 1.904e+03, threshold=1.522e+03, percent-clipped=4.0 +2023-04-01 18:21:18,283 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68271.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:21:41,179 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 18:21:41,644 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 18:21:44,653 INFO [train.py:903] (3/4) Epoch 11, batch 0, loss[loss=0.1989, simple_loss=0.2841, pruned_loss=0.05682, over 19666.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2841, pruned_loss=0.05682, over 19666.00 frames. ], batch size: 55, lr: 7.77e-03, grad_scale: 8.0 +2023-04-01 18:21:44,654 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 18:21:56,757 INFO [train.py:937] (3/4) Epoch 11, validation: loss=0.181, simple_loss=0.2818, pruned_loss=0.04012, over 944034.00 frames. +2023-04-01 18:21:56,758 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 18:22:09,357 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 18:22:22,222 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-04-01 18:22:36,682 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8389, 1.6550, 1.5265, 1.8193, 1.7013, 1.6731, 1.4961, 1.7835], + device='cuda:3'), covar=tensor([0.0916, 0.1416, 0.1333, 0.0973, 0.1156, 0.0520, 0.1188, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0355, 0.0290, 0.0240, 0.0297, 0.0246, 0.0277, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:22:57,941 INFO [train.py:903] (3/4) Epoch 11, batch 50, loss[loss=0.1856, simple_loss=0.2684, pruned_loss=0.05139, over 19774.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.3086, pruned_loss=0.08103, over 861265.24 frames. ], batch size: 46, lr: 7.76e-03, grad_scale: 8.0 +2023-04-01 18:23:15,262 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68343.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:23:35,567 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 18:23:46,885 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.915e+02 5.747e+02 7.027e+02 9.557e+02 1.564e+03, threshold=1.405e+03, percent-clipped=1.0 +2023-04-01 18:24:00,207 INFO [train.py:903] (3/4) Epoch 11, batch 100, loss[loss=0.1995, simple_loss=0.2703, pruned_loss=0.06433, over 19818.00 frames. ], tot_loss[loss=0.2303, simple_loss=0.3037, pruned_loss=0.07843, over 1522644.01 frames. ], batch size: 49, lr: 7.76e-03, grad_scale: 8.0 +2023-04-01 18:24:13,685 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 18:24:32,241 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5634, 1.6318, 1.8319, 2.1054, 1.3490, 1.8169, 2.0231, 1.7184], + device='cuda:3'), covar=tensor([0.3218, 0.2704, 0.1421, 0.1600, 0.2914, 0.1476, 0.3505, 0.2569], + device='cuda:3'), in_proj_covar=tensor([0.0773, 0.0789, 0.0641, 0.0880, 0.0771, 0.0695, 0.0779, 0.0696], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:24:42,555 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68413.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:24:47,972 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68417.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:24:57,054 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68425.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:25:01,044 INFO [train.py:903] (3/4) Epoch 11, batch 150, loss[loss=0.2297, simple_loss=0.3009, pruned_loss=0.07928, over 19610.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3082, pruned_loss=0.08145, over 2048022.82 frames. ], batch size: 50, lr: 7.76e-03, grad_scale: 8.0 +2023-04-01 18:25:11,956 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68438.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:25:25,314 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68449.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:25:26,576 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68450.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:25:36,631 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68458.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:25:47,780 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.848e+02 5.281e+02 6.719e+02 8.986e+02 1.619e+03, threshold=1.344e+03, percent-clipped=5.0 +2023-04-01 18:26:00,681 INFO [train.py:903] (3/4) Epoch 11, batch 200, loss[loss=0.2501, simple_loss=0.3221, pruned_loss=0.089, over 19734.00 frames. ], tot_loss[loss=0.2382, simple_loss=0.3104, pruned_loss=0.08302, over 2438066.85 frames. ], batch size: 63, lr: 7.76e-03, grad_scale: 8.0 +2023-04-01 18:26:02,034 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 18:27:03,595 INFO [train.py:903] (3/4) Epoch 11, batch 250, loss[loss=0.1965, simple_loss=0.2768, pruned_loss=0.05806, over 19618.00 frames. ], tot_loss[loss=0.238, simple_loss=0.3099, pruned_loss=0.08306, over 2725421.42 frames. ], batch size: 50, lr: 7.75e-03, grad_scale: 8.0 +2023-04-01 18:27:46,871 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68564.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:27:50,251 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4162, 1.5388, 2.0172, 1.9042, 3.1228, 4.0377, 3.9961, 4.3588], + device='cuda:3'), covar=tensor([0.1571, 0.3231, 0.2840, 0.1773, 0.0552, 0.0260, 0.0174, 0.0171], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0295, 0.0321, 0.0251, 0.0212, 0.0154, 0.0204, 0.0193], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 18:27:51,031 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.001e+02 5.512e+02 6.613e+02 8.406e+02 1.798e+03, threshold=1.323e+03, percent-clipped=1.0 +2023-04-01 18:28:07,068 INFO [train.py:903] (3/4) Epoch 11, batch 300, loss[loss=0.2384, simple_loss=0.3137, pruned_loss=0.08156, over 19522.00 frames. ], tot_loss[loss=0.2379, simple_loss=0.3099, pruned_loss=0.08293, over 2974967.78 frames. ], batch size: 56, lr: 7.75e-03, grad_scale: 8.0 +2023-04-01 18:28:27,756 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68596.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:28:30,584 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-01 18:29:09,811 INFO [train.py:903] (3/4) Epoch 11, batch 350, loss[loss=0.181, simple_loss=0.2616, pruned_loss=0.05024, over 19754.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3092, pruned_loss=0.08239, over 3169358.91 frames. ], batch size: 46, lr: 7.75e-03, grad_scale: 8.0 +2023-04-01 18:29:16,649 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 18:29:40,788 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3567, 2.1466, 1.6518, 1.4512, 2.0644, 1.1765, 1.2390, 1.7515], + device='cuda:3'), covar=tensor([0.0919, 0.0644, 0.0870, 0.0642, 0.0425, 0.1028, 0.0662, 0.0417], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0294, 0.0325, 0.0243, 0.0232, 0.0316, 0.0285, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:29:57,402 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.055e+02 5.409e+02 6.235e+02 7.842e+02 1.948e+03, threshold=1.247e+03, percent-clipped=7.0 +2023-04-01 18:30:09,872 INFO [train.py:903] (3/4) Epoch 11, batch 400, loss[loss=0.2008, simple_loss=0.2754, pruned_loss=0.06308, over 19398.00 frames. ], tot_loss[loss=0.2379, simple_loss=0.3098, pruned_loss=0.08302, over 3302588.91 frames. ], batch size: 47, lr: 7.74e-03, grad_scale: 8.0 +2023-04-01 18:30:10,350 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8706, 1.3179, 1.0462, 0.9912, 1.1604, 0.9542, 0.9854, 1.1982], + device='cuda:3'), covar=tensor([0.0565, 0.0785, 0.1074, 0.0559, 0.0493, 0.1183, 0.0508, 0.0430], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0294, 0.0326, 0.0244, 0.0233, 0.0317, 0.0286, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:30:40,470 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68703.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:30:54,278 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68714.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:31:11,172 INFO [train.py:903] (3/4) Epoch 11, batch 450, loss[loss=0.2337, simple_loss=0.3134, pruned_loss=0.07698, over 19600.00 frames. ], tot_loss[loss=0.2357, simple_loss=0.3075, pruned_loss=0.08198, over 3426568.44 frames. ], batch size: 61, lr: 7.74e-03, grad_scale: 8.0 +2023-04-01 18:31:25,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68739.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:31:49,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 18:31:50,646 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 18:31:51,685 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68761.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:31:59,523 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.166e+02 5.180e+02 6.491e+02 8.751e+02 1.660e+03, threshold=1.298e+03, percent-clipped=7.0 +2023-04-01 18:32:13,193 INFO [train.py:903] (3/4) Epoch 11, batch 500, loss[loss=0.1843, simple_loss=0.2627, pruned_loss=0.05295, over 19594.00 frames. ], tot_loss[loss=0.2358, simple_loss=0.3079, pruned_loss=0.08183, over 3516610.95 frames. ], batch size: 50, lr: 7.74e-03, grad_scale: 8.0 +2023-04-01 18:32:31,353 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4604, 1.5329, 1.8877, 1.6679, 2.7214, 2.1560, 2.8176, 1.4385], + device='cuda:3'), covar=tensor([0.1935, 0.3282, 0.1973, 0.1596, 0.1298, 0.1765, 0.1355, 0.3255], + device='cuda:3'), in_proj_covar=tensor([0.0479, 0.0565, 0.0588, 0.0432, 0.0585, 0.0485, 0.0644, 0.0485], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:33:02,552 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.89 vs. limit=5.0 +2023-04-01 18:33:05,570 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68820.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:33:11,248 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5473, 1.4487, 1.4118, 1.5821, 3.0729, 1.1109, 2.0452, 3.3718], + device='cuda:3'), covar=tensor([0.0484, 0.2450, 0.2651, 0.1676, 0.0767, 0.2418, 0.1379, 0.0304], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0336, 0.0351, 0.0319, 0.0347, 0.0335, 0.0331, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:33:17,990 INFO [train.py:903] (3/4) Epoch 11, batch 550, loss[loss=0.2259, simple_loss=0.3014, pruned_loss=0.07515, over 19832.00 frames. ], tot_loss[loss=0.2372, simple_loss=0.3091, pruned_loss=0.08264, over 3580287.89 frames. ], batch size: 52, lr: 7.74e-03, grad_scale: 8.0 +2023-04-01 18:33:36,713 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68845.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:34:06,690 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.095e+02 5.206e+02 6.523e+02 8.559e+02 1.532e+03, threshold=1.305e+03, percent-clipped=5.0 +2023-04-01 18:34:18,091 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68876.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:34:21,109 INFO [train.py:903] (3/4) Epoch 11, batch 600, loss[loss=0.287, simple_loss=0.3358, pruned_loss=0.1191, over 13759.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.3079, pruned_loss=0.08152, over 3632801.55 frames. ], batch size: 135, lr: 7.73e-03, grad_scale: 8.0 +2023-04-01 18:35:07,353 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 18:35:23,584 INFO [train.py:903] (3/4) Epoch 11, batch 650, loss[loss=0.1883, simple_loss=0.2662, pruned_loss=0.05513, over 19059.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.3085, pruned_loss=0.08185, over 3687390.04 frames. ], batch size: 42, lr: 7.73e-03, grad_scale: 4.0 +2023-04-01 18:35:36,884 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=68940.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:36:11,531 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0136, 1.7819, 2.1730, 1.8966, 4.3710, 1.1738, 2.5735, 4.7730], + device='cuda:3'), covar=tensor([0.0370, 0.2700, 0.2460, 0.1693, 0.0777, 0.2644, 0.1186, 0.0192], + device='cuda:3'), in_proj_covar=tensor([0.0353, 0.0336, 0.0351, 0.0318, 0.0346, 0.0332, 0.0330, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:36:14,706 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.863e+02 5.561e+02 7.157e+02 8.945e+02 2.089e+03, threshold=1.431e+03, percent-clipped=8.0 +2023-04-01 18:36:26,509 INFO [train.py:903] (3/4) Epoch 11, batch 700, loss[loss=0.2482, simple_loss=0.3262, pruned_loss=0.08511, over 19742.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.3082, pruned_loss=0.08144, over 3727282.84 frames. ], batch size: 63, lr: 7.73e-03, grad_scale: 4.0 +2023-04-01 18:36:41,605 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.39 vs. limit=5.0 +2023-04-01 18:37:30,478 INFO [train.py:903] (3/4) Epoch 11, batch 750, loss[loss=0.2581, simple_loss=0.329, pruned_loss=0.09362, over 19299.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.3087, pruned_loss=0.08172, over 3757341.08 frames. ], batch size: 66, lr: 7.72e-03, grad_scale: 4.0 +2023-04-01 18:37:44,831 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0209, 1.9418, 1.7264, 1.5531, 1.3025, 1.5297, 0.4419, 0.8369], + device='cuda:3'), covar=tensor([0.0411, 0.0422, 0.0309, 0.0515, 0.0981, 0.0605, 0.0849, 0.0763], + device='cuda:3'), in_proj_covar=tensor([0.0333, 0.0329, 0.0330, 0.0352, 0.0425, 0.0349, 0.0309, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 18:37:52,539 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69047.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:38:01,898 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69055.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:38:10,049 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69061.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:38:20,856 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.322e+02 5.289e+02 6.423e+02 8.063e+02 1.861e+03, threshold=1.285e+03, percent-clipped=2.0 +2023-04-01 18:38:33,460 INFO [train.py:903] (3/4) Epoch 11, batch 800, loss[loss=0.2683, simple_loss=0.3315, pruned_loss=0.1026, over 19601.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.3079, pruned_loss=0.0814, over 3778570.43 frames. ], batch size: 61, lr: 7.72e-03, grad_scale: 8.0 +2023-04-01 18:38:49,967 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 18:39:34,220 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5618, 4.0885, 2.4857, 3.6766, 0.9946, 3.8732, 3.8404, 3.9853], + device='cuda:3'), covar=tensor([0.0579, 0.1070, 0.2101, 0.0779, 0.3876, 0.0735, 0.0763, 0.0979], + device='cuda:3'), in_proj_covar=tensor([0.0421, 0.0362, 0.0429, 0.0315, 0.0373, 0.0361, 0.0349, 0.0382], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 18:39:35,201 INFO [train.py:903] (3/4) Epoch 11, batch 850, loss[loss=0.2256, simple_loss=0.3056, pruned_loss=0.07281, over 19615.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3086, pruned_loss=0.08157, over 3805211.46 frames. ], batch size: 57, lr: 7.72e-03, grad_scale: 8.0 +2023-04-01 18:39:39,344 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69132.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:39:41,078 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-04-01 18:40:04,946 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7378, 1.8042, 1.9764, 2.4613, 1.5778, 2.2376, 2.2201, 1.8587], + device='cuda:3'), covar=tensor([0.3352, 0.2772, 0.1392, 0.1639, 0.3178, 0.1427, 0.3204, 0.2568], + device='cuda:3'), in_proj_covar=tensor([0.0779, 0.0795, 0.0641, 0.0886, 0.0771, 0.0698, 0.0780, 0.0704], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 18:40:09,379 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2315, 2.9275, 2.0062, 2.6754, 0.7045, 2.8565, 2.7708, 2.8503], + device='cuda:3'), covar=tensor([0.1061, 0.1306, 0.2143, 0.0915, 0.3751, 0.1051, 0.0916, 0.1399], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0359, 0.0427, 0.0314, 0.0371, 0.0360, 0.0348, 0.0380], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 18:40:12,647 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69157.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:40:18,350 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69162.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:40:26,036 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.965e+02 5.808e+02 7.359e+02 9.451e+02 2.011e+03, threshold=1.472e+03, percent-clipped=12.0 +2023-04-01 18:40:32,113 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 18:40:38,101 INFO [train.py:903] (3/4) Epoch 11, batch 900, loss[loss=0.2132, simple_loss=0.2904, pruned_loss=0.06796, over 19662.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3081, pruned_loss=0.08116, over 3812698.75 frames. ], batch size: 53, lr: 7.72e-03, grad_scale: 8.0 +2023-04-01 18:40:39,673 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69180.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:40:48,343 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69186.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:41:28,530 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-01 18:41:42,669 INFO [train.py:903] (3/4) Epoch 11, batch 950, loss[loss=0.2099, simple_loss=0.2915, pruned_loss=0.06415, over 19680.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.3086, pruned_loss=0.08121, over 3816304.22 frames. ], batch size: 53, lr: 7.71e-03, grad_scale: 8.0 +2023-04-01 18:41:49,490 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 18:42:01,319 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69243.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:42:11,247 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-01 18:42:24,851 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69263.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:42:33,739 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.446e+02 5.204e+02 6.315e+02 7.601e+02 2.315e+03, threshold=1.263e+03, percent-clipped=1.0 +2023-04-01 18:42:47,233 INFO [train.py:903] (3/4) Epoch 11, batch 1000, loss[loss=0.2888, simple_loss=0.3441, pruned_loss=0.1168, over 13571.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.3087, pruned_loss=0.08168, over 3810817.55 frames. ], batch size: 136, lr: 7.71e-03, grad_scale: 8.0 +2023-04-01 18:43:01,215 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69291.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:43:27,049 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69311.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:43:43,482 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 18:43:48,901 INFO [train.py:903] (3/4) Epoch 11, batch 1050, loss[loss=0.2346, simple_loss=0.3116, pruned_loss=0.07875, over 19659.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.3083, pruned_loss=0.08131, over 3806111.69 frames. ], batch size: 55, lr: 7.71e-03, grad_scale: 8.0 +2023-04-01 18:43:57,031 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69336.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:43:59,174 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3820, 3.9574, 2.4846, 3.5399, 1.1465, 3.7272, 3.7161, 3.7658], + device='cuda:3'), covar=tensor([0.0623, 0.0905, 0.2007, 0.0811, 0.3766, 0.0892, 0.0829, 0.1113], + device='cuda:3'), in_proj_covar=tensor([0.0423, 0.0361, 0.0428, 0.0312, 0.0375, 0.0361, 0.0350, 0.0383], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 18:43:59,263 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69338.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:43:59,385 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9208, 1.9479, 1.6992, 1.4335, 1.2806, 1.5365, 0.3980, 0.8564], + device='cuda:3'), covar=tensor([0.0636, 0.0580, 0.0395, 0.0665, 0.1197, 0.0712, 0.0954, 0.0940], + device='cuda:3'), in_proj_covar=tensor([0.0336, 0.0329, 0.0331, 0.0352, 0.0425, 0.0348, 0.0309, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 18:44:02,815 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2554, 3.6999, 2.2553, 2.2521, 3.4300, 2.0462, 1.5264, 1.9955], + device='cuda:3'), covar=tensor([0.1039, 0.0426, 0.0829, 0.0724, 0.0364, 0.0943, 0.0845, 0.0642], + device='cuda:3'), in_proj_covar=tensor([0.0285, 0.0301, 0.0328, 0.0246, 0.0237, 0.0318, 0.0289, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:44:24,568 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 18:44:38,149 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.334e+02 5.222e+02 6.475e+02 7.672e+02 1.315e+03, threshold=1.295e+03, percent-clipped=1.0 +2023-04-01 18:44:49,372 INFO [train.py:903] (3/4) Epoch 11, batch 1100, loss[loss=0.2166, simple_loss=0.305, pruned_loss=0.06414, over 19769.00 frames. ], tot_loss[loss=0.2357, simple_loss=0.3085, pruned_loss=0.08147, over 3806447.22 frames. ], batch size: 54, lr: 7.70e-03, grad_scale: 8.0 +2023-04-01 18:44:57,866 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69386.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:45:24,084 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69405.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:45:27,992 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2511, 1.3403, 1.2273, 1.0260, 1.0791, 1.1139, 0.0482, 0.3592], + device='cuda:3'), covar=tensor([0.0477, 0.0424, 0.0290, 0.0379, 0.0876, 0.0372, 0.0770, 0.0702], + device='cuda:3'), in_proj_covar=tensor([0.0333, 0.0327, 0.0330, 0.0350, 0.0423, 0.0346, 0.0307, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 18:45:32,698 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69412.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:45:39,614 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69418.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:45:53,079 INFO [train.py:903] (3/4) Epoch 11, batch 1150, loss[loss=0.2193, simple_loss=0.296, pruned_loss=0.07131, over 19671.00 frames. ], tot_loss[loss=0.2375, simple_loss=0.3102, pruned_loss=0.08238, over 3817715.22 frames. ], batch size: 55, lr: 7.70e-03, grad_scale: 8.0 +2023-04-01 18:46:11,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69443.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:46:41,981 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.975e+02 5.885e+02 7.181e+02 8.495e+02 1.651e+03, threshold=1.436e+03, percent-clipped=3.0 +2023-04-01 18:46:55,767 INFO [train.py:903] (3/4) Epoch 11, batch 1200, loss[loss=0.3033, simple_loss=0.3536, pruned_loss=0.1265, over 13262.00 frames. ], tot_loss[loss=0.2372, simple_loss=0.31, pruned_loss=0.08216, over 3818457.35 frames. ], batch size: 136, lr: 7.70e-03, grad_scale: 8.0 +2023-04-01 18:46:58,595 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.89 vs. limit=5.0 +2023-04-01 18:47:01,621 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7005, 4.2059, 4.4230, 4.4091, 1.4104, 4.1244, 3.5718, 4.1133], + device='cuda:3'), covar=tensor([0.1428, 0.0784, 0.0530, 0.0576, 0.5493, 0.0559, 0.0644, 0.1036], + device='cuda:3'), in_proj_covar=tensor([0.0662, 0.0594, 0.0783, 0.0668, 0.0723, 0.0541, 0.0486, 0.0725], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 18:47:27,277 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 18:47:48,950 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69520.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:47:53,431 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69524.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:47:59,004 INFO [train.py:903] (3/4) Epoch 11, batch 1250, loss[loss=0.2593, simple_loss=0.3329, pruned_loss=0.09286, over 19727.00 frames. ], tot_loss[loss=0.2376, simple_loss=0.3104, pruned_loss=0.08233, over 3805716.62 frames. ], batch size: 63, lr: 7.70e-03, grad_scale: 8.0 +2023-04-01 18:48:00,223 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69530.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:48:49,514 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.132e+02 5.575e+02 6.899e+02 8.428e+02 1.860e+03, threshold=1.380e+03, percent-clipped=3.0 +2023-04-01 18:49:00,826 INFO [train.py:903] (3/4) Epoch 11, batch 1300, loss[loss=0.2133, simple_loss=0.303, pruned_loss=0.06181, over 18825.00 frames. ], tot_loss[loss=0.2375, simple_loss=0.3104, pruned_loss=0.08236, over 3827113.22 frames. ], batch size: 74, lr: 7.69e-03, grad_scale: 8.0 +2023-04-01 18:49:10,211 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69587.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:49:38,641 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69607.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:50:04,326 INFO [train.py:903] (3/4) Epoch 11, batch 1350, loss[loss=0.2338, simple_loss=0.3055, pruned_loss=0.08107, over 17273.00 frames. ], tot_loss[loss=0.2377, simple_loss=0.3109, pruned_loss=0.08226, over 3809965.96 frames. ], batch size: 101, lr: 7.69e-03, grad_scale: 8.0 +2023-04-01 18:50:13,535 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69635.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:50:19,167 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69639.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:50:26,772 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69645.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:50:30,832 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.43 vs. limit=5.0 +2023-04-01 18:50:54,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.470e+02 5.709e+02 6.895e+02 8.476e+02 1.895e+03, threshold=1.379e+03, percent-clipped=5.0 +2023-04-01 18:51:08,229 INFO [train.py:903] (3/4) Epoch 11, batch 1400, loss[loss=0.213, simple_loss=0.2743, pruned_loss=0.07587, over 19322.00 frames. ], tot_loss[loss=0.2364, simple_loss=0.3094, pruned_loss=0.08171, over 3817437.17 frames. ], batch size: 44, lr: 7.69e-03, grad_scale: 8.0 +2023-04-01 18:51:12,914 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69682.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:51:24,977 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69692.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:51:36,551 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69702.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:52:03,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69722.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:52:09,959 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 18:52:10,948 INFO [train.py:903] (3/4) Epoch 11, batch 1450, loss[loss=0.2761, simple_loss=0.3249, pruned_loss=0.1136, over 19744.00 frames. ], tot_loss[loss=0.2373, simple_loss=0.3102, pruned_loss=0.08223, over 3822748.70 frames. ], batch size: 51, lr: 7.69e-03, grad_scale: 8.0 +2023-04-01 18:52:12,413 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69730.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:52:36,751 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:52:44,834 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=69756.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:53:01,499 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.562e+02 5.344e+02 6.531e+02 8.357e+02 2.062e+03, threshold=1.306e+03, percent-clipped=2.0 +2023-04-01 18:53:10,372 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69776.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:53:13,430 INFO [train.py:903] (3/4) Epoch 11, batch 1500, loss[loss=0.2225, simple_loss=0.2986, pruned_loss=0.07316, over 19590.00 frames. ], tot_loss[loss=0.2366, simple_loss=0.3094, pruned_loss=0.08194, over 3835202.19 frames. ], batch size: 61, lr: 7.68e-03, grad_scale: 8.0 +2023-04-01 18:53:36,797 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69797.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:53:43,105 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69801.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:54:15,941 INFO [train.py:903] (3/4) Epoch 11, batch 1550, loss[loss=0.2006, simple_loss=0.2675, pruned_loss=0.0669, over 19755.00 frames. ], tot_loss[loss=0.2364, simple_loss=0.309, pruned_loss=0.08187, over 3827914.52 frames. ], batch size: 46, lr: 7.68e-03, grad_scale: 8.0 +2023-04-01 18:54:38,378 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69845.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:55:07,633 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.582e+02 5.814e+02 6.999e+02 8.951e+02 2.972e+03, threshold=1.400e+03, percent-clipped=5.0 +2023-04-01 18:55:09,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69871.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:55:20,439 INFO [train.py:903] (3/4) Epoch 11, batch 1600, loss[loss=0.1882, simple_loss=0.2611, pruned_loss=0.05763, over 19324.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.3085, pruned_loss=0.08185, over 3829945.40 frames. ], batch size: 44, lr: 7.68e-03, grad_scale: 8.0 +2023-04-01 18:55:38,103 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-01 18:55:41,190 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69895.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:55:46,716 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 18:55:48,146 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69901.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:56:12,826 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69920.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:56:20,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69926.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:56:23,826 INFO [train.py:903] (3/4) Epoch 11, batch 1650, loss[loss=0.2283, simple_loss=0.2934, pruned_loss=0.08156, over 19380.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.3079, pruned_loss=0.08138, over 3827284.69 frames. ], batch size: 47, lr: 7.67e-03, grad_scale: 8.0 +2023-04-01 18:56:33,416 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1625, 1.7591, 1.5954, 2.1520, 1.9776, 1.7886, 1.5683, 1.9096], + device='cuda:3'), covar=tensor([0.0828, 0.1514, 0.1414, 0.0842, 0.1178, 0.0484, 0.1316, 0.0645], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0345, 0.0286, 0.0236, 0.0295, 0.0241, 0.0275, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 18:57:00,776 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69958.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:57:16,657 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.105e+02 5.294e+02 6.504e+02 8.314e+02 1.576e+03, threshold=1.301e+03, percent-clipped=2.0 +2023-04-01 18:57:26,345 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69978.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:57:27,010 INFO [train.py:903] (3/4) Epoch 11, batch 1700, loss[loss=0.2069, simple_loss=0.2771, pruned_loss=0.06832, over 19335.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3083, pruned_loss=0.08149, over 3825039.85 frames. ], batch size: 44, lr: 7.67e-03, grad_scale: 8.0 +2023-04-01 18:57:32,087 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69983.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:57:59,102 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70003.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 18:58:03,177 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70006.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:58:09,483 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 18:58:29,307 INFO [train.py:903] (3/4) Epoch 11, batch 1750, loss[loss=0.2624, simple_loss=0.3343, pruned_loss=0.09527, over 19669.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3093, pruned_loss=0.08232, over 3820448.84 frames. ], batch size: 58, lr: 7.67e-03, grad_scale: 8.0 +2023-04-01 18:58:31,987 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70031.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:58:40,221 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=70036.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 18:59:02,458 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70053.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:59:11,019 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-01 18:59:22,247 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.188e+02 6.480e+02 8.127e+02 9.904e+02 1.581e+03, threshold=1.625e+03, percent-clipped=5.0 +2023-04-01 18:59:34,189 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70078.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 18:59:34,847 INFO [train.py:903] (3/4) Epoch 11, batch 1800, loss[loss=0.2241, simple_loss=0.3009, pruned_loss=0.07371, over 19389.00 frames. ], tot_loss[loss=0.2378, simple_loss=0.3103, pruned_loss=0.0827, over 3815713.35 frames. ], batch size: 48, lr: 7.67e-03, grad_scale: 8.0 +2023-04-01 19:00:02,666 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70101.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:00:08,716 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-04-01 19:00:34,705 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70126.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:00:35,503 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 19:00:36,753 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70127.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:00:37,763 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=70128.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:00:38,700 INFO [train.py:903] (3/4) Epoch 11, batch 1850, loss[loss=0.2589, simple_loss=0.3342, pruned_loss=0.0918, over 19539.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3098, pruned_loss=0.08213, over 3814332.93 frames. ], batch size: 56, lr: 7.66e-03, grad_scale: 8.0 +2023-04-01 19:01:05,172 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70151.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 19:01:06,370 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70152.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:01:13,693 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 19:01:30,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.636e+02 5.171e+02 6.551e+02 7.968e+02 1.780e+03, threshold=1.310e+03, percent-clipped=1.0 +2023-04-01 19:01:41,532 INFO [train.py:903] (3/4) Epoch 11, batch 1900, loss[loss=0.2152, simple_loss=0.2972, pruned_loss=0.06661, over 19666.00 frames. ], tot_loss[loss=0.2363, simple_loss=0.3094, pruned_loss=0.08164, over 3821413.21 frames. ], batch size: 53, lr: 7.66e-03, grad_scale: 8.0 +2023-04-01 19:01:57,150 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 19:02:04,776 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 19:02:28,859 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 19:02:42,952 INFO [train.py:903] (3/4) Epoch 11, batch 1950, loss[loss=0.2498, simple_loss=0.3244, pruned_loss=0.08765, over 18926.00 frames. ], tot_loss[loss=0.2369, simple_loss=0.3097, pruned_loss=0.08205, over 3822741.33 frames. ], batch size: 74, lr: 7.66e-03, grad_scale: 8.0 +2023-04-01 19:03:35,381 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.353e+02 4.987e+02 6.114e+02 7.804e+02 3.131e+03, threshold=1.223e+03, percent-clipped=9.0 +2023-04-01 19:03:37,258 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-04-01 19:03:43,676 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3894, 2.2513, 1.7911, 1.4530, 2.0881, 1.3223, 1.2786, 1.9376], + device='cuda:3'), covar=tensor([0.0784, 0.0565, 0.0801, 0.0722, 0.0402, 0.0997, 0.0627, 0.0334], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0293, 0.0321, 0.0242, 0.0231, 0.0314, 0.0284, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:03:45,798 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=70278.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:03:46,664 INFO [train.py:903] (3/4) Epoch 11, batch 2000, loss[loss=0.2514, simple_loss=0.3235, pruned_loss=0.08971, over 19375.00 frames. ], tot_loss[loss=0.2372, simple_loss=0.31, pruned_loss=0.08223, over 3813582.94 frames. ], batch size: 70, lr: 7.66e-03, grad_scale: 8.0 +2023-04-01 19:04:48,767 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 19:04:51,058 INFO [train.py:903] (3/4) Epoch 11, batch 2050, loss[loss=0.2657, simple_loss=0.3329, pruned_loss=0.09923, over 18816.00 frames. ], tot_loss[loss=0.2367, simple_loss=0.3098, pruned_loss=0.0818, over 3818280.82 frames. ], batch size: 74, lr: 7.65e-03, grad_scale: 8.0 +2023-04-01 19:05:07,049 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 19:05:08,148 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 19:05:27,796 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 19:05:43,673 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.125e+02 5.481e+02 7.126e+02 9.531e+02 2.002e+03, threshold=1.425e+03, percent-clipped=10.0 +2023-04-01 19:05:54,410 INFO [train.py:903] (3/4) Epoch 11, batch 2100, loss[loss=0.3509, simple_loss=0.3838, pruned_loss=0.159, over 13747.00 frames. ], tot_loss[loss=0.2381, simple_loss=0.3107, pruned_loss=0.08271, over 3816355.40 frames. ], batch size: 136, lr: 7.65e-03, grad_scale: 8.0 +2023-04-01 19:06:24,811 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 19:06:30,870 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70407.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 19:06:48,465 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 19:06:57,561 INFO [train.py:903] (3/4) Epoch 11, batch 2150, loss[loss=0.1956, simple_loss=0.2691, pruned_loss=0.06107, over 19773.00 frames. ], tot_loss[loss=0.2379, simple_loss=0.3106, pruned_loss=0.08262, over 3819095.08 frames. ], batch size: 47, lr: 7.65e-03, grad_scale: 8.0 +2023-04-01 19:07:01,543 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70432.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 19:07:51,367 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.544e+02 5.585e+02 6.733e+02 8.759e+02 1.859e+03, threshold=1.347e+03, percent-clipped=4.0 +2023-04-01 19:07:52,877 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=70472.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:08:02,119 INFO [train.py:903] (3/4) Epoch 11, batch 2200, loss[loss=0.2468, simple_loss=0.3213, pruned_loss=0.08618, over 19532.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.3099, pruned_loss=0.08216, over 3818970.30 frames. ], batch size: 54, lr: 7.64e-03, grad_scale: 4.0 +2023-04-01 19:09:06,388 INFO [train.py:903] (3/4) Epoch 11, batch 2250, loss[loss=0.2381, simple_loss=0.3135, pruned_loss=0.08135, over 19558.00 frames. ], tot_loss[loss=0.2367, simple_loss=0.3096, pruned_loss=0.08186, over 3814250.55 frames. ], batch size: 61, lr: 7.64e-03, grad_scale: 4.0 +2023-04-01 19:10:01,000 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.382e+02 4.990e+02 6.839e+02 8.349e+02 1.575e+03, threshold=1.368e+03, percent-clipped=2.0 +2023-04-01 19:10:10,416 INFO [train.py:903] (3/4) Epoch 11, batch 2300, loss[loss=0.23, simple_loss=0.3044, pruned_loss=0.07776, over 19660.00 frames. ], tot_loss[loss=0.2368, simple_loss=0.3099, pruned_loss=0.08185, over 3819304.37 frames. ], batch size: 55, lr: 7.64e-03, grad_scale: 4.0 +2023-04-01 19:10:19,835 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70587.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:10:23,148 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 19:11:05,025 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=70622.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:11:13,191 INFO [train.py:903] (3/4) Epoch 11, batch 2350, loss[loss=0.2329, simple_loss=0.3085, pruned_loss=0.07868, over 19543.00 frames. ], tot_loss[loss=0.2343, simple_loss=0.3078, pruned_loss=0.08047, over 3819510.73 frames. ], batch size: 56, lr: 7.64e-03, grad_scale: 4.0 +2023-04-01 19:11:27,849 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9922, 2.0902, 2.3144, 2.9608, 1.9373, 2.6758, 2.6583, 2.0914], + device='cuda:3'), covar=tensor([0.3666, 0.3203, 0.1375, 0.1721, 0.3511, 0.1488, 0.3073, 0.2655], + device='cuda:3'), in_proj_covar=tensor([0.0788, 0.0799, 0.0644, 0.0887, 0.0776, 0.0694, 0.0779, 0.0706], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 19:11:56,876 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 19:12:06,972 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.408e+02 5.270e+02 6.524e+02 8.351e+02 2.247e+03, threshold=1.305e+03, percent-clipped=5.0 +2023-04-01 19:12:12,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 19:12:16,797 INFO [train.py:903] (3/4) Epoch 11, batch 2400, loss[loss=0.204, simple_loss=0.2839, pruned_loss=0.06199, over 19483.00 frames. ], tot_loss[loss=0.2329, simple_loss=0.3064, pruned_loss=0.07972, over 3819924.43 frames. ], batch size: 49, lr: 7.63e-03, grad_scale: 8.0 +2023-04-01 19:12:58,761 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5964, 1.6727, 2.0329, 1.6433, 2.5871, 3.0024, 2.8731, 3.1262], + device='cuda:3'), covar=tensor([0.1289, 0.2801, 0.2411, 0.2022, 0.1055, 0.0353, 0.0210, 0.0251], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0295, 0.0320, 0.0251, 0.0215, 0.0155, 0.0205, 0.0196], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 19:13:20,247 INFO [train.py:903] (3/4) Epoch 11, batch 2450, loss[loss=0.2486, simple_loss=0.3287, pruned_loss=0.08423, over 19495.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3088, pruned_loss=0.08119, over 3811932.51 frames. ], batch size: 64, lr: 7.63e-03, grad_scale: 8.0 +2023-04-01 19:13:32,426 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70737.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:13:44,507 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8710, 1.5982, 1.5764, 1.9129, 1.7509, 1.6604, 1.6068, 1.8160], + device='cuda:3'), covar=tensor([0.0851, 0.1295, 0.1228, 0.0776, 0.1026, 0.0477, 0.1127, 0.0608], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0347, 0.0287, 0.0238, 0.0296, 0.0242, 0.0276, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:14:14,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.733e+02 5.717e+02 6.752e+02 8.721e+02 4.215e+03, threshold=1.350e+03, percent-clipped=8.0 +2023-04-01 19:14:25,713 INFO [train.py:903] (3/4) Epoch 11, batch 2500, loss[loss=0.2218, simple_loss=0.2857, pruned_loss=0.07898, over 19485.00 frames. ], tot_loss[loss=0.236, simple_loss=0.309, pruned_loss=0.08147, over 3810142.12 frames. ], batch size: 49, lr: 7.63e-03, grad_scale: 8.0 +2023-04-01 19:15:27,560 INFO [train.py:903] (3/4) Epoch 11, batch 2550, loss[loss=0.2457, simple_loss=0.3163, pruned_loss=0.08749, over 19656.00 frames. ], tot_loss[loss=0.2365, simple_loss=0.3096, pruned_loss=0.08171, over 3813509.52 frames. ], batch size: 60, lr: 7.63e-03, grad_scale: 8.0 +2023-04-01 19:15:46,682 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70843.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:16:17,534 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70868.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:16:20,506 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.586e+02 5.428e+02 6.511e+02 7.969e+02 1.423e+03, threshold=1.302e+03, percent-clipped=4.0 +2023-04-01 19:16:26,492 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 19:16:30,128 INFO [train.py:903] (3/4) Epoch 11, batch 2600, loss[loss=0.2043, simple_loss=0.2747, pruned_loss=0.06697, over 19061.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.309, pruned_loss=0.08165, over 3815994.61 frames. ], batch size: 42, lr: 7.62e-03, grad_scale: 8.0 +2023-04-01 19:17:34,823 INFO [train.py:903] (3/4) Epoch 11, batch 2650, loss[loss=0.2297, simple_loss=0.311, pruned_loss=0.0742, over 19675.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3086, pruned_loss=0.08165, over 3807427.35 frames. ], batch size: 58, lr: 7.62e-03, grad_scale: 8.0 +2023-04-01 19:17:56,637 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 19:18:28,065 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.326e+02 5.322e+02 6.264e+02 9.085e+02 2.401e+03, threshold=1.253e+03, percent-clipped=10.0 +2023-04-01 19:18:39,374 INFO [train.py:903] (3/4) Epoch 11, batch 2700, loss[loss=0.2025, simple_loss=0.2903, pruned_loss=0.0573, over 19355.00 frames. ], tot_loss[loss=0.235, simple_loss=0.3079, pruned_loss=0.08102, over 3822487.18 frames. ], batch size: 70, lr: 7.62e-03, grad_scale: 8.0 +2023-04-01 19:18:56,261 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70993.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:19:18,237 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71010.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:19:30,052 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71018.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:19:43,444 INFO [train.py:903] (3/4) Epoch 11, batch 2750, loss[loss=0.1922, simple_loss=0.2726, pruned_loss=0.05588, over 19468.00 frames. ], tot_loss[loss=0.2348, simple_loss=0.3078, pruned_loss=0.08096, over 3811339.78 frames. ], batch size: 49, lr: 7.62e-03, grad_scale: 8.0 +2023-04-01 19:20:08,271 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71048.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:20:37,009 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 5.301e+02 6.565e+02 7.951e+02 1.458e+03, threshold=1.313e+03, percent-clipped=6.0 +2023-04-01 19:20:45,875 INFO [train.py:903] (3/4) Epoch 11, batch 2800, loss[loss=0.2491, simple_loss=0.3207, pruned_loss=0.08875, over 19722.00 frames. ], tot_loss[loss=0.236, simple_loss=0.3089, pruned_loss=0.08158, over 3809780.69 frames. ], batch size: 51, lr: 7.61e-03, grad_scale: 8.0 +2023-04-01 19:21:29,710 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-01 19:21:51,261 INFO [train.py:903] (3/4) Epoch 11, batch 2850, loss[loss=0.2367, simple_loss=0.3138, pruned_loss=0.07976, over 19514.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3091, pruned_loss=0.08139, over 3817723.87 frames. ], batch size: 64, lr: 7.61e-03, grad_scale: 8.0 +2023-04-01 19:22:45,074 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.239e+02 5.466e+02 6.713e+02 8.640e+02 2.236e+03, threshold=1.343e+03, percent-clipped=7.0 +2023-04-01 19:22:54,322 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 19:22:55,470 INFO [train.py:903] (3/4) Epoch 11, batch 2900, loss[loss=0.226, simple_loss=0.3015, pruned_loss=0.07527, over 19783.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.309, pruned_loss=0.08136, over 3826420.01 frames. ], batch size: 54, lr: 7.61e-03, grad_scale: 8.0 +2023-04-01 19:23:03,498 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.77 vs. limit=5.0 +2023-04-01 19:23:09,064 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9997, 3.6066, 2.4908, 3.2953, 0.9817, 3.3873, 3.3813, 3.4759], + device='cuda:3'), covar=tensor([0.0716, 0.1091, 0.1888, 0.0864, 0.3827, 0.0892, 0.0847, 0.1065], + device='cuda:3'), in_proj_covar=tensor([0.0431, 0.0366, 0.0440, 0.0321, 0.0382, 0.0368, 0.0356, 0.0393], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:24:00,112 INFO [train.py:903] (3/4) Epoch 11, batch 2950, loss[loss=0.2586, simple_loss=0.3273, pruned_loss=0.09498, over 19568.00 frames. ], tot_loss[loss=0.2342, simple_loss=0.3074, pruned_loss=0.08051, over 3842722.78 frames. ], batch size: 61, lr: 7.60e-03, grad_scale: 8.0 +2023-04-01 19:24:53,527 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.649e+02 5.471e+02 6.960e+02 8.532e+02 1.699e+03, threshold=1.392e+03, percent-clipped=7.0 +2023-04-01 19:25:02,841 INFO [train.py:903] (3/4) Epoch 11, batch 3000, loss[loss=0.2284, simple_loss=0.2919, pruned_loss=0.08241, over 19030.00 frames. ], tot_loss[loss=0.235, simple_loss=0.3081, pruned_loss=0.08092, over 3833916.42 frames. ], batch size: 42, lr: 7.60e-03, grad_scale: 8.0 +2023-04-01 19:25:02,841 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 19:25:16,077 INFO [train.py:937] (3/4) Epoch 11, validation: loss=0.1785, simple_loss=0.2793, pruned_loss=0.0389, over 944034.00 frames. +2023-04-01 19:25:16,079 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 19:25:20,545 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 19:25:26,652 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2490, 1.3064, 1.2113, 1.0284, 1.0405, 1.1360, 0.0326, 0.3827], + device='cuda:3'), covar=tensor([0.0451, 0.0479, 0.0295, 0.0373, 0.0969, 0.0423, 0.0867, 0.0766], + device='cuda:3'), in_proj_covar=tensor([0.0342, 0.0335, 0.0335, 0.0356, 0.0429, 0.0358, 0.0312, 0.0326], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 19:25:51,487 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2086, 2.1462, 2.3543, 3.3170, 2.1811, 3.1288, 2.8352, 2.2464], + device='cuda:3'), covar=tensor([0.3574, 0.2999, 0.1426, 0.1807, 0.3656, 0.1425, 0.3053, 0.2541], + device='cuda:3'), in_proj_covar=tensor([0.0789, 0.0803, 0.0647, 0.0890, 0.0781, 0.0701, 0.0784, 0.0705], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 19:26:01,967 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6551, 1.3325, 1.3694, 2.0288, 1.6835, 1.9905, 2.0940, 1.7713], + device='cuda:3'), covar=tensor([0.0808, 0.1023, 0.1076, 0.0834, 0.0857, 0.0661, 0.0803, 0.0666], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0226, 0.0223, 0.0250, 0.0236, 0.0215, 0.0197, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 19:26:20,107 INFO [train.py:903] (3/4) Epoch 11, batch 3050, loss[loss=0.2285, simple_loss=0.3073, pruned_loss=0.07482, over 19686.00 frames. ], tot_loss[loss=0.2345, simple_loss=0.3076, pruned_loss=0.08073, over 3828657.47 frames. ], batch size: 59, lr: 7.60e-03, grad_scale: 8.0 +2023-04-01 19:26:51,684 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=71354.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:27:07,121 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8399, 1.9351, 2.1233, 2.6418, 1.8285, 2.5183, 2.3669, 1.9532], + device='cuda:3'), covar=tensor([0.3473, 0.2912, 0.1366, 0.1698, 0.3310, 0.1429, 0.3290, 0.2613], + device='cuda:3'), in_proj_covar=tensor([0.0786, 0.0802, 0.0645, 0.0888, 0.0777, 0.0699, 0.0783, 0.0703], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 19:27:13,588 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.278e+02 5.484e+02 7.005e+02 9.170e+02 2.820e+03, threshold=1.401e+03, percent-clipped=8.0 +2023-04-01 19:27:22,999 INFO [train.py:903] (3/4) Epoch 11, batch 3100, loss[loss=0.2858, simple_loss=0.3405, pruned_loss=0.1155, over 13264.00 frames. ], tot_loss[loss=0.2362, simple_loss=0.3092, pruned_loss=0.08158, over 3811927.55 frames. ], batch size: 136, lr: 7.60e-03, grad_scale: 8.0 +2023-04-01 19:27:27,833 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1889, 1.6587, 1.7798, 2.6336, 2.0904, 2.5680, 2.5809, 2.2765], + device='cuda:3'), covar=tensor([0.0735, 0.0927, 0.0991, 0.0878, 0.0861, 0.0616, 0.0871, 0.0639], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0228, 0.0224, 0.0251, 0.0238, 0.0215, 0.0198, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 19:27:40,191 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=71392.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:28:26,276 INFO [train.py:903] (3/4) Epoch 11, batch 3150, loss[loss=0.2262, simple_loss=0.2873, pruned_loss=0.08257, over 19056.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3096, pruned_loss=0.08217, over 3808658.25 frames. ], batch size: 42, lr: 7.59e-03, grad_scale: 8.0 +2023-04-01 19:28:55,925 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 19:28:57,864 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-01 19:29:17,641 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=71469.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:29:19,592 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.562e+02 5.450e+02 6.508e+02 8.826e+02 1.534e+03, threshold=1.302e+03, percent-clipped=1.0 +2023-04-01 19:29:31,374 INFO [train.py:903] (3/4) Epoch 11, batch 3200, loss[loss=0.2018, simple_loss=0.2751, pruned_loss=0.06429, over 19785.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.3082, pruned_loss=0.08125, over 3808380.40 frames. ], batch size: 47, lr: 7.59e-03, grad_scale: 8.0 +2023-04-01 19:30:05,880 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=71507.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:30:32,440 INFO [train.py:903] (3/4) Epoch 11, batch 3250, loss[loss=0.2928, simple_loss=0.3529, pruned_loss=0.1164, over 19529.00 frames. ], tot_loss[loss=0.2369, simple_loss=0.3094, pruned_loss=0.08214, over 3808016.65 frames. ], batch size: 54, lr: 7.59e-03, grad_scale: 8.0 +2023-04-01 19:30:35,032 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2404, 1.3992, 1.8164, 1.4416, 2.6647, 2.1627, 2.6742, 1.0459], + device='cuda:3'), covar=tensor([0.2259, 0.3841, 0.2194, 0.1761, 0.1337, 0.1850, 0.1497, 0.3787], + device='cuda:3'), in_proj_covar=tensor([0.0490, 0.0575, 0.0601, 0.0441, 0.0595, 0.0493, 0.0649, 0.0497], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 19:31:26,708 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.701e+02 5.157e+02 6.415e+02 8.641e+02 1.397e+03, threshold=1.283e+03, percent-clipped=1.0 +2023-04-01 19:31:36,054 INFO [train.py:903] (3/4) Epoch 11, batch 3300, loss[loss=0.2574, simple_loss=0.3211, pruned_loss=0.09686, over 19650.00 frames. ], tot_loss[loss=0.2375, simple_loss=0.3099, pruned_loss=0.08255, over 3800692.80 frames. ], batch size: 60, lr: 7.59e-03, grad_scale: 8.0 +2023-04-01 19:31:41,724 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 19:32:32,427 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-01 19:32:39,937 INFO [train.py:903] (3/4) Epoch 11, batch 3350, loss[loss=0.2531, simple_loss=0.3277, pruned_loss=0.08932, over 19324.00 frames. ], tot_loss[loss=0.2381, simple_loss=0.3107, pruned_loss=0.08281, over 3811064.82 frames. ], batch size: 70, lr: 7.58e-03, grad_scale: 8.0 +2023-04-01 19:32:47,341 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71634.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:33:34,251 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.542e+02 5.202e+02 6.815e+02 8.173e+02 2.322e+03, threshold=1.363e+03, percent-clipped=5.0 +2023-04-01 19:33:44,446 INFO [train.py:903] (3/4) Epoch 11, batch 3400, loss[loss=0.2064, simple_loss=0.2777, pruned_loss=0.06757, over 19306.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.31, pruned_loss=0.08207, over 3822045.58 frames. ], batch size: 44, lr: 7.58e-03, grad_scale: 8.0 +2023-04-01 19:34:40,425 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8272, 1.2019, 0.9383, 0.8884, 1.0799, 0.8978, 0.8730, 1.1302], + device='cuda:3'), covar=tensor([0.0535, 0.0700, 0.0932, 0.0552, 0.0445, 0.0971, 0.0518, 0.0400], + device='cuda:3'), in_proj_covar=tensor([0.0291, 0.0301, 0.0329, 0.0250, 0.0235, 0.0321, 0.0289, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:34:44,138 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71725.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:34:48,505 INFO [train.py:903] (3/4) Epoch 11, batch 3450, loss[loss=0.2226, simple_loss=0.3033, pruned_loss=0.07097, over 17302.00 frames. ], tot_loss[loss=0.2381, simple_loss=0.3109, pruned_loss=0.08266, over 3812894.68 frames. ], batch size: 101, lr: 7.58e-03, grad_scale: 8.0 +2023-04-01 19:34:52,161 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 19:35:16,030 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:35:33,077 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71763.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:35:42,113 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.543e+02 5.649e+02 6.835e+02 8.667e+02 1.565e+03, threshold=1.367e+03, percent-clipped=4.0 +2023-04-01 19:35:52,284 INFO [train.py:903] (3/4) Epoch 11, batch 3500, loss[loss=0.2494, simple_loss=0.3373, pruned_loss=0.08073, over 19664.00 frames. ], tot_loss[loss=0.2382, simple_loss=0.3112, pruned_loss=0.08259, over 3815899.90 frames. ], batch size: 58, lr: 7.58e-03, grad_scale: 8.0 +2023-04-01 19:36:04,530 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71788.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:36:33,912 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8252, 1.8053, 2.0170, 2.4414, 1.7686, 2.4150, 2.1872, 1.9049], + device='cuda:3'), covar=tensor([0.3308, 0.2878, 0.1533, 0.1548, 0.2913, 0.1334, 0.3633, 0.2736], + device='cuda:3'), in_proj_covar=tensor([0.0787, 0.0805, 0.0648, 0.0891, 0.0779, 0.0703, 0.0785, 0.0709], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 19:36:56,412 INFO [train.py:903] (3/4) Epoch 11, batch 3550, loss[loss=0.1946, simple_loss=0.2856, pruned_loss=0.05182, over 19695.00 frames. ], tot_loss[loss=0.237, simple_loss=0.31, pruned_loss=0.08199, over 3805516.69 frames. ], batch size: 53, lr: 7.57e-03, grad_scale: 8.0 +2023-04-01 19:37:00,714 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.84 vs. limit=5.0 +2023-04-01 19:37:49,026 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.173e+02 5.129e+02 6.236e+02 8.069e+02 1.994e+03, threshold=1.247e+03, percent-clipped=3.0 +2023-04-01 19:37:58,900 INFO [train.py:903] (3/4) Epoch 11, batch 3600, loss[loss=0.3396, simple_loss=0.3793, pruned_loss=0.1499, over 19702.00 frames. ], tot_loss[loss=0.2377, simple_loss=0.3102, pruned_loss=0.08264, over 3804245.40 frames. ], batch size: 59, lr: 7.57e-03, grad_scale: 8.0 +2023-04-01 19:39:03,005 INFO [train.py:903] (3/4) Epoch 11, batch 3650, loss[loss=0.1883, simple_loss=0.2701, pruned_loss=0.05324, over 19611.00 frames. ], tot_loss[loss=0.2374, simple_loss=0.3101, pruned_loss=0.0824, over 3802783.54 frames. ], batch size: 50, lr: 7.57e-03, grad_scale: 8.0 +2023-04-01 19:39:13,492 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71937.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:39:55,936 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.339e+02 5.422e+02 6.726e+02 7.997e+02 1.955e+03, threshold=1.345e+03, percent-clipped=5.0 +2023-04-01 19:40:05,239 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=71978.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:40:06,256 INFO [train.py:903] (3/4) Epoch 11, batch 3700, loss[loss=0.1885, simple_loss=0.2682, pruned_loss=0.05439, over 19754.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.3096, pruned_loss=0.08225, over 3808130.34 frames. ], batch size: 54, lr: 7.57e-03, grad_scale: 8.0 +2023-04-01 19:41:02,232 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-01 19:41:11,922 INFO [train.py:903] (3/4) Epoch 11, batch 3750, loss[loss=0.2146, simple_loss=0.2811, pruned_loss=0.07402, over 19396.00 frames. ], tot_loss[loss=0.2365, simple_loss=0.3091, pruned_loss=0.08197, over 3815631.67 frames. ], batch size: 48, lr: 7.56e-03, grad_scale: 8.0 +2023-04-01 19:42:06,342 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.116e+02 5.141e+02 6.053e+02 7.322e+02 1.150e+03, threshold=1.211e+03, percent-clipped=0.0 +2023-04-01 19:42:17,092 INFO [train.py:903] (3/4) Epoch 11, batch 3800, loss[loss=0.2515, simple_loss=0.325, pruned_loss=0.08895, over 19527.00 frames. ], tot_loss[loss=0.2364, simple_loss=0.3088, pruned_loss=0.08198, over 3814220.42 frames. ], batch size: 54, lr: 7.56e-03, grad_scale: 8.0 +2023-04-01 19:42:22,597 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-04-01 19:42:35,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72093.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:42:46,960 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 19:43:21,150 INFO [train.py:903] (3/4) Epoch 11, batch 3850, loss[loss=0.254, simple_loss=0.3318, pruned_loss=0.08813, over 19543.00 frames. ], tot_loss[loss=0.2365, simple_loss=0.3093, pruned_loss=0.08186, over 3821603.62 frames. ], batch size: 56, lr: 7.56e-03, grad_scale: 8.0 +2023-04-01 19:43:45,136 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72147.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:44:16,373 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.216e+02 5.600e+02 6.929e+02 8.545e+02 2.074e+03, threshold=1.386e+03, percent-clipped=5.0 +2023-04-01 19:44:25,938 INFO [train.py:903] (3/4) Epoch 11, batch 3900, loss[loss=0.2818, simple_loss=0.3523, pruned_loss=0.1057, over 19448.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.308, pruned_loss=0.0812, over 3824311.47 frames. ], batch size: 64, lr: 7.55e-03, grad_scale: 8.0 +2023-04-01 19:45:04,323 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0674, 1.4216, 2.0972, 1.7455, 3.1806, 4.6771, 4.5650, 5.0794], + device='cuda:3'), covar=tensor([0.1630, 0.3234, 0.2750, 0.1812, 0.0436, 0.0140, 0.0141, 0.0106], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0294, 0.0321, 0.0251, 0.0213, 0.0154, 0.0206, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 19:45:30,663 INFO [train.py:903] (3/4) Epoch 11, batch 3950, loss[loss=0.2953, simple_loss=0.3554, pruned_loss=0.1176, over 19680.00 frames. ], tot_loss[loss=0.2347, simple_loss=0.3077, pruned_loss=0.08084, over 3830783.45 frames. ], batch size: 58, lr: 7.55e-03, grad_scale: 8.0 +2023-04-01 19:45:31,939 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 19:46:23,817 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.514e+02 5.760e+02 7.487e+02 9.416e+02 2.541e+03, threshold=1.497e+03, percent-clipped=9.0 +2023-04-01 19:46:34,098 INFO [train.py:903] (3/4) Epoch 11, batch 4000, loss[loss=0.2742, simple_loss=0.33, pruned_loss=0.1092, over 19678.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3097, pruned_loss=0.0821, over 3818250.95 frames. ], batch size: 60, lr: 7.55e-03, grad_scale: 8.0 +2023-04-01 19:46:36,456 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=72281.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:47:22,644 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 19:47:37,789 INFO [train.py:903] (3/4) Epoch 11, batch 4050, loss[loss=0.2514, simple_loss=0.3281, pruned_loss=0.08732, over 19534.00 frames. ], tot_loss[loss=0.2357, simple_loss=0.3088, pruned_loss=0.08129, over 3816326.23 frames. ], batch size: 54, lr: 7.55e-03, grad_scale: 8.0 +2023-04-01 19:47:57,886 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72344.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:48:05,163 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72349.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:48:33,044 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.561e+02 4.924e+02 6.616e+02 8.581e+02 2.031e+03, threshold=1.323e+03, percent-clipped=3.0 +2023-04-01 19:48:37,092 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72374.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:48:43,657 INFO [train.py:903] (3/4) Epoch 11, batch 4100, loss[loss=0.2562, simple_loss=0.3324, pruned_loss=0.08998, over 19305.00 frames. ], tot_loss[loss=0.236, simple_loss=0.3092, pruned_loss=0.08139, over 3806614.10 frames. ], batch size: 66, lr: 7.54e-03, grad_scale: 8.0 +2023-04-01 19:49:05,298 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72396.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:49:18,919 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 19:49:48,308 INFO [train.py:903] (3/4) Epoch 11, batch 4150, loss[loss=0.2986, simple_loss=0.3495, pruned_loss=0.1239, over 13769.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.3071, pruned_loss=0.08009, over 3815109.30 frames. ], batch size: 136, lr: 7.54e-03, grad_scale: 8.0 +2023-04-01 19:50:01,691 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2391, 3.7651, 3.8697, 3.8714, 1.5034, 3.6584, 3.1859, 3.5856], + device='cuda:3'), covar=tensor([0.1448, 0.0791, 0.0629, 0.0690, 0.4885, 0.0687, 0.0661, 0.1171], + device='cuda:3'), in_proj_covar=tensor([0.0665, 0.0601, 0.0793, 0.0674, 0.0723, 0.0550, 0.0485, 0.0733], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 19:50:42,219 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.080e+02 4.902e+02 6.015e+02 8.099e+02 1.569e+03, threshold=1.203e+03, percent-clipped=3.0 +2023-04-01 19:50:42,505 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.8380, 5.1925, 2.9916, 4.5416, 1.2458, 5.1753, 5.1951, 5.3370], + device='cuda:3'), covar=tensor([0.0401, 0.0913, 0.1870, 0.0642, 0.3720, 0.0566, 0.0649, 0.0817], + device='cuda:3'), in_proj_covar=tensor([0.0433, 0.0370, 0.0440, 0.0320, 0.0379, 0.0373, 0.0358, 0.0393], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:50:47,126 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9600, 4.3571, 4.6640, 4.6521, 1.8322, 4.2985, 3.8035, 4.3261], + device='cuda:3'), covar=tensor([0.1390, 0.0842, 0.0515, 0.0563, 0.4872, 0.0704, 0.0576, 0.1023], + device='cuda:3'), in_proj_covar=tensor([0.0668, 0.0602, 0.0795, 0.0676, 0.0724, 0.0552, 0.0485, 0.0734], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 19:50:51,449 INFO [train.py:903] (3/4) Epoch 11, batch 4200, loss[loss=0.2152, simple_loss=0.2848, pruned_loss=0.07277, over 19801.00 frames. ], tot_loss[loss=0.2325, simple_loss=0.3059, pruned_loss=0.07954, over 3811659.63 frames. ], batch size: 48, lr: 7.54e-03, grad_scale: 16.0 +2023-04-01 19:50:57,210 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 19:51:07,242 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=72491.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:51:56,566 INFO [train.py:903] (3/4) Epoch 11, batch 4250, loss[loss=0.2781, simple_loss=0.3416, pruned_loss=0.1073, over 19585.00 frames. ], tot_loss[loss=0.2326, simple_loss=0.3063, pruned_loss=0.0794, over 3829435.89 frames. ], batch size: 61, lr: 7.54e-03, grad_scale: 8.0 +2023-04-01 19:52:17,367 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 19:52:28,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 19:52:33,232 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5791, 1.1297, 1.3435, 1.2221, 2.1893, 0.9653, 1.9586, 2.3341], + device='cuda:3'), covar=tensor([0.0644, 0.2606, 0.2554, 0.1492, 0.0837, 0.1943, 0.0948, 0.0509], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0339, 0.0352, 0.0322, 0.0348, 0.0333, 0.0336, 0.0354], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:52:51,987 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.976e+02 5.433e+02 6.415e+02 7.675e+02 1.468e+03, threshold=1.283e+03, percent-clipped=2.0 +2023-04-01 19:53:01,646 INFO [train.py:903] (3/4) Epoch 11, batch 4300, loss[loss=0.2026, simple_loss=0.2776, pruned_loss=0.06383, over 19752.00 frames. ], tot_loss[loss=0.2317, simple_loss=0.3055, pruned_loss=0.07896, over 3828494.67 frames. ], batch size: 45, lr: 7.53e-03, grad_scale: 8.0 +2023-04-01 19:53:36,425 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72606.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:54:00,227 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 19:54:04,816 INFO [train.py:903] (3/4) Epoch 11, batch 4350, loss[loss=0.2581, simple_loss=0.3242, pruned_loss=0.09598, over 19343.00 frames. ], tot_loss[loss=0.2337, simple_loss=0.3072, pruned_loss=0.08008, over 3823637.91 frames. ], batch size: 66, lr: 7.53e-03, grad_scale: 8.0 +2023-04-01 19:54:34,068 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72652.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:54:58,904 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 5.206e+02 6.524e+02 8.372e+02 1.509e+03, threshold=1.305e+03, percent-clipped=4.0 +2023-04-01 19:55:06,355 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72677.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:55:08,342 INFO [train.py:903] (3/4) Epoch 11, batch 4400, loss[loss=0.2302, simple_loss=0.3084, pruned_loss=0.07604, over 18035.00 frames. ], tot_loss[loss=0.2339, simple_loss=0.3074, pruned_loss=0.08022, over 3815566.80 frames. ], batch size: 83, lr: 7.53e-03, grad_scale: 8.0 +2023-04-01 19:55:20,145 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=72688.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:55:33,215 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6947, 1.3904, 1.5151, 1.4863, 3.1775, 0.9661, 2.2779, 3.6028], + device='cuda:3'), covar=tensor([0.0399, 0.2671, 0.2618, 0.1816, 0.0679, 0.2650, 0.1261, 0.0257], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0337, 0.0351, 0.0321, 0.0344, 0.0331, 0.0335, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:55:37,549 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 19:55:47,769 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 19:56:12,261 INFO [train.py:903] (3/4) Epoch 11, batch 4450, loss[loss=0.2467, simple_loss=0.3242, pruned_loss=0.08463, over 19681.00 frames. ], tot_loss[loss=0.2333, simple_loss=0.3068, pruned_loss=0.07994, over 3817417.84 frames. ], batch size: 59, lr: 7.53e-03, grad_scale: 8.0 +2023-04-01 19:57:06,050 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.681e+02 5.500e+02 6.922e+02 8.863e+02 1.965e+03, threshold=1.384e+03, percent-clipped=10.0 +2023-04-01 19:57:14,277 INFO [train.py:903] (3/4) Epoch 11, batch 4500, loss[loss=0.2303, simple_loss=0.312, pruned_loss=0.07429, over 19431.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.3082, pruned_loss=0.08121, over 3788983.22 frames. ], batch size: 70, lr: 7.52e-03, grad_scale: 8.0 +2023-04-01 19:57:46,499 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72803.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:58:19,535 INFO [train.py:903] (3/4) Epoch 11, batch 4550, loss[loss=0.2246, simple_loss=0.2879, pruned_loss=0.08065, over 19373.00 frames. ], tot_loss[loss=0.2347, simple_loss=0.3075, pruned_loss=0.08101, over 3798305.16 frames. ], batch size: 47, lr: 7.52e-03, grad_scale: 8.0 +2023-04-01 19:58:28,663 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 19:58:33,821 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9105, 4.4131, 3.0115, 3.9639, 1.2602, 4.1946, 4.2584, 4.4321], + device='cuda:3'), covar=tensor([0.0548, 0.1029, 0.1701, 0.0703, 0.3711, 0.0747, 0.0706, 0.0976], + device='cuda:3'), in_proj_covar=tensor([0.0432, 0.0366, 0.0436, 0.0319, 0.0378, 0.0374, 0.0359, 0.0392], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 19:58:53,316 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 19:59:02,258 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72862.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 19:59:11,645 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-01 19:59:15,577 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.061e+02 5.895e+02 7.318e+02 9.053e+02 1.617e+03, threshold=1.464e+03, percent-clipped=1.0 +2023-04-01 19:59:23,898 INFO [train.py:903] (3/4) Epoch 11, batch 4600, loss[loss=0.2251, simple_loss=0.3049, pruned_loss=0.07262, over 19677.00 frames. ], tot_loss[loss=0.2342, simple_loss=0.3074, pruned_loss=0.08052, over 3803772.03 frames. ], batch size: 58, lr: 7.52e-03, grad_scale: 8.0 +2023-04-01 19:59:34,635 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72887.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:00:04,228 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-01 20:00:06,142 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72912.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:00:27,758 INFO [train.py:903] (3/4) Epoch 11, batch 4650, loss[loss=0.2651, simple_loss=0.3369, pruned_loss=0.09665, over 19780.00 frames. ], tot_loss[loss=0.2346, simple_loss=0.308, pruned_loss=0.08067, over 3816021.48 frames. ], batch size: 63, lr: 7.52e-03, grad_scale: 8.0 +2023-04-01 20:00:46,742 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 20:00:57,786 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 20:01:22,036 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.839e+02 5.234e+02 6.593e+02 8.122e+02 1.324e+03, threshold=1.319e+03, percent-clipped=0.0 +2023-04-01 20:01:30,179 INFO [train.py:903] (3/4) Epoch 11, batch 4700, loss[loss=0.2332, simple_loss=0.3115, pruned_loss=0.07745, over 18843.00 frames. ], tot_loss[loss=0.236, simple_loss=0.3087, pruned_loss=0.0816, over 3810050.04 frames. ], batch size: 74, lr: 7.51e-03, grad_scale: 8.0 +2023-04-01 20:01:34,994 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72982.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 20:01:55,752 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 20:02:34,858 INFO [train.py:903] (3/4) Epoch 11, batch 4750, loss[loss=0.2401, simple_loss=0.3202, pruned_loss=0.08001, over 19510.00 frames. ], tot_loss[loss=0.2367, simple_loss=0.3093, pruned_loss=0.08206, over 3807208.88 frames. ], batch size: 64, lr: 7.51e-03, grad_scale: 8.0 +2023-04-01 20:03:12,008 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73059.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:03:28,793 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.570e+02 5.728e+02 6.958e+02 8.518e+02 2.013e+03, threshold=1.392e+03, percent-clipped=2.0 +2023-04-01 20:03:37,835 INFO [train.py:903] (3/4) Epoch 11, batch 4800, loss[loss=0.1839, simple_loss=0.2625, pruned_loss=0.05264, over 19768.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3079, pruned_loss=0.0812, over 3816358.86 frames. ], batch size: 47, lr: 7.51e-03, grad_scale: 8.0 +2023-04-01 20:03:38,480 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 20:03:45,156 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73084.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:04:18,448 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1302, 5.4338, 3.0359, 4.8452, 1.2660, 5.5896, 5.5078, 5.6578], + device='cuda:3'), covar=tensor([0.0364, 0.0783, 0.1677, 0.0541, 0.3595, 0.0468, 0.0594, 0.0662], + device='cuda:3'), in_proj_covar=tensor([0.0430, 0.0364, 0.0433, 0.0315, 0.0376, 0.0370, 0.0356, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:04:26,318 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2122, 1.2361, 1.5894, 1.3857, 2.1873, 1.8523, 2.1608, 0.8831], + device='cuda:3'), covar=tensor([0.2412, 0.4247, 0.2450, 0.1992, 0.1498, 0.2264, 0.1495, 0.4001], + device='cuda:3'), in_proj_covar=tensor([0.0490, 0.0577, 0.0601, 0.0438, 0.0591, 0.0493, 0.0645, 0.0497], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 20:04:40,996 INFO [train.py:903] (3/4) Epoch 11, batch 4850, loss[loss=0.2213, simple_loss=0.3017, pruned_loss=0.0704, over 19535.00 frames. ], tot_loss[loss=0.2347, simple_loss=0.3077, pruned_loss=0.08087, over 3814059.16 frames. ], batch size: 54, lr: 7.51e-03, grad_scale: 8.0 +2023-04-01 20:05:01,955 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 20:05:22,855 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 20:05:29,940 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 20:05:29,972 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 20:05:35,782 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.053e+02 5.592e+02 6.791e+02 8.044e+02 1.982e+03, threshold=1.358e+03, percent-clipped=5.0 +2023-04-01 20:05:39,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 20:05:44,088 INFO [train.py:903] (3/4) Epoch 11, batch 4900, loss[loss=0.2338, simple_loss=0.3225, pruned_loss=0.07253, over 19663.00 frames. ], tot_loss[loss=0.2358, simple_loss=0.3084, pruned_loss=0.08157, over 3812230.42 frames. ], batch size: 59, lr: 7.50e-03, grad_scale: 8.0 +2023-04-01 20:05:57,049 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3445, 1.0688, 1.3754, 1.4783, 2.8173, 1.0551, 2.0385, 3.1424], + device='cuda:3'), covar=tensor([0.0505, 0.3022, 0.2965, 0.1819, 0.0829, 0.2567, 0.1332, 0.0378], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0337, 0.0352, 0.0321, 0.0344, 0.0332, 0.0331, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:06:01,217 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 20:06:48,395 INFO [train.py:903] (3/4) Epoch 11, batch 4950, loss[loss=0.2411, simple_loss=0.3232, pruned_loss=0.07952, over 17314.00 frames. ], tot_loss[loss=0.2364, simple_loss=0.309, pruned_loss=0.08189, over 3815630.65 frames. ], batch size: 101, lr: 7.50e-03, grad_scale: 8.0 +2023-04-01 20:07:00,958 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 20:07:21,299 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=73256.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:07:24,664 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 20:07:42,674 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.726e+02 5.448e+02 7.218e+02 1.051e+03 2.092e+03, threshold=1.444e+03, percent-clipped=5.0 +2023-04-01 20:07:51,012 INFO [train.py:903] (3/4) Epoch 11, batch 5000, loss[loss=0.1952, simple_loss=0.2687, pruned_loss=0.06086, over 19765.00 frames. ], tot_loss[loss=0.2366, simple_loss=0.3093, pruned_loss=0.08199, over 3816147.04 frames. ], batch size: 46, lr: 7.50e-03, grad_scale: 8.0 +2023-04-01 20:07:55,873 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 20:08:08,617 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 20:08:13,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6237, 1.3803, 1.5090, 2.0336, 1.5820, 1.8380, 1.9945, 1.6572], + device='cuda:3'), covar=tensor([0.0868, 0.1050, 0.1042, 0.0795, 0.0898, 0.0777, 0.0851, 0.0703], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0223, 0.0250, 0.0235, 0.0213, 0.0194, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 20:08:32,077 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73311.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:08:50,645 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=73326.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 20:08:54,854 INFO [train.py:903] (3/4) Epoch 11, batch 5050, loss[loss=0.216, simple_loss=0.288, pruned_loss=0.07204, over 19494.00 frames. ], tot_loss[loss=0.2349, simple_loss=0.3076, pruned_loss=0.08114, over 3831566.51 frames. ], batch size: 49, lr: 7.50e-03, grad_scale: 8.0 +2023-04-01 20:09:28,954 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 20:09:47,530 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73371.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:09:48,284 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.996e+02 5.337e+02 6.487e+02 8.820e+02 1.986e+03, threshold=1.297e+03, percent-clipped=3.0 +2023-04-01 20:09:56,474 INFO [train.py:903] (3/4) Epoch 11, batch 5100, loss[loss=0.2421, simple_loss=0.3097, pruned_loss=0.0872, over 19053.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.3081, pruned_loss=0.08139, over 3828967.42 frames. ], batch size: 69, lr: 7.49e-03, grad_scale: 8.0 +2023-04-01 20:10:04,910 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 20:10:10,270 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 20:10:13,776 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 20:10:22,146 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2342, 5.5420, 3.0786, 4.8779, 1.6187, 5.5765, 5.5244, 5.7518], + device='cuda:3'), covar=tensor([0.0436, 0.0981, 0.1934, 0.0575, 0.3651, 0.0577, 0.0634, 0.0705], + device='cuda:3'), in_proj_covar=tensor([0.0432, 0.0363, 0.0437, 0.0317, 0.0379, 0.0373, 0.0357, 0.0390], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:11:00,244 INFO [train.py:903] (3/4) Epoch 11, batch 5150, loss[loss=0.2435, simple_loss=0.3206, pruned_loss=0.08321, over 19693.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3087, pruned_loss=0.08156, over 3809654.14 frames. ], batch size: 59, lr: 7.49e-03, grad_scale: 8.0 +2023-04-01 20:11:09,675 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 20:11:16,054 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73441.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 20:11:31,599 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5517, 1.3549, 1.3886, 1.2762, 3.1050, 0.9346, 2.1413, 3.4404], + device='cuda:3'), covar=tensor([0.0467, 0.2450, 0.2719, 0.1907, 0.0714, 0.2562, 0.1213, 0.0280], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0334, 0.0348, 0.0320, 0.0344, 0.0331, 0.0328, 0.0350], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:11:43,042 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 20:11:50,721 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-01 20:11:54,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.282e+02 5.910e+02 7.077e+02 9.438e+02 2.777e+03, threshold=1.415e+03, percent-clipped=4.0 +2023-04-01 20:12:04,004 INFO [train.py:903] (3/4) Epoch 11, batch 5200, loss[loss=0.2373, simple_loss=0.3149, pruned_loss=0.07988, over 19609.00 frames. ], tot_loss[loss=0.237, simple_loss=0.3099, pruned_loss=0.08204, over 3807996.05 frames. ], batch size: 57, lr: 7.49e-03, grad_scale: 8.0 +2023-04-01 20:12:16,631 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 20:12:29,016 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3354, 2.3041, 1.7666, 1.5976, 2.2666, 1.2027, 1.1910, 1.8603], + device='cuda:3'), covar=tensor([0.0952, 0.0604, 0.0951, 0.0677, 0.0365, 0.1118, 0.0804, 0.0402], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0304, 0.0332, 0.0251, 0.0235, 0.0324, 0.0290, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:13:00,027 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 20:13:07,097 INFO [train.py:903] (3/4) Epoch 11, batch 5250, loss[loss=0.2111, simple_loss=0.2925, pruned_loss=0.06491, over 19690.00 frames. ], tot_loss[loss=0.236, simple_loss=0.3093, pruned_loss=0.08134, over 3798850.00 frames. ], batch size: 59, lr: 7.49e-03, grad_scale: 8.0 +2023-04-01 20:14:02,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.872e+02 5.338e+02 6.658e+02 8.757e+02 1.866e+03, threshold=1.332e+03, percent-clipped=3.0 +2023-04-01 20:14:11,099 INFO [train.py:903] (3/4) Epoch 11, batch 5300, loss[loss=0.2648, simple_loss=0.3337, pruned_loss=0.09789, over 19666.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3086, pruned_loss=0.08131, over 3816277.85 frames. ], batch size: 60, lr: 7.48e-03, grad_scale: 8.0 +2023-04-01 20:14:26,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 20:14:49,242 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73608.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:14:50,946 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-01 20:15:13,596 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73627.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:15:15,330 INFO [train.py:903] (3/4) Epoch 11, batch 5350, loss[loss=0.3298, simple_loss=0.3722, pruned_loss=0.1437, over 13118.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3084, pruned_loss=0.08106, over 3811875.06 frames. ], batch size: 135, lr: 7.48e-03, grad_scale: 8.0 +2023-04-01 20:15:46,116 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73652.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:15:46,874 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 20:15:49,349 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=73655.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:16:09,903 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.174e+02 5.285e+02 6.573e+02 8.256e+02 1.333e+03, threshold=1.315e+03, percent-clipped=1.0 +2023-04-01 20:16:19,917 INFO [train.py:903] (3/4) Epoch 11, batch 5400, loss[loss=0.2823, simple_loss=0.3367, pruned_loss=0.1139, over 13456.00 frames. ], tot_loss[loss=0.2341, simple_loss=0.3079, pruned_loss=0.08018, over 3810585.04 frames. ], batch size: 136, lr: 7.48e-03, grad_scale: 8.0 +2023-04-01 20:16:41,417 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73697.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 20:17:13,034 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73722.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 20:17:20,604 INFO [train.py:903] (3/4) Epoch 11, batch 5450, loss[loss=0.23, simple_loss=0.3093, pruned_loss=0.07541, over 19779.00 frames. ], tot_loss[loss=0.2342, simple_loss=0.3079, pruned_loss=0.08025, over 3830188.20 frames. ], batch size: 56, lr: 7.48e-03, grad_scale: 8.0 +2023-04-01 20:17:45,902 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:18:12,070 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73770.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:18:14,003 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.246e+02 5.707e+02 7.323e+02 9.356e+02 2.024e+03, threshold=1.465e+03, percent-clipped=8.0 +2023-04-01 20:18:23,353 INFO [train.py:903] (3/4) Epoch 11, batch 5500, loss[loss=0.2568, simple_loss=0.33, pruned_loss=0.09186, over 19572.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3083, pruned_loss=0.08104, over 3831499.55 frames. ], batch size: 61, lr: 7.47e-03, grad_scale: 8.0 +2023-04-01 20:18:50,412 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 20:19:25,475 INFO [train.py:903] (3/4) Epoch 11, batch 5550, loss[loss=0.2436, simple_loss=0.3166, pruned_loss=0.08531, over 19666.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3085, pruned_loss=0.08132, over 3837108.09 frames. ], batch size: 58, lr: 7.47e-03, grad_scale: 8.0 +2023-04-01 20:19:35,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 20:20:04,784 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1256, 1.2577, 1.6816, 0.9634, 2.3236, 2.9089, 2.6792, 3.1115], + device='cuda:3'), covar=tensor([0.1562, 0.3397, 0.2909, 0.2332, 0.0543, 0.0282, 0.0266, 0.0260], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0292, 0.0321, 0.0248, 0.0212, 0.0156, 0.0205, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 20:20:20,661 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.349e+02 5.761e+02 6.752e+02 8.536e+02 1.570e+03, threshold=1.350e+03, percent-clipped=1.0 +2023-04-01 20:20:26,366 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 20:20:29,863 INFO [train.py:903] (3/4) Epoch 11, batch 5600, loss[loss=0.2941, simple_loss=0.3566, pruned_loss=0.1158, over 13122.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3088, pruned_loss=0.08147, over 3809441.77 frames. ], batch size: 136, lr: 7.47e-03, grad_scale: 8.0 +2023-04-01 20:21:03,558 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-01 20:21:34,086 INFO [train.py:903] (3/4) Epoch 11, batch 5650, loss[loss=0.2507, simple_loss=0.3337, pruned_loss=0.08392, over 19591.00 frames. ], tot_loss[loss=0.2356, simple_loss=0.3087, pruned_loss=0.08128, over 3825253.59 frames. ], batch size: 61, lr: 7.47e-03, grad_scale: 8.0 +2023-04-01 20:21:48,845 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5694, 1.2130, 1.2200, 1.4185, 1.0521, 1.3736, 1.2356, 1.3664], + device='cuda:3'), covar=tensor([0.0998, 0.1175, 0.1439, 0.0930, 0.1189, 0.0554, 0.1281, 0.0821], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0351, 0.0289, 0.0239, 0.0295, 0.0244, 0.0279, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:21:50,105 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3397, 1.3939, 1.7988, 1.6426, 2.9711, 2.4753, 3.3404, 1.3704], + device='cuda:3'), covar=tensor([0.2261, 0.3832, 0.2393, 0.1682, 0.1605, 0.1922, 0.1732, 0.3663], + device='cuda:3'), in_proj_covar=tensor([0.0494, 0.0578, 0.0603, 0.0439, 0.0597, 0.0494, 0.0648, 0.0496], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 20:21:59,920 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.59 vs. limit=2.0 +2023-04-01 20:22:03,008 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=73952.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:22:23,885 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 20:22:28,535 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.512e+02 5.880e+02 7.006e+02 8.632e+02 1.687e+03, threshold=1.401e+03, percent-clipped=2.0 +2023-04-01 20:22:37,797 INFO [train.py:903] (3/4) Epoch 11, batch 5700, loss[loss=0.2493, simple_loss=0.3308, pruned_loss=0.08395, over 19532.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.309, pruned_loss=0.08103, over 3823179.92 frames. ], batch size: 56, lr: 7.46e-03, grad_scale: 8.0 +2023-04-01 20:23:38,871 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74026.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:23:42,022 INFO [train.py:903] (3/4) Epoch 11, batch 5750, loss[loss=0.227, simple_loss=0.2944, pruned_loss=0.07981, over 19328.00 frames. ], tot_loss[loss=0.2344, simple_loss=0.3079, pruned_loss=0.08047, over 3825280.55 frames. ], batch size: 47, lr: 7.46e-03, grad_scale: 8.0 +2023-04-01 20:23:43,290 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 20:23:51,340 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 20:23:56,806 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 20:24:08,305 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-04-01 20:24:11,336 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74051.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:24:17,091 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.39 vs. limit=5.0 +2023-04-01 20:24:30,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74067.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:24:36,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.134e+02 5.413e+02 6.547e+02 7.617e+02 1.298e+03, threshold=1.309e+03, percent-clipped=0.0 +2023-04-01 20:24:37,154 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.74 vs. limit=2.0 +2023-04-01 20:24:45,231 INFO [train.py:903] (3/4) Epoch 11, batch 5800, loss[loss=0.2412, simple_loss=0.3022, pruned_loss=0.09014, over 19407.00 frames. ], tot_loss[loss=0.2335, simple_loss=0.3076, pruned_loss=0.07971, over 3832716.77 frames. ], batch size: 48, lr: 7.46e-03, grad_scale: 8.0 +2023-04-01 20:24:50,607 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74082.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:25:04,163 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=74094.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:25:48,486 INFO [train.py:903] (3/4) Epoch 11, batch 5850, loss[loss=0.231, simple_loss=0.3068, pruned_loss=0.07764, over 19605.00 frames. ], tot_loss[loss=0.2347, simple_loss=0.3086, pruned_loss=0.08042, over 3841282.25 frames. ], batch size: 57, lr: 7.46e-03, grad_scale: 8.0 +2023-04-01 20:25:55,540 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9064, 1.5223, 1.5751, 2.2089, 1.7240, 2.1309, 2.1265, 2.0439], + device='cuda:3'), covar=tensor([0.0738, 0.0970, 0.0976, 0.0791, 0.0879, 0.0638, 0.0880, 0.0578], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0230, 0.0225, 0.0248, 0.0237, 0.0214, 0.0195, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 20:26:42,000 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.778e+02 5.734e+02 7.251e+02 9.100e+02 2.751e+03, threshold=1.450e+03, percent-clipped=7.0 +2023-04-01 20:26:51,311 INFO [train.py:903] (3/4) Epoch 11, batch 5900, loss[loss=0.201, simple_loss=0.2737, pruned_loss=0.0642, over 19295.00 frames. ], tot_loss[loss=0.2363, simple_loss=0.3095, pruned_loss=0.0815, over 3830255.06 frames. ], batch size: 44, lr: 7.45e-03, grad_scale: 8.0 +2023-04-01 20:26:53,560 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 20:27:12,262 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.94 vs. limit=5.0 +2023-04-01 20:27:15,048 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 20:27:30,897 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74209.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:27:55,168 INFO [train.py:903] (3/4) Epoch 11, batch 5950, loss[loss=0.2505, simple_loss=0.3207, pruned_loss=0.09012, over 13501.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.3088, pruned_loss=0.08112, over 3811840.44 frames. ], batch size: 136, lr: 7.45e-03, grad_scale: 8.0 +2023-04-01 20:28:33,398 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74258.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:28:49,684 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.166e+02 4.883e+02 6.160e+02 7.607e+02 1.521e+03, threshold=1.232e+03, percent-clipped=1.0 +2023-04-01 20:28:59,177 INFO [train.py:903] (3/4) Epoch 11, batch 6000, loss[loss=0.2143, simple_loss=0.2926, pruned_loss=0.06798, over 19680.00 frames. ], tot_loss[loss=0.2338, simple_loss=0.3076, pruned_loss=0.07998, over 3825993.16 frames. ], batch size: 53, lr: 7.45e-03, grad_scale: 8.0 +2023-04-01 20:28:59,177 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 20:29:11,904 INFO [train.py:937] (3/4) Epoch 11, validation: loss=0.1778, simple_loss=0.2787, pruned_loss=0.03847, over 944034.00 frames. +2023-04-01 20:29:11,906 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 20:30:09,794 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74323.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:30:18,373 INFO [train.py:903] (3/4) Epoch 11, batch 6050, loss[loss=0.204, simple_loss=0.2837, pruned_loss=0.0622, over 19613.00 frames. ], tot_loss[loss=0.2332, simple_loss=0.3072, pruned_loss=0.07964, over 3814500.28 frames. ], batch size: 50, lr: 7.45e-03, grad_scale: 8.0 +2023-04-01 20:30:42,852 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74348.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:31:12,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.290e+02 5.870e+02 7.295e+02 8.913e+02 1.728e+03, threshold=1.459e+03, percent-clipped=8.0 +2023-04-01 20:31:21,969 INFO [train.py:903] (3/4) Epoch 11, batch 6100, loss[loss=0.1858, simple_loss=0.2639, pruned_loss=0.05388, over 19734.00 frames. ], tot_loss[loss=0.2331, simple_loss=0.3071, pruned_loss=0.07954, over 3816190.56 frames. ], batch size: 47, lr: 7.44e-03, grad_scale: 8.0 +2023-04-01 20:32:22,432 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=74426.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:32:25,901 INFO [train.py:903] (3/4) Epoch 11, batch 6150, loss[loss=0.1984, simple_loss=0.2774, pruned_loss=0.05975, over 19592.00 frames. ], tot_loss[loss=0.2335, simple_loss=0.3071, pruned_loss=0.07996, over 3814888.58 frames. ], batch size: 52, lr: 7.44e-03, grad_scale: 8.0 +2023-04-01 20:32:37,704 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8828, 1.9879, 2.1352, 2.6839, 1.8299, 2.5230, 2.3611, 2.0511], + device='cuda:3'), covar=tensor([0.3429, 0.2840, 0.1422, 0.1871, 0.3276, 0.1437, 0.3420, 0.2522], + device='cuda:3'), in_proj_covar=tensor([0.0790, 0.0811, 0.0644, 0.0894, 0.0779, 0.0699, 0.0783, 0.0708], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 20:32:54,707 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 20:33:12,618 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74465.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:33:20,927 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.809e+02 5.576e+02 6.382e+02 8.608e+02 2.367e+03, threshold=1.276e+03, percent-clipped=4.0 +2023-04-01 20:33:29,315 INFO [train.py:903] (3/4) Epoch 11, batch 6200, loss[loss=0.2664, simple_loss=0.3327, pruned_loss=0.1, over 19778.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.3082, pruned_loss=0.08125, over 3823765.87 frames. ], batch size: 54, lr: 7.44e-03, grad_scale: 8.0 +2023-04-01 20:33:43,962 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74490.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:34:31,909 INFO [train.py:903] (3/4) Epoch 11, batch 6250, loss[loss=0.234, simple_loss=0.3113, pruned_loss=0.07835, over 19824.00 frames. ], tot_loss[loss=0.2344, simple_loss=0.3077, pruned_loss=0.08054, over 3829893.00 frames. ], batch size: 49, lr: 7.44e-03, grad_scale: 8.0 +2023-04-01 20:34:48,642 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74541.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:35:02,958 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 20:35:30,746 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.915e+02 5.334e+02 6.699e+02 8.918e+02 1.691e+03, threshold=1.340e+03, percent-clipped=6.0 +2023-04-01 20:35:38,652 INFO [train.py:903] (3/4) Epoch 11, batch 6300, loss[loss=0.2689, simple_loss=0.3302, pruned_loss=0.1038, over 19680.00 frames. ], tot_loss[loss=0.2329, simple_loss=0.3065, pruned_loss=0.07961, over 3824881.62 frames. ], batch size: 60, lr: 7.43e-03, grad_scale: 8.0 +2023-04-01 20:36:06,824 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=74602.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:36:21,252 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74613.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:36:40,495 INFO [train.py:903] (3/4) Epoch 11, batch 6350, loss[loss=0.3293, simple_loss=0.3801, pruned_loss=0.1392, over 19665.00 frames. ], tot_loss[loss=0.2334, simple_loss=0.3069, pruned_loss=0.07996, over 3838217.92 frames. ], batch size: 58, lr: 7.43e-03, grad_scale: 8.0 +2023-04-01 20:37:36,496 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.847e+02 5.622e+02 6.880e+02 9.256e+02 3.776e+03, threshold=1.376e+03, percent-clipped=5.0 +2023-04-01 20:37:44,833 INFO [train.py:903] (3/4) Epoch 11, batch 6400, loss[loss=0.2263, simple_loss=0.2958, pruned_loss=0.07844, over 19628.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.307, pruned_loss=0.08009, over 3831037.79 frames. ], batch size: 50, lr: 7.43e-03, grad_scale: 8.0 +2023-04-01 20:37:56,012 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 20:38:02,527 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6443, 2.2278, 2.2060, 2.8228, 2.7293, 2.1904, 2.2511, 2.7713], + device='cuda:3'), covar=tensor([0.0783, 0.1563, 0.1252, 0.0903, 0.1078, 0.0483, 0.1069, 0.0539], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0351, 0.0291, 0.0238, 0.0298, 0.0243, 0.0278, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:38:33,504 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74717.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:38:46,864 INFO [train.py:903] (3/4) Epoch 11, batch 6450, loss[loss=0.2529, simple_loss=0.3305, pruned_loss=0.08761, over 19467.00 frames. ], tot_loss[loss=0.2349, simple_loss=0.3082, pruned_loss=0.08083, over 3826239.64 frames. ], batch size: 64, lr: 7.43e-03, grad_scale: 8.0 +2023-04-01 20:39:25,311 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1685, 1.8392, 1.7471, 2.1388, 2.0927, 1.8905, 1.7201, 2.1578], + device='cuda:3'), covar=tensor([0.0881, 0.1556, 0.1391, 0.0931, 0.1186, 0.0475, 0.1186, 0.0639], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0351, 0.0293, 0.0239, 0.0298, 0.0244, 0.0279, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:39:28,492 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 20:39:44,916 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.542e+02 5.590e+02 7.165e+02 9.175e+02 2.194e+03, threshold=1.433e+03, percent-clipped=7.0 +2023-04-01 20:39:53,331 INFO [train.py:903] (3/4) Epoch 11, batch 6500, loss[loss=0.1974, simple_loss=0.2822, pruned_loss=0.05631, over 19696.00 frames. ], tot_loss[loss=0.2339, simple_loss=0.3073, pruned_loss=0.08019, over 3829175.25 frames. ], batch size: 58, lr: 7.42e-03, grad_scale: 8.0 +2023-04-01 20:39:54,543 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 20:40:16,083 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74797.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:40:48,464 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74822.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:40:56,368 INFO [train.py:903] (3/4) Epoch 11, batch 6550, loss[loss=0.2192, simple_loss=0.29, pruned_loss=0.07422, over 19730.00 frames. ], tot_loss[loss=0.2359, simple_loss=0.3092, pruned_loss=0.08133, over 3813672.47 frames. ], batch size: 51, lr: 7.42e-03, grad_scale: 8.0 +2023-04-01 20:40:56,801 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7807, 1.7573, 1.5382, 1.3405, 1.3279, 1.3730, 0.1488, 0.6406], + device='cuda:3'), covar=tensor([0.0429, 0.0416, 0.0268, 0.0428, 0.0843, 0.0476, 0.0818, 0.0758], + device='cuda:3'), in_proj_covar=tensor([0.0341, 0.0331, 0.0333, 0.0354, 0.0429, 0.0353, 0.0312, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 20:41:52,115 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.386e+02 5.036e+02 6.201e+02 8.020e+02 1.527e+03, threshold=1.240e+03, percent-clipped=1.0 +2023-04-01 20:41:59,119 INFO [train.py:903] (3/4) Epoch 11, batch 6600, loss[loss=0.2264, simple_loss=0.309, pruned_loss=0.07189, over 19621.00 frames. ], tot_loss[loss=0.2367, simple_loss=0.31, pruned_loss=0.08168, over 3814526.61 frames. ], batch size: 57, lr: 7.42e-03, grad_scale: 8.0 +2023-04-01 20:42:42,902 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-01 20:43:01,253 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74927.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:43:03,149 INFO [train.py:903] (3/4) Epoch 11, batch 6650, loss[loss=0.211, simple_loss=0.2921, pruned_loss=0.06495, over 19762.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3082, pruned_loss=0.08112, over 3818543.02 frames. ], batch size: 54, lr: 7.42e-03, grad_scale: 8.0 +2023-04-01 20:43:40,282 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=74957.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:44:00,890 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.349e+02 5.506e+02 7.046e+02 9.043e+02 1.602e+03, threshold=1.409e+03, percent-clipped=2.0 +2023-04-01 20:44:01,364 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74973.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:44:07,813 INFO [train.py:903] (3/4) Epoch 11, batch 6700, loss[loss=0.2443, simple_loss=0.3181, pruned_loss=0.08532, over 18829.00 frames. ], tot_loss[loss=0.2357, simple_loss=0.3088, pruned_loss=0.08128, over 3817522.41 frames. ], batch size: 74, lr: 7.41e-03, grad_scale: 8.0 +2023-04-01 20:44:12,478 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74982.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:44:29,486 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4161, 1.9441, 2.0311, 2.9171, 2.2164, 2.6797, 2.6387, 2.5528], + device='cuda:3'), covar=tensor([0.0692, 0.0928, 0.0910, 0.0737, 0.0820, 0.0624, 0.0837, 0.0567], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0227, 0.0224, 0.0247, 0.0237, 0.0214, 0.0197, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 20:44:31,738 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74998.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:44:46,277 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3160, 3.9116, 2.6635, 3.4459, 0.9899, 3.6891, 3.6836, 3.8069], + device='cuda:3'), covar=tensor([0.0720, 0.1075, 0.1903, 0.0845, 0.3911, 0.0767, 0.0761, 0.1234], + device='cuda:3'), in_proj_covar=tensor([0.0437, 0.0369, 0.0438, 0.0317, 0.0379, 0.0369, 0.0357, 0.0393], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:45:06,536 INFO [train.py:903] (3/4) Epoch 11, batch 6750, loss[loss=0.2017, simple_loss=0.2728, pruned_loss=0.06537, over 19721.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.3083, pruned_loss=0.08115, over 3826439.85 frames. ], batch size: 51, lr: 7.41e-03, grad_scale: 8.0 +2023-04-01 20:45:55,399 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75072.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:45:56,179 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.565e+02 5.649e+02 7.002e+02 8.709e+02 1.691e+03, threshold=1.400e+03, percent-clipped=2.0 +2023-04-01 20:46:04,203 INFO [train.py:903] (3/4) Epoch 11, batch 6800, loss[loss=0.2313, simple_loss=0.3119, pruned_loss=0.07533, over 19584.00 frames. ], tot_loss[loss=0.2346, simple_loss=0.3074, pruned_loss=0.08087, over 3826836.77 frames. ], batch size: 57, lr: 7.41e-03, grad_scale: 8.0 +2023-04-01 20:46:51,849 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 20:46:52,893 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 20:46:56,120 INFO [train.py:903] (3/4) Epoch 12, batch 0, loss[loss=0.2547, simple_loss=0.3075, pruned_loss=0.1009, over 19762.00 frames. ], tot_loss[loss=0.2547, simple_loss=0.3075, pruned_loss=0.1009, over 19762.00 frames. ], batch size: 46, lr: 7.10e-03, grad_scale: 8.0 +2023-04-01 20:46:56,121 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 20:47:04,905 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7784, 3.3635, 2.7148, 3.1651, 0.8568, 3.1660, 3.1678, 3.4948], + device='cuda:3'), covar=tensor([0.0842, 0.0878, 0.1907, 0.0967, 0.4247, 0.1174, 0.0875, 0.1105], + device='cuda:3'), in_proj_covar=tensor([0.0430, 0.0361, 0.0432, 0.0312, 0.0374, 0.0364, 0.0352, 0.0385], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:3') +2023-04-01 20:47:08,129 INFO [train.py:937] (3/4) Epoch 12, validation: loss=0.1777, simple_loss=0.2788, pruned_loss=0.03825, over 944034.00 frames. +2023-04-01 20:47:08,129 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 20:47:20,781 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 20:48:11,519 INFO [train.py:903] (3/4) Epoch 12, batch 50, loss[loss=0.2447, simple_loss=0.3192, pruned_loss=0.08512, over 19758.00 frames. ], tot_loss[loss=0.235, simple_loss=0.3081, pruned_loss=0.08092, over 859389.28 frames. ], batch size: 63, lr: 7.09e-03, grad_scale: 8.0 +2023-04-01 20:48:29,486 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.226e+02 5.251e+02 6.823e+02 1.011e+03 3.055e+03, threshold=1.365e+03, percent-clipped=9.0 +2023-04-01 20:48:42,837 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 20:48:54,694 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 20:49:13,862 INFO [train.py:903] (3/4) Epoch 12, batch 100, loss[loss=0.2056, simple_loss=0.2916, pruned_loss=0.05981, over 19686.00 frames. ], tot_loss[loss=0.2339, simple_loss=0.3074, pruned_loss=0.08015, over 1514333.49 frames. ], batch size: 58, lr: 7.09e-03, grad_scale: 8.0 +2023-04-01 20:49:22,068 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 20:50:02,890 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75246.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:50:16,769 INFO [train.py:903] (3/4) Epoch 12, batch 150, loss[loss=0.2598, simple_loss=0.3341, pruned_loss=0.09274, over 19673.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3043, pruned_loss=0.07852, over 2033183.40 frames. ], batch size: 58, lr: 7.09e-03, grad_scale: 8.0 +2023-04-01 20:50:33,998 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=75271.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:50:36,119 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.416e+02 5.206e+02 6.373e+02 8.343e+02 1.576e+03, threshold=1.275e+03, percent-clipped=5.0 +2023-04-01 20:50:47,873 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0994, 1.7671, 1.7700, 2.1023, 2.0920, 1.9757, 1.6273, 2.0621], + device='cuda:3'), covar=tensor([0.0921, 0.1597, 0.1340, 0.0932, 0.1124, 0.0457, 0.1215, 0.0657], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0350, 0.0288, 0.0239, 0.0293, 0.0242, 0.0276, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:50:56,167 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75288.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:51:15,337 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 20:51:18,701 INFO [train.py:903] (3/4) Epoch 12, batch 200, loss[loss=0.2323, simple_loss=0.3117, pruned_loss=0.07646, over 19543.00 frames. ], tot_loss[loss=0.2311, simple_loss=0.3043, pruned_loss=0.07897, over 2428017.31 frames. ], batch size: 56, lr: 7.09e-03, grad_scale: 8.0 +2023-04-01 20:51:42,624 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=75326.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:51:45,294 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75328.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:52:02,495 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3640, 2.1398, 1.9697, 1.8563, 1.5555, 1.7136, 0.4954, 1.1907], + device='cuda:3'), covar=tensor([0.0371, 0.0413, 0.0323, 0.0536, 0.0827, 0.0646, 0.0867, 0.0723], + device='cuda:3'), in_proj_covar=tensor([0.0341, 0.0331, 0.0330, 0.0357, 0.0430, 0.0353, 0.0309, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 20:52:16,217 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75353.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:52:22,162 INFO [train.py:903] (3/4) Epoch 12, batch 250, loss[loss=0.229, simple_loss=0.3034, pruned_loss=0.07728, over 19720.00 frames. ], tot_loss[loss=0.2332, simple_loss=0.3067, pruned_loss=0.07988, over 2741508.68 frames. ], batch size: 51, lr: 7.08e-03, grad_scale: 8.0 +2023-04-01 20:52:41,713 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.835e+02 5.714e+02 6.835e+02 8.470e+02 1.829e+03, threshold=1.367e+03, percent-clipped=2.0 +2023-04-01 20:52:56,688 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75386.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:53:25,602 INFO [train.py:903] (3/4) Epoch 12, batch 300, loss[loss=0.1905, simple_loss=0.2646, pruned_loss=0.05815, over 19758.00 frames. ], tot_loss[loss=0.2331, simple_loss=0.3063, pruned_loss=0.07995, over 2996669.23 frames. ], batch size: 46, lr: 7.08e-03, grad_scale: 8.0 +2023-04-01 20:54:07,363 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75441.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:54:28,450 INFO [train.py:903] (3/4) Epoch 12, batch 350, loss[loss=0.2529, simple_loss=0.3264, pruned_loss=0.08968, over 19318.00 frames. ], tot_loss[loss=0.2326, simple_loss=0.3064, pruned_loss=0.07942, over 3184738.64 frames. ], batch size: 66, lr: 7.08e-03, grad_scale: 8.0 +2023-04-01 20:54:31,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 20:54:45,903 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.322e+02 5.463e+02 6.814e+02 8.627e+02 1.955e+03, threshold=1.363e+03, percent-clipped=3.0 +2023-04-01 20:55:30,817 INFO [train.py:903] (3/4) Epoch 12, batch 400, loss[loss=0.1838, simple_loss=0.264, pruned_loss=0.05178, over 19383.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3047, pruned_loss=0.07836, over 3338608.82 frames. ], batch size: 47, lr: 7.08e-03, grad_scale: 8.0 +2023-04-01 20:56:16,875 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2664, 1.4222, 1.6773, 1.5100, 2.6367, 2.2864, 2.8434, 1.1674], + device='cuda:3'), covar=tensor([0.2217, 0.3660, 0.2182, 0.1711, 0.1389, 0.1731, 0.1449, 0.3564], + device='cuda:3'), in_proj_covar=tensor([0.0492, 0.0577, 0.0604, 0.0438, 0.0598, 0.0493, 0.0646, 0.0492], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 20:56:32,374 INFO [train.py:903] (3/4) Epoch 12, batch 450, loss[loss=0.2487, simple_loss=0.3241, pruned_loss=0.0867, over 17545.00 frames. ], tot_loss[loss=0.2324, simple_loss=0.3065, pruned_loss=0.07912, over 3450916.59 frames. ], batch size: 101, lr: 7.07e-03, grad_scale: 8.0 +2023-04-01 20:56:51,546 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.480e+02 5.172e+02 6.423e+02 7.976e+02 2.291e+03, threshold=1.285e+03, percent-clipped=4.0 +2023-04-01 20:57:09,484 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 20:57:10,707 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 20:57:13,237 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=75590.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:57:17,035 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2499, 3.7830, 3.8673, 3.8660, 1.3979, 3.6205, 3.1548, 3.5713], + device='cuda:3'), covar=tensor([0.1372, 0.0910, 0.0653, 0.0653, 0.5243, 0.0766, 0.0712, 0.1155], + device='cuda:3'), in_proj_covar=tensor([0.0677, 0.0603, 0.0804, 0.0688, 0.0731, 0.0559, 0.0492, 0.0741], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 20:57:36,550 INFO [train.py:903] (3/4) Epoch 12, batch 500, loss[loss=0.2209, simple_loss=0.3081, pruned_loss=0.06687, over 19748.00 frames. ], tot_loss[loss=0.2305, simple_loss=0.3045, pruned_loss=0.07826, over 3546757.30 frames. ], batch size: 63, lr: 7.07e-03, grad_scale: 8.0 +2023-04-01 20:58:06,133 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=75632.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:58:18,028 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75642.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:58:38,553 INFO [train.py:903] (3/4) Epoch 12, batch 550, loss[loss=0.2231, simple_loss=0.2961, pruned_loss=0.07504, over 19590.00 frames. ], tot_loss[loss=0.2318, simple_loss=0.3056, pruned_loss=0.079, over 3617982.71 frames. ], batch size: 52, lr: 7.07e-03, grad_scale: 8.0 +2023-04-01 20:58:50,380 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75667.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:58:56,744 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.239e+02 5.816e+02 6.939e+02 9.327e+02 2.224e+03, threshold=1.388e+03, percent-clipped=13.0 +2023-04-01 20:59:09,128 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-01 20:59:24,533 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8371, 1.3213, 1.5077, 1.7086, 3.3374, 1.1543, 2.2113, 3.7324], + device='cuda:3'), covar=tensor([0.0426, 0.2659, 0.2697, 0.1697, 0.0745, 0.2506, 0.1396, 0.0243], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0340, 0.0351, 0.0319, 0.0347, 0.0330, 0.0333, 0.0354], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 20:59:26,890 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75697.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:59:38,265 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75705.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 20:59:41,424 INFO [train.py:903] (3/4) Epoch 12, batch 600, loss[loss=0.234, simple_loss=0.3068, pruned_loss=0.08061, over 19465.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3049, pruned_loss=0.07843, over 3668942.94 frames. ], batch size: 49, lr: 7.07e-03, grad_scale: 8.0 +2023-04-01 20:59:57,556 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75722.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:00:22,877 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 21:00:30,160 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75747.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:00:42,504 INFO [train.py:903] (3/4) Epoch 12, batch 650, loss[loss=0.2051, simple_loss=0.2878, pruned_loss=0.06119, over 19695.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3054, pruned_loss=0.07888, over 3704918.72 frames. ], batch size: 59, lr: 7.07e-03, grad_scale: 8.0 +2023-04-01 21:01:01,250 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.249e+02 5.224e+02 6.354e+02 7.929e+02 1.382e+03, threshold=1.271e+03, percent-clipped=0.0 +2023-04-01 21:01:22,981 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3512, 2.1324, 1.5319, 1.3806, 1.9729, 1.1219, 1.1565, 1.7945], + device='cuda:3'), covar=tensor([0.0866, 0.0643, 0.0959, 0.0658, 0.0422, 0.1143, 0.0739, 0.0403], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0304, 0.0329, 0.0249, 0.0237, 0.0320, 0.0292, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:01:24,570 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.56 vs. limit=5.0 +2023-04-01 21:01:29,736 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1397, 1.2494, 1.7332, 1.1717, 2.7669, 3.6800, 3.4285, 3.8515], + device='cuda:3'), covar=tensor([0.1548, 0.3502, 0.3038, 0.2119, 0.0471, 0.0141, 0.0193, 0.0192], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0293, 0.0323, 0.0250, 0.0214, 0.0156, 0.0206, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:01:45,574 INFO [train.py:903] (3/4) Epoch 12, batch 700, loss[loss=0.2415, simple_loss=0.3184, pruned_loss=0.08233, over 19786.00 frames. ], tot_loss[loss=0.2319, simple_loss=0.3058, pruned_loss=0.07901, over 3736030.44 frames. ], batch size: 56, lr: 7.06e-03, grad_scale: 8.0 +2023-04-01 21:02:04,533 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.90 vs. limit=5.0 +2023-04-01 21:02:46,461 INFO [train.py:903] (3/4) Epoch 12, batch 750, loss[loss=0.226, simple_loss=0.3098, pruned_loss=0.07107, over 19663.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3061, pruned_loss=0.07896, over 3755689.91 frames. ], batch size: 55, lr: 7.06e-03, grad_scale: 8.0 +2023-04-01 21:03:05,206 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.592e+02 5.547e+02 6.814e+02 8.571e+02 2.504e+03, threshold=1.363e+03, percent-clipped=8.0 +2023-04-01 21:03:49,642 INFO [train.py:903] (3/4) Epoch 12, batch 800, loss[loss=0.2279, simple_loss=0.3076, pruned_loss=0.07409, over 19771.00 frames. ], tot_loss[loss=0.2296, simple_loss=0.3039, pruned_loss=0.07761, over 3775786.05 frames. ], batch size: 54, lr: 7.06e-03, grad_scale: 8.0 +2023-04-01 21:04:07,030 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 21:04:50,830 INFO [train.py:903] (3/4) Epoch 12, batch 850, loss[loss=0.2661, simple_loss=0.3288, pruned_loss=0.1017, over 13579.00 frames. ], tot_loss[loss=0.2301, simple_loss=0.3048, pruned_loss=0.07775, over 3784050.32 frames. ], batch size: 135, lr: 7.06e-03, grad_scale: 8.0 +2023-04-01 21:04:54,859 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75961.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:05:10,028 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.002e+02 5.208e+02 6.361e+02 7.722e+02 1.579e+03, threshold=1.272e+03, percent-clipped=2.0 +2023-04-01 21:05:25,738 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75986.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:05:29,799 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1035, 1.1397, 1.5713, 1.1137, 2.5903, 3.4892, 3.1916, 3.6990], + device='cuda:3'), covar=tensor([0.1627, 0.3586, 0.3184, 0.2183, 0.0511, 0.0157, 0.0221, 0.0194], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0294, 0.0324, 0.0251, 0.0214, 0.0157, 0.0206, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:05:46,881 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 21:05:48,555 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=76003.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:05:53,893 INFO [train.py:903] (3/4) Epoch 12, batch 900, loss[loss=0.2014, simple_loss=0.2664, pruned_loss=0.06815, over 19714.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3051, pruned_loss=0.07814, over 3799236.83 frames. ], batch size: 46, lr: 7.05e-03, grad_scale: 8.0 +2023-04-01 21:06:13,774 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6541, 1.4123, 1.4420, 2.0028, 1.6194, 1.9546, 1.9468, 1.7088], + device='cuda:3'), covar=tensor([0.0779, 0.0960, 0.1019, 0.0712, 0.0823, 0.0638, 0.0850, 0.0669], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0222, 0.0244, 0.0235, 0.0212, 0.0195, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 21:06:20,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=76028.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:06:57,933 INFO [train.py:903] (3/4) Epoch 12, batch 950, loss[loss=0.2591, simple_loss=0.3364, pruned_loss=0.09092, over 18267.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3059, pruned_loss=0.07858, over 3807321.71 frames. ], batch size: 83, lr: 7.05e-03, grad_scale: 8.0 +2023-04-01 21:07:02,614 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 21:07:17,511 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.890e+02 5.102e+02 6.493e+02 8.387e+02 1.985e+03, threshold=1.299e+03, percent-clipped=2.0 +2023-04-01 21:07:59,855 INFO [train.py:903] (3/4) Epoch 12, batch 1000, loss[loss=0.2037, simple_loss=0.2722, pruned_loss=0.06764, over 19770.00 frames. ], tot_loss[loss=0.2311, simple_loss=0.3052, pruned_loss=0.07845, over 3814222.80 frames. ], batch size: 47, lr: 7.05e-03, grad_scale: 8.0 +2023-04-01 21:08:42,444 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0154, 3.3307, 1.8590, 2.0258, 3.0341, 1.5045, 1.2598, 1.9177], + device='cuda:3'), covar=tensor([0.1188, 0.0492, 0.0980, 0.0662, 0.0401, 0.1074, 0.0978, 0.0658], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0303, 0.0326, 0.0247, 0.0233, 0.0314, 0.0289, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:08:54,934 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 21:09:02,681 INFO [train.py:903] (3/4) Epoch 12, batch 1050, loss[loss=0.2148, simple_loss=0.2914, pruned_loss=0.06911, over 19655.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3054, pruned_loss=0.07878, over 3811378.95 frames. ], batch size: 53, lr: 7.05e-03, grad_scale: 8.0 +2023-04-01 21:09:20,680 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.538e+02 5.178e+02 6.428e+02 8.624e+02 1.751e+03, threshold=1.286e+03, percent-clipped=4.0 +2023-04-01 21:09:35,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 21:10:05,079 INFO [train.py:903] (3/4) Epoch 12, batch 1100, loss[loss=0.2269, simple_loss=0.3016, pruned_loss=0.07611, over 19838.00 frames. ], tot_loss[loss=0.2326, simple_loss=0.3063, pruned_loss=0.07952, over 3802404.41 frames. ], batch size: 52, lr: 7.04e-03, grad_scale: 8.0 +2023-04-01 21:10:13,330 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8884, 1.4941, 1.4936, 1.8936, 1.7146, 1.6654, 1.4935, 1.7759], + device='cuda:3'), covar=tensor([0.0885, 0.1475, 0.1323, 0.0908, 0.1079, 0.0491, 0.1259, 0.0656], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0352, 0.0292, 0.0241, 0.0296, 0.0243, 0.0280, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:11:08,884 INFO [train.py:903] (3/4) Epoch 12, batch 1150, loss[loss=0.2265, simple_loss=0.3018, pruned_loss=0.07561, over 19458.00 frames. ], tot_loss[loss=0.234, simple_loss=0.3076, pruned_loss=0.08017, over 3788059.09 frames. ], batch size: 49, lr: 7.04e-03, grad_scale: 8.0 +2023-04-01 21:11:27,482 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.258e+02 5.432e+02 6.664e+02 8.568e+02 1.731e+03, threshold=1.333e+03, percent-clipped=3.0 +2023-04-01 21:11:44,829 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76287.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:12:10,715 INFO [train.py:903] (3/4) Epoch 12, batch 1200, loss[loss=0.2473, simple_loss=0.3212, pruned_loss=0.08667, over 19502.00 frames. ], tot_loss[loss=0.2346, simple_loss=0.3082, pruned_loss=0.08051, over 3810754.99 frames. ], batch size: 64, lr: 7.04e-03, grad_scale: 8.0 +2023-04-01 21:12:15,531 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76312.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:12:45,725 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 21:12:52,915 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76341.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 21:13:14,060 INFO [train.py:903] (3/4) Epoch 12, batch 1250, loss[loss=0.2502, simple_loss=0.3294, pruned_loss=0.08552, over 19520.00 frames. ], tot_loss[loss=0.2333, simple_loss=0.3073, pruned_loss=0.07962, over 3832196.59 frames. ], batch size: 54, lr: 7.04e-03, grad_scale: 8.0 +2023-04-01 21:13:31,221 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.467e+02 5.242e+02 6.279e+02 7.669e+02 1.575e+03, threshold=1.256e+03, percent-clipped=2.0 +2023-04-01 21:14:15,616 INFO [train.py:903] (3/4) Epoch 12, batch 1300, loss[loss=0.1758, simple_loss=0.2616, pruned_loss=0.04499, over 19615.00 frames. ], tot_loss[loss=0.2321, simple_loss=0.3059, pruned_loss=0.07913, over 3839827.58 frames. ], batch size: 50, lr: 7.04e-03, grad_scale: 8.0 +2023-04-01 21:15:18,800 INFO [train.py:903] (3/4) Epoch 12, batch 1350, loss[loss=0.3105, simple_loss=0.3575, pruned_loss=0.1317, over 13253.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3059, pruned_loss=0.07911, over 3831211.11 frames. ], batch size: 135, lr: 7.03e-03, grad_scale: 8.0 +2023-04-01 21:15:37,067 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.121e+02 5.247e+02 6.486e+02 7.910e+02 1.390e+03, threshold=1.297e+03, percent-clipped=4.0 +2023-04-01 21:16:20,445 INFO [train.py:903] (3/4) Epoch 12, batch 1400, loss[loss=0.1944, simple_loss=0.27, pruned_loss=0.05944, over 19762.00 frames. ], tot_loss[loss=0.2319, simple_loss=0.3058, pruned_loss=0.07905, over 3832298.93 frames. ], batch size: 47, lr: 7.03e-03, grad_scale: 16.0 +2023-04-01 21:16:50,821 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2186, 2.0767, 2.0046, 2.2542, 2.2236, 1.9598, 1.9343, 2.1766], + device='cuda:3'), covar=tensor([0.0696, 0.1133, 0.0975, 0.0730, 0.0815, 0.0418, 0.0969, 0.0514], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0354, 0.0292, 0.0242, 0.0298, 0.0245, 0.0279, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:17:01,818 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.8159, 5.2526, 3.0735, 4.5468, 1.3136, 5.0286, 5.1484, 5.1958], + device='cuda:3'), covar=tensor([0.0412, 0.0773, 0.1738, 0.0656, 0.3739, 0.0670, 0.0624, 0.0920], + device='cuda:3'), in_proj_covar=tensor([0.0442, 0.0367, 0.0442, 0.0319, 0.0380, 0.0374, 0.0358, 0.0392], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:17:23,469 INFO [train.py:903] (3/4) Epoch 12, batch 1450, loss[loss=0.2058, simple_loss=0.2922, pruned_loss=0.05963, over 19855.00 frames. ], tot_loss[loss=0.2311, simple_loss=0.3054, pruned_loss=0.07837, over 3824215.74 frames. ], batch size: 52, lr: 7.03e-03, grad_scale: 16.0 +2023-04-01 21:17:25,881 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 21:17:40,980 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.252e+02 4.935e+02 6.510e+02 8.250e+02 1.774e+03, threshold=1.302e+03, percent-clipped=3.0 +2023-04-01 21:18:24,878 INFO [train.py:903] (3/4) Epoch 12, batch 1500, loss[loss=0.2037, simple_loss=0.2818, pruned_loss=0.06283, over 19365.00 frames. ], tot_loss[loss=0.2313, simple_loss=0.3055, pruned_loss=0.07861, over 3810581.69 frames. ], batch size: 47, lr: 7.03e-03, grad_scale: 8.0 +2023-04-01 21:18:53,783 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=76631.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:19:24,250 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=76656.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:19:27,185 INFO [train.py:903] (3/4) Epoch 12, batch 1550, loss[loss=0.2861, simple_loss=0.3485, pruned_loss=0.1118, over 18191.00 frames. ], tot_loss[loss=0.2308, simple_loss=0.305, pruned_loss=0.07829, over 3810801.48 frames. ], batch size: 83, lr: 7.02e-03, grad_scale: 8.0 +2023-04-01 21:19:46,334 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.911e+02 5.709e+02 6.905e+02 9.210e+02 1.884e+03, threshold=1.381e+03, percent-clipped=4.0 +2023-04-01 21:20:01,260 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=76685.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:20:19,383 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-04-01 21:20:23,685 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-04-01 21:20:29,938 INFO [train.py:903] (3/4) Epoch 12, batch 1600, loss[loss=0.2096, simple_loss=0.2806, pruned_loss=0.06927, over 19348.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3048, pruned_loss=0.07832, over 3812357.93 frames. ], batch size: 47, lr: 7.02e-03, grad_scale: 8.0 +2023-04-01 21:20:54,022 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 21:21:16,449 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76746.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:21:23,382 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2700, 2.1410, 1.6303, 1.1765, 2.0380, 1.0013, 1.1378, 1.9948], + device='cuda:3'), covar=tensor([0.0855, 0.0616, 0.0854, 0.0841, 0.0439, 0.1156, 0.0758, 0.0309], + device='cuda:3'), in_proj_covar=tensor([0.0289, 0.0300, 0.0321, 0.0243, 0.0231, 0.0316, 0.0289, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:21:29,664 INFO [train.py:903] (3/4) Epoch 12, batch 1650, loss[loss=0.2027, simple_loss=0.2727, pruned_loss=0.06639, over 19618.00 frames. ], tot_loss[loss=0.2304, simple_loss=0.3045, pruned_loss=0.07813, over 3816900.93 frames. ], batch size: 50, lr: 7.02e-03, grad_scale: 8.0 +2023-04-01 21:21:30,367 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-01 21:21:46,972 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76771.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:21:49,949 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.672e+02 5.088e+02 6.360e+02 7.707e+02 1.579e+03, threshold=1.272e+03, percent-clipped=3.0 +2023-04-01 21:22:06,280 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4435, 1.4769, 1.4173, 2.0025, 1.5331, 1.9371, 1.6028, 1.3140], + device='cuda:3'), covar=tensor([0.2942, 0.2591, 0.1679, 0.1611, 0.2495, 0.1184, 0.3163, 0.2858], + device='cuda:3'), in_proj_covar=tensor([0.0791, 0.0816, 0.0646, 0.0889, 0.0781, 0.0706, 0.0779, 0.0709], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:22:22,696 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76800.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:22:33,401 INFO [train.py:903] (3/4) Epoch 12, batch 1700, loss[loss=0.2375, simple_loss=0.3104, pruned_loss=0.08236, over 19565.00 frames. ], tot_loss[loss=0.2305, simple_loss=0.3044, pruned_loss=0.07833, over 3821382.83 frames. ], batch size: 61, lr: 7.02e-03, grad_scale: 8.0 +2023-04-01 21:22:59,229 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3348, 1.7776, 1.9608, 2.8060, 2.1663, 2.6400, 2.6065, 2.4403], + device='cuda:3'), covar=tensor([0.0704, 0.0893, 0.0928, 0.0860, 0.0917, 0.0638, 0.0875, 0.0565], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0225, 0.0224, 0.0246, 0.0236, 0.0213, 0.0195, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 21:23:10,318 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 21:23:16,945 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0990, 2.0217, 1.8847, 1.7756, 1.6988, 1.8552, 1.0590, 1.4919], + device='cuda:3'), covar=tensor([0.0323, 0.0422, 0.0304, 0.0458, 0.0686, 0.0590, 0.0805, 0.0626], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0329, 0.0332, 0.0356, 0.0427, 0.0354, 0.0309, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:23:33,278 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76857.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:23:34,184 INFO [train.py:903] (3/4) Epoch 12, batch 1750, loss[loss=0.1988, simple_loss=0.273, pruned_loss=0.06225, over 19314.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3051, pruned_loss=0.07912, over 3814498.32 frames. ], batch size: 44, lr: 7.01e-03, grad_scale: 8.0 +2023-04-01 21:23:36,383 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-01 21:23:53,639 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.317e+02 5.587e+02 7.065e+02 8.864e+02 2.096e+03, threshold=1.413e+03, percent-clipped=6.0 +2023-04-01 21:24:37,142 INFO [train.py:903] (3/4) Epoch 12, batch 1800, loss[loss=0.2003, simple_loss=0.285, pruned_loss=0.05776, over 19770.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3048, pruned_loss=0.07847, over 3823198.70 frames. ], batch size: 54, lr: 7.01e-03, grad_scale: 8.0 +2023-04-01 21:24:44,484 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4807, 2.3344, 1.7955, 1.6023, 2.2169, 1.3302, 1.3559, 1.9953], + device='cuda:3'), covar=tensor([0.0941, 0.0711, 0.0876, 0.0730, 0.0390, 0.1048, 0.0717, 0.0366], + device='cuda:3'), in_proj_covar=tensor([0.0287, 0.0300, 0.0320, 0.0243, 0.0231, 0.0317, 0.0289, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:25:32,729 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 21:25:38,683 INFO [train.py:903] (3/4) Epoch 12, batch 1850, loss[loss=0.1931, simple_loss=0.274, pruned_loss=0.05606, over 19465.00 frames. ], tot_loss[loss=0.2323, simple_loss=0.3058, pruned_loss=0.07944, over 3812226.75 frames. ], batch size: 49, lr: 7.01e-03, grad_scale: 8.0 +2023-04-01 21:25:45,874 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.06 vs. limit=5.0 +2023-04-01 21:25:59,469 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.186e+02 5.640e+02 7.026e+02 8.457e+02 1.689e+03, threshold=1.405e+03, percent-clipped=1.0 +2023-04-01 21:26:12,301 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 21:26:34,355 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77002.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:26:41,889 INFO [train.py:903] (3/4) Epoch 12, batch 1900, loss[loss=0.2441, simple_loss=0.3212, pruned_loss=0.0835, over 19665.00 frames. ], tot_loss[loss=0.2323, simple_loss=0.3059, pruned_loss=0.07936, over 3826817.20 frames. ], batch size: 60, lr: 7.01e-03, grad_scale: 8.0 +2023-04-01 21:27:00,201 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 21:27:04,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 21:27:06,424 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77027.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:27:06,441 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77027.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:27:28,494 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 21:27:36,569 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77052.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:27:42,275 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77056.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:27:43,956 INFO [train.py:903] (3/4) Epoch 12, batch 1950, loss[loss=0.2465, simple_loss=0.3261, pruned_loss=0.08349, over 19389.00 frames. ], tot_loss[loss=0.2337, simple_loss=0.3071, pruned_loss=0.08017, over 3823496.75 frames. ], batch size: 70, lr: 7.01e-03, grad_scale: 8.0 +2023-04-01 21:28:03,224 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.700e+02 5.468e+02 7.092e+02 9.065e+02 1.810e+03, threshold=1.418e+03, percent-clipped=4.0 +2023-04-01 21:28:11,862 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77081.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 21:28:45,468 INFO [train.py:903] (3/4) Epoch 12, batch 2000, loss[loss=0.2428, simple_loss=0.3147, pruned_loss=0.08543, over 18162.00 frames. ], tot_loss[loss=0.2348, simple_loss=0.3081, pruned_loss=0.08072, over 3815243.95 frames. ], batch size: 83, lr: 7.00e-03, grad_scale: 8.0 +2023-04-01 21:28:51,490 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77113.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:29:15,566 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5407, 1.2845, 1.1380, 1.4378, 1.0987, 1.3313, 1.1920, 1.3494], + device='cuda:3'), covar=tensor([0.0993, 0.1141, 0.1540, 0.0935, 0.1194, 0.0551, 0.1285, 0.0761], + device='cuda:3'), in_proj_covar=tensor([0.0252, 0.0349, 0.0292, 0.0238, 0.0294, 0.0242, 0.0278, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:29:43,047 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 21:29:46,529 INFO [train.py:903] (3/4) Epoch 12, batch 2050, loss[loss=0.2509, simple_loss=0.3186, pruned_loss=0.09162, over 17640.00 frames. ], tot_loss[loss=0.2352, simple_loss=0.3086, pruned_loss=0.08097, over 3801636.06 frames. ], batch size: 101, lr: 7.00e-03, grad_scale: 8.0 +2023-04-01 21:30:02,210 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 21:30:03,436 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 21:30:06,833 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.225e+02 5.837e+02 7.308e+02 9.815e+02 2.165e+03, threshold=1.462e+03, percent-clipped=5.0 +2023-04-01 21:30:26,544 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 21:30:40,585 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77201.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:30:49,629 INFO [train.py:903] (3/4) Epoch 12, batch 2100, loss[loss=0.2208, simple_loss=0.3034, pruned_loss=0.06907, over 19766.00 frames. ], tot_loss[loss=0.235, simple_loss=0.3083, pruned_loss=0.08084, over 3797944.86 frames. ], batch size: 63, lr: 7.00e-03, grad_scale: 4.0 +2023-04-01 21:31:10,324 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 21:31:12,334 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8575, 4.4938, 3.2047, 4.0104, 1.9584, 4.2026, 4.2111, 4.3147], + device='cuda:3'), covar=tensor([0.0503, 0.0824, 0.1682, 0.0760, 0.2850, 0.0725, 0.0780, 0.0990], + device='cuda:3'), in_proj_covar=tensor([0.0445, 0.0369, 0.0444, 0.0321, 0.0384, 0.0378, 0.0364, 0.0397], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:31:17,043 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77229.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:31:20,315 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 21:31:28,734 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6847, 1.3913, 1.3867, 2.1754, 1.7466, 2.1315, 2.0965, 1.7976], + device='cuda:3'), covar=tensor([0.0807, 0.0965, 0.1055, 0.0815, 0.0867, 0.0642, 0.0870, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0226, 0.0226, 0.0249, 0.0236, 0.0215, 0.0197, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 21:31:40,667 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 21:31:52,752 INFO [train.py:903] (3/4) Epoch 12, batch 2150, loss[loss=0.3022, simple_loss=0.3541, pruned_loss=0.1252, over 13271.00 frames. ], tot_loss[loss=0.2343, simple_loss=0.3075, pruned_loss=0.08051, over 3787024.47 frames. ], batch size: 136, lr: 7.00e-03, grad_scale: 4.0 +2023-04-01 21:32:13,139 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.202e+02 5.327e+02 7.168e+02 9.347e+02 2.125e+03, threshold=1.434e+03, percent-clipped=4.0 +2023-04-01 21:32:33,277 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77291.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:32:39,476 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77295.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:32:51,912 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.67 vs. limit=5.0 +2023-04-01 21:32:53,871 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0405, 1.2432, 1.6893, 0.8979, 2.2875, 3.0613, 2.7691, 3.1868], + device='cuda:3'), covar=tensor([0.1571, 0.3401, 0.2907, 0.2299, 0.0487, 0.0175, 0.0233, 0.0232], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0295, 0.0324, 0.0250, 0.0215, 0.0158, 0.0205, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:32:55,903 INFO [train.py:903] (3/4) Epoch 12, batch 2200, loss[loss=0.2223, simple_loss=0.3075, pruned_loss=0.06857, over 19792.00 frames. ], tot_loss[loss=0.2334, simple_loss=0.307, pruned_loss=0.07988, over 3805720.43 frames. ], batch size: 56, lr: 6.99e-03, grad_scale: 4.0 +2023-04-01 21:33:05,541 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77316.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:33:25,220 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-01 21:33:57,353 INFO [train.py:903] (3/4) Epoch 12, batch 2250, loss[loss=0.2211, simple_loss=0.2948, pruned_loss=0.07367, over 19685.00 frames. ], tot_loss[loss=0.2326, simple_loss=0.3066, pruned_loss=0.07925, over 3822701.21 frames. ], batch size: 53, lr: 6.99e-03, grad_scale: 4.0 +2023-04-01 21:34:08,714 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0280, 1.2276, 1.6896, 1.0446, 2.5000, 3.3198, 3.0029, 3.4931], + device='cuda:3'), covar=tensor([0.1610, 0.3325, 0.2922, 0.2216, 0.0493, 0.0142, 0.0232, 0.0203], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0294, 0.0325, 0.0250, 0.0215, 0.0158, 0.0206, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:34:18,136 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.729e+02 5.298e+02 6.341e+02 7.997e+02 1.542e+03, threshold=1.268e+03, percent-clipped=2.0 +2023-04-01 21:34:27,396 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7043, 4.1746, 4.4033, 4.4075, 1.6362, 4.1100, 3.5963, 4.0656], + device='cuda:3'), covar=tensor([0.1397, 0.0896, 0.0543, 0.0556, 0.5187, 0.0769, 0.0631, 0.1071], + device='cuda:3'), in_proj_covar=tensor([0.0683, 0.0613, 0.0812, 0.0690, 0.0738, 0.0565, 0.0498, 0.0749], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 21:34:35,674 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 21:34:58,607 INFO [train.py:903] (3/4) Epoch 12, batch 2300, loss[loss=0.2652, simple_loss=0.3339, pruned_loss=0.09829, over 19666.00 frames. ], tot_loss[loss=0.2325, simple_loss=0.3063, pruned_loss=0.07933, over 3819606.50 frames. ], batch size: 60, lr: 6.99e-03, grad_scale: 4.0 +2023-04-01 21:35:10,957 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 21:35:59,780 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77457.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:36:00,814 INFO [train.py:903] (3/4) Epoch 12, batch 2350, loss[loss=0.1877, simple_loss=0.2611, pruned_loss=0.0572, over 19377.00 frames. ], tot_loss[loss=0.2314, simple_loss=0.3048, pruned_loss=0.07896, over 3812917.12 frames. ], batch size: 47, lr: 6.99e-03, grad_scale: 4.0 +2023-04-01 21:36:22,285 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 5.245e+02 6.457e+02 8.417e+02 4.507e+03, threshold=1.291e+03, percent-clipped=6.0 +2023-04-01 21:36:41,224 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 21:36:59,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 21:37:02,530 INFO [train.py:903] (3/4) Epoch 12, batch 2400, loss[loss=0.3063, simple_loss=0.3673, pruned_loss=0.1227, over 19332.00 frames. ], tot_loss[loss=0.2327, simple_loss=0.306, pruned_loss=0.0797, over 3827453.00 frames. ], batch size: 66, lr: 6.99e-03, grad_scale: 8.0 +2023-04-01 21:37:17,465 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8389, 1.5349, 1.4371, 2.0671, 1.6274, 2.2432, 2.2430, 1.9573], + device='cuda:3'), covar=tensor([0.0724, 0.0840, 0.0983, 0.0794, 0.0835, 0.0538, 0.0739, 0.0571], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0224, 0.0224, 0.0248, 0.0234, 0.0214, 0.0196, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 21:37:36,455 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5715, 1.6459, 1.8204, 2.0895, 1.4275, 1.8421, 2.0000, 1.7432], + device='cuda:3'), covar=tensor([0.3361, 0.2758, 0.1429, 0.1585, 0.2970, 0.1455, 0.3501, 0.2557], + device='cuda:3'), in_proj_covar=tensor([0.0802, 0.0823, 0.0653, 0.0897, 0.0788, 0.0711, 0.0789, 0.0719], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:38:04,501 INFO [train.py:903] (3/4) Epoch 12, batch 2450, loss[loss=0.2296, simple_loss=0.3101, pruned_loss=0.07456, over 19777.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.307, pruned_loss=0.08012, over 3822131.59 frames. ], batch size: 56, lr: 6.98e-03, grad_scale: 8.0 +2023-04-01 21:38:21,469 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77572.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:38:21,542 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77572.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:38:22,404 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77573.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:38:24,483 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 5.530e+02 6.529e+02 8.263e+02 1.639e+03, threshold=1.306e+03, percent-clipped=2.0 +2023-04-01 21:38:52,868 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77597.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:38:56,110 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.8535, 5.3021, 3.1035, 4.6730, 1.5213, 5.1939, 5.1958, 5.3548], + device='cuda:3'), covar=tensor([0.0372, 0.0777, 0.1754, 0.0593, 0.3497, 0.0517, 0.0614, 0.0910], + device='cuda:3'), in_proj_covar=tensor([0.0440, 0.0365, 0.0440, 0.0319, 0.0380, 0.0373, 0.0360, 0.0392], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:39:05,885 INFO [train.py:903] (3/4) Epoch 12, batch 2500, loss[loss=0.2245, simple_loss=0.3155, pruned_loss=0.06679, over 19661.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.3071, pruned_loss=0.08004, over 3813055.72 frames. ], batch size: 55, lr: 6.98e-03, grad_scale: 8.0 +2023-04-01 21:39:27,255 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1737, 1.1199, 1.1015, 1.3484, 0.9896, 1.3162, 1.3781, 1.2605], + device='cuda:3'), covar=tensor([0.0896, 0.1009, 0.1147, 0.0715, 0.0925, 0.0805, 0.0803, 0.0733], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0221, 0.0222, 0.0244, 0.0232, 0.0211, 0.0193, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 21:39:39,400 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77634.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:39:40,550 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77635.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:39:40,772 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8373, 1.7320, 1.4399, 1.8394, 1.8643, 1.4010, 1.4390, 1.6980], + device='cuda:3'), covar=tensor([0.1115, 0.1673, 0.1791, 0.1151, 0.1298, 0.0972, 0.1671, 0.0924], + device='cuda:3'), in_proj_covar=tensor([0.0250, 0.0347, 0.0290, 0.0237, 0.0292, 0.0240, 0.0277, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:39:45,169 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77639.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:40:08,013 INFO [train.py:903] (3/4) Epoch 12, batch 2550, loss[loss=0.2241, simple_loss=0.295, pruned_loss=0.0766, over 19540.00 frames. ], tot_loss[loss=0.2337, simple_loss=0.3074, pruned_loss=0.07999, over 3817938.97 frames. ], batch size: 56, lr: 6.98e-03, grad_scale: 8.0 +2023-04-01 21:40:30,544 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.096e+02 5.340e+02 6.709e+02 8.508e+02 1.809e+03, threshold=1.342e+03, percent-clipped=3.0 +2023-04-01 21:40:46,852 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77688.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:40:54,796 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77695.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:41:05,578 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-01 21:41:10,777 INFO [train.py:903] (3/4) Epoch 12, batch 2600, loss[loss=0.2966, simple_loss=0.3518, pruned_loss=0.1207, over 13626.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.3072, pruned_loss=0.08003, over 3816955.25 frames. ], batch size: 136, lr: 6.98e-03, grad_scale: 8.0 +2023-04-01 21:41:54,848 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-01 21:42:03,120 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77750.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:42:07,423 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77754.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:42:11,588 INFO [train.py:903] (3/4) Epoch 12, batch 2650, loss[loss=0.2071, simple_loss=0.2713, pruned_loss=0.07144, over 19301.00 frames. ], tot_loss[loss=0.233, simple_loss=0.3069, pruned_loss=0.07959, over 3834096.91 frames. ], batch size: 44, lr: 6.97e-03, grad_scale: 8.0 +2023-04-01 21:42:32,176 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.366e+02 5.775e+02 6.860e+02 8.613e+02 1.817e+03, threshold=1.372e+03, percent-clipped=5.0 +2023-04-01 21:42:33,297 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-01 21:43:06,971 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9848, 3.6301, 2.2516, 3.3010, 0.9107, 3.4347, 3.4275, 3.5211], + device='cuda:3'), covar=tensor([0.0840, 0.1264, 0.2292, 0.0821, 0.3932, 0.0890, 0.0868, 0.1142], + device='cuda:3'), in_proj_covar=tensor([0.0449, 0.0374, 0.0450, 0.0326, 0.0386, 0.0379, 0.0367, 0.0400], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:43:12,310 INFO [train.py:903] (3/4) Epoch 12, batch 2700, loss[loss=0.2838, simple_loss=0.3426, pruned_loss=0.1125, over 18129.00 frames. ], tot_loss[loss=0.2332, simple_loss=0.3072, pruned_loss=0.07961, over 3836169.59 frames. ], batch size: 83, lr: 6.97e-03, grad_scale: 8.0 +2023-04-01 21:43:38,341 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77828.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:44:08,828 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77853.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:44:14,414 INFO [train.py:903] (3/4) Epoch 12, batch 2750, loss[loss=0.2332, simple_loss=0.315, pruned_loss=0.07566, over 19788.00 frames. ], tot_loss[loss=0.2323, simple_loss=0.3065, pruned_loss=0.07904, over 3834600.94 frames. ], batch size: 56, lr: 6.97e-03, grad_scale: 8.0 +2023-04-01 21:44:36,446 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.412e+02 5.785e+02 6.935e+02 8.560e+02 1.739e+03, threshold=1.387e+03, percent-clipped=4.0 +2023-04-01 21:45:15,207 INFO [train.py:903] (3/4) Epoch 12, batch 2800, loss[loss=0.2301, simple_loss=0.3129, pruned_loss=0.0737, over 19342.00 frames. ], tot_loss[loss=0.2323, simple_loss=0.3065, pruned_loss=0.07908, over 3828347.46 frames. ], batch size: 70, lr: 6.97e-03, grad_scale: 8.0 +2023-04-01 21:46:01,263 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77944.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:46:19,707 INFO [train.py:903] (3/4) Epoch 12, batch 2850, loss[loss=0.2479, simple_loss=0.3226, pruned_loss=0.08665, over 19643.00 frames. ], tot_loss[loss=0.2329, simple_loss=0.3068, pruned_loss=0.07946, over 3813364.86 frames. ], batch size: 55, lr: 6.97e-03, grad_scale: 4.0 +2023-04-01 21:46:22,391 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1409, 2.3227, 2.4679, 2.3230, 4.6786, 1.4131, 2.8768, 4.8966], + device='cuda:3'), covar=tensor([0.0342, 0.2216, 0.2155, 0.1559, 0.0669, 0.2571, 0.1102, 0.0187], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0333, 0.0345, 0.0315, 0.0340, 0.0330, 0.0331, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 21:46:32,683 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77969.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:46:41,208 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.493e+02 5.663e+02 6.634e+02 8.773e+02 3.814e+03, threshold=1.327e+03, percent-clipped=6.0 +2023-04-01 21:46:43,750 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=77978.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:47:20,362 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78006.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:47:22,309 INFO [train.py:903] (3/4) Epoch 12, batch 2900, loss[loss=0.173, simple_loss=0.2518, pruned_loss=0.0471, over 19802.00 frames. ], tot_loss[loss=0.2328, simple_loss=0.3069, pruned_loss=0.07935, over 3801925.62 frames. ], batch size: 48, lr: 6.96e-03, grad_scale: 4.0 +2023-04-01 21:47:22,329 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-01 21:47:25,128 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78010.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:47:39,756 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5939, 1.9188, 2.2555, 1.8768, 3.2215, 2.7780, 3.6197, 1.5160], + device='cuda:3'), covar=tensor([0.2001, 0.3335, 0.2189, 0.1623, 0.1369, 0.1694, 0.1411, 0.3404], + device='cuda:3'), in_proj_covar=tensor([0.0489, 0.0578, 0.0608, 0.0437, 0.0594, 0.0489, 0.0644, 0.0495], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:47:53,543 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78031.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:47:57,901 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78035.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:48:02,345 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=78039.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:48:25,036 INFO [train.py:903] (3/4) Epoch 12, batch 2950, loss[loss=0.2154, simple_loss=0.2872, pruned_loss=0.07179, over 19842.00 frames. ], tot_loss[loss=0.2324, simple_loss=0.3063, pruned_loss=0.07926, over 3792249.37 frames. ], batch size: 52, lr: 6.96e-03, grad_scale: 4.0 +2023-04-01 21:48:48,724 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.235e+02 5.583e+02 6.866e+02 9.102e+02 1.641e+03, threshold=1.373e+03, percent-clipped=7.0 +2023-04-01 21:48:57,555 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.53 vs. limit=2.0 +2023-04-01 21:49:10,007 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78093.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:49:28,178 INFO [train.py:903] (3/4) Epoch 12, batch 3000, loss[loss=0.2024, simple_loss=0.2776, pruned_loss=0.06357, over 19779.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3053, pruned_loss=0.07891, over 3797722.02 frames. ], batch size: 47, lr: 6.96e-03, grad_scale: 4.0 +2023-04-01 21:49:28,178 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 21:49:40,665 INFO [train.py:937] (3/4) Epoch 12, validation: loss=0.1772, simple_loss=0.2779, pruned_loss=0.0383, over 944034.00 frames. +2023-04-01 21:49:40,666 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 21:49:45,492 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-01 21:50:04,820 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-01 21:50:38,059 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78154.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:50:42,292 INFO [train.py:903] (3/4) Epoch 12, batch 3050, loss[loss=0.1666, simple_loss=0.2529, pruned_loss=0.04016, over 19757.00 frames. ], tot_loss[loss=0.2313, simple_loss=0.3052, pruned_loss=0.07871, over 3801848.39 frames. ], batch size: 47, lr: 6.96e-03, grad_scale: 4.0 +2023-04-01 21:50:57,239 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1966, 1.1968, 1.7210, 1.2510, 2.5539, 3.5801, 3.3929, 3.8216], + device='cuda:3'), covar=tensor([0.1560, 0.3586, 0.2914, 0.2139, 0.0529, 0.0149, 0.0185, 0.0186], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0296, 0.0325, 0.0252, 0.0216, 0.0159, 0.0205, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:51:04,785 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.897e+02 5.238e+02 6.566e+02 8.426e+02 1.854e+03, threshold=1.313e+03, percent-clipped=6.0 +2023-04-01 21:51:43,597 INFO [train.py:903] (3/4) Epoch 12, batch 3100, loss[loss=0.2368, simple_loss=0.3153, pruned_loss=0.07916, over 19731.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3059, pruned_loss=0.07898, over 3812219.63 frames. ], batch size: 63, lr: 6.95e-03, grad_scale: 4.0 +2023-04-01 21:52:07,085 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-01 21:52:36,438 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:52:45,567 INFO [train.py:903] (3/4) Epoch 12, batch 3150, loss[loss=0.3227, simple_loss=0.3735, pruned_loss=0.136, over 13105.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3057, pruned_loss=0.07875, over 3804640.88 frames. ], batch size: 135, lr: 6.95e-03, grad_scale: 4.0 +2023-04-01 21:53:07,629 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.745e+02 5.622e+02 7.364e+02 8.683e+02 1.879e+03, threshold=1.473e+03, percent-clipped=3.0 +2023-04-01 21:53:15,182 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-01 21:53:47,075 INFO [train.py:903] (3/4) Epoch 12, batch 3200, loss[loss=0.2344, simple_loss=0.2976, pruned_loss=0.08558, over 19734.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3057, pruned_loss=0.0787, over 3808633.35 frames. ], batch size: 51, lr: 6.95e-03, grad_scale: 8.0 +2023-04-01 21:54:39,760 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78349.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:54:50,789 INFO [train.py:903] (3/4) Epoch 12, batch 3250, loss[loss=0.2149, simple_loss=0.2994, pruned_loss=0.06518, over 19660.00 frames. ], tot_loss[loss=0.2321, simple_loss=0.3061, pruned_loss=0.07906, over 3804142.41 frames. ], batch size: 55, lr: 6.95e-03, grad_scale: 8.0 +2023-04-01 21:54:57,746 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-04-01 21:55:10,601 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0038, 2.0816, 2.3073, 2.8212, 2.0007, 2.6571, 2.5310, 2.1681], + device='cuda:3'), covar=tensor([0.3528, 0.3056, 0.1373, 0.1931, 0.3438, 0.1539, 0.3468, 0.2535], + device='cuda:3'), in_proj_covar=tensor([0.0798, 0.0823, 0.0652, 0.0895, 0.0785, 0.0711, 0.0788, 0.0717], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:55:11,649 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78374.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:55:13,651 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.865e+02 5.227e+02 6.164e+02 7.465e+02 1.234e+03, threshold=1.233e+03, percent-clipped=0.0 +2023-04-01 21:55:54,291 INFO [train.py:903] (3/4) Epoch 12, batch 3300, loss[loss=0.2172, simple_loss=0.2787, pruned_loss=0.07785, over 19371.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3061, pruned_loss=0.07899, over 3797523.65 frames. ], batch size: 47, lr: 6.95e-03, grad_scale: 8.0 +2023-04-01 21:55:57,153 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78410.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:55:57,916 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-01 21:56:26,541 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78435.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:56:55,606 INFO [train.py:903] (3/4) Epoch 12, batch 3350, loss[loss=0.2299, simple_loss=0.3053, pruned_loss=0.07731, over 19583.00 frames. ], tot_loss[loss=0.2324, simple_loss=0.3064, pruned_loss=0.07919, over 3797811.90 frames. ], batch size: 52, lr: 6.94e-03, grad_scale: 8.0 +2023-04-01 21:56:58,304 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2985, 1.4131, 1.6364, 1.4844, 2.3937, 2.1369, 2.5170, 0.9044], + device='cuda:3'), covar=tensor([0.2109, 0.3684, 0.2246, 0.1747, 0.1408, 0.1839, 0.1339, 0.3701], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0570, 0.0598, 0.0430, 0.0587, 0.0483, 0.0636, 0.0486], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 21:57:18,016 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.479e+02 5.251e+02 6.145e+02 6.896e+02 1.617e+03, threshold=1.229e+03, percent-clipped=1.0 +2023-04-01 21:57:21,542 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78478.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:57:37,005 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.15 vs. limit=5.0 +2023-04-01 21:57:57,207 INFO [train.py:903] (3/4) Epoch 12, batch 3400, loss[loss=0.2269, simple_loss=0.302, pruned_loss=0.07592, over 19790.00 frames. ], tot_loss[loss=0.2321, simple_loss=0.3059, pruned_loss=0.07908, over 3806076.50 frames. ], batch size: 56, lr: 6.94e-03, grad_scale: 8.0 +2023-04-01 21:58:04,410 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2721, 2.4142, 2.4603, 3.2733, 2.3999, 3.2095, 2.8073, 2.3851], + device='cuda:3'), covar=tensor([0.3614, 0.2962, 0.1434, 0.1863, 0.3492, 0.1389, 0.3398, 0.2590], + device='cuda:3'), in_proj_covar=tensor([0.0800, 0.0824, 0.0652, 0.0895, 0.0784, 0.0711, 0.0789, 0.0717], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:59:00,759 INFO [train.py:903] (3/4) Epoch 12, batch 3450, loss[loss=0.251, simple_loss=0.3245, pruned_loss=0.08871, over 19318.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3058, pruned_loss=0.07868, over 3814243.25 frames. ], batch size: 66, lr: 6.94e-03, grad_scale: 8.0 +2023-04-01 21:59:06,198 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-01 21:59:06,691 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8350, 1.8959, 2.0724, 2.7735, 1.9179, 2.6965, 2.3676, 1.9086], + device='cuda:3'), covar=tensor([0.3611, 0.3076, 0.1468, 0.1683, 0.3290, 0.1332, 0.3580, 0.2844], + device='cuda:3'), in_proj_covar=tensor([0.0800, 0.0827, 0.0652, 0.0894, 0.0787, 0.0710, 0.0788, 0.0718], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 21:59:21,409 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78574.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 21:59:23,277 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.785e+02 5.692e+02 7.472e+02 9.495e+02 2.057e+03, threshold=1.494e+03, percent-clipped=9.0 +2023-04-01 21:59:44,692 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=78594.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:00:03,311 INFO [train.py:903] (3/4) Epoch 12, batch 3500, loss[loss=0.2517, simple_loss=0.3241, pruned_loss=0.08964, over 18208.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3055, pruned_loss=0.07877, over 3807932.13 frames. ], batch size: 83, lr: 6.94e-03, grad_scale: 8.0 +2023-04-01 22:00:54,070 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5767, 1.4029, 1.4057, 1.7946, 1.4375, 1.7799, 1.8429, 1.6145], + device='cuda:3'), covar=tensor([0.0806, 0.0975, 0.1024, 0.0715, 0.0844, 0.0724, 0.0782, 0.0671], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0226, 0.0225, 0.0248, 0.0236, 0.0213, 0.0196, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:01:05,327 INFO [train.py:903] (3/4) Epoch 12, batch 3550, loss[loss=0.2315, simple_loss=0.3091, pruned_loss=0.07699, over 19767.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3053, pruned_loss=0.0781, over 3808861.04 frames. ], batch size: 63, lr: 6.93e-03, grad_scale: 8.0 +2023-04-01 22:01:09,484 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-04-01 22:01:26,741 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.897e+02 5.737e+02 7.042e+02 1.025e+03 1.962e+03, threshold=1.408e+03, percent-clipped=6.0 +2023-04-01 22:02:07,313 INFO [train.py:903] (3/4) Epoch 12, batch 3600, loss[loss=0.2217, simple_loss=0.3075, pruned_loss=0.06796, over 19698.00 frames. ], tot_loss[loss=0.23, simple_loss=0.3047, pruned_loss=0.07761, over 3821053.73 frames. ], batch size: 59, lr: 6.93e-03, grad_scale: 8.0 +2023-04-01 22:02:08,953 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78709.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:02:13,718 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-01 22:03:09,171 INFO [train.py:903] (3/4) Epoch 12, batch 3650, loss[loss=0.1831, simple_loss=0.2637, pruned_loss=0.05124, over 19499.00 frames. ], tot_loss[loss=0.231, simple_loss=0.3054, pruned_loss=0.07835, over 3815772.28 frames. ], batch size: 49, lr: 6.93e-03, grad_scale: 8.0 +2023-04-01 22:03:29,490 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8173, 3.1817, 3.2663, 3.3095, 1.3363, 3.1683, 2.8199, 3.0110], + device='cuda:3'), covar=tensor([0.1417, 0.0971, 0.0765, 0.0772, 0.4826, 0.0855, 0.0729, 0.1274], + device='cuda:3'), in_proj_covar=tensor([0.0674, 0.0612, 0.0801, 0.0685, 0.0729, 0.0569, 0.0488, 0.0738], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 22:03:33,802 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.478e+02 5.318e+02 6.562e+02 8.219e+02 2.478e+03, threshold=1.312e+03, percent-clipped=4.0 +2023-04-01 22:04:14,200 INFO [train.py:903] (3/4) Epoch 12, batch 3700, loss[loss=0.1976, simple_loss=0.2856, pruned_loss=0.05477, over 19659.00 frames. ], tot_loss[loss=0.2298, simple_loss=0.3043, pruned_loss=0.07764, over 3832883.65 frames. ], batch size: 53, lr: 6.93e-03, grad_scale: 8.0 +2023-04-01 22:04:31,405 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=78822.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:04:31,618 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5892, 1.3895, 1.3772, 1.9149, 1.7046, 1.8646, 1.9527, 1.7779], + device='cuda:3'), covar=tensor([0.0898, 0.1022, 0.1097, 0.0943, 0.0871, 0.0750, 0.0879, 0.0686], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0226, 0.0224, 0.0249, 0.0237, 0.0214, 0.0196, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:04:34,438 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-01 22:05:15,856 INFO [train.py:903] (3/4) Epoch 12, batch 3750, loss[loss=0.2023, simple_loss=0.2871, pruned_loss=0.05874, over 19521.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.3041, pruned_loss=0.07765, over 3827617.20 frames. ], batch size: 56, lr: 6.93e-03, grad_scale: 8.0 +2023-04-01 22:05:37,741 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.608e+02 5.233e+02 6.179e+02 8.242e+02 1.500e+03, threshold=1.236e+03, percent-clipped=2.0 +2023-04-01 22:05:59,317 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.10 vs. limit=2.0 +2023-04-01 22:06:05,769 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.61 vs. limit=2.0 +2023-04-01 22:06:16,426 INFO [train.py:903] (3/4) Epoch 12, batch 3800, loss[loss=0.2709, simple_loss=0.3392, pruned_loss=0.1013, over 13401.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.3043, pruned_loss=0.07755, over 3822788.57 frames. ], batch size: 136, lr: 6.92e-03, grad_scale: 8.0 +2023-04-01 22:06:29,250 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=78918.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:06:53,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-01 22:06:54,049 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78937.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:07:17,949 INFO [train.py:903] (3/4) Epoch 12, batch 3850, loss[loss=0.2679, simple_loss=0.3401, pruned_loss=0.09784, over 18839.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.306, pruned_loss=0.07864, over 3811947.14 frames. ], batch size: 74, lr: 6.92e-03, grad_scale: 8.0 +2023-04-01 22:07:28,336 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78965.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:07:40,746 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.425e+02 5.982e+02 6.977e+02 9.373e+02 2.137e+03, threshold=1.395e+03, percent-clipped=8.0 +2023-04-01 22:07:49,869 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78983.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:07:58,104 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78990.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:08:21,051 INFO [train.py:903] (3/4) Epoch 12, batch 3900, loss[loss=0.2245, simple_loss=0.3052, pruned_loss=0.07189, over 19606.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3064, pruned_loss=0.07878, over 3817965.85 frames. ], batch size: 61, lr: 6.92e-03, grad_scale: 8.0 +2023-04-01 22:08:51,038 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79033.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:09:15,079 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7878, 1.8362, 2.0994, 2.4271, 1.6376, 2.3046, 2.2763, 1.9439], + device='cuda:3'), covar=tensor([0.3296, 0.3035, 0.1464, 0.1660, 0.3229, 0.1484, 0.3504, 0.2652], + device='cuda:3'), in_proj_covar=tensor([0.0802, 0.0826, 0.0655, 0.0900, 0.0787, 0.0716, 0.0794, 0.0718], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 22:09:22,288 INFO [train.py:903] (3/4) Epoch 12, batch 3950, loss[loss=0.1975, simple_loss=0.2658, pruned_loss=0.06463, over 19763.00 frames. ], tot_loss[loss=0.2312, simple_loss=0.3055, pruned_loss=0.07845, over 3812015.42 frames. ], batch size: 47, lr: 6.92e-03, grad_scale: 8.0 +2023-04-01 22:09:24,092 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-01 22:09:29,035 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-01 22:09:29,441 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3300, 2.0827, 2.0130, 1.7296, 1.4299, 1.6782, 0.6622, 1.2180], + device='cuda:3'), covar=tensor([0.0436, 0.0497, 0.0342, 0.0691, 0.1013, 0.0687, 0.0945, 0.0772], + device='cuda:3'), in_proj_covar=tensor([0.0346, 0.0334, 0.0336, 0.0361, 0.0431, 0.0360, 0.0316, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:09:43,634 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.615e+02 6.132e+02 6.993e+02 8.356e+02 2.478e+03, threshold=1.399e+03, percent-clipped=5.0 +2023-04-01 22:10:22,895 INFO [train.py:903] (3/4) Epoch 12, batch 4000, loss[loss=0.2563, simple_loss=0.3307, pruned_loss=0.09096, over 19707.00 frames. ], tot_loss[loss=0.2328, simple_loss=0.3069, pruned_loss=0.07937, over 3812824.07 frames. ], batch size: 63, lr: 6.91e-03, grad_scale: 8.0 +2023-04-01 22:11:13,118 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-01 22:11:21,652 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.31 vs. limit=5.0 +2023-04-01 22:11:24,605 INFO [train.py:903] (3/4) Epoch 12, batch 4050, loss[loss=0.23, simple_loss=0.312, pruned_loss=0.07398, over 18119.00 frames. ], tot_loss[loss=0.2322, simple_loss=0.3061, pruned_loss=0.07916, over 3801839.18 frames. ], batch size: 83, lr: 6.91e-03, grad_scale: 8.0 +2023-04-01 22:11:47,130 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.257e+02 5.067e+02 6.190e+02 7.758e+02 2.001e+03, threshold=1.238e+03, percent-clipped=2.0 +2023-04-01 22:11:53,113 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.09 vs. limit=2.0 +2023-04-01 22:12:05,411 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3169, 2.0843, 1.9379, 1.7048, 1.5514, 1.7893, 0.5245, 1.1643], + device='cuda:3'), covar=tensor([0.0383, 0.0484, 0.0354, 0.0649, 0.0970, 0.0685, 0.1016, 0.0825], + device='cuda:3'), in_proj_covar=tensor([0.0346, 0.0336, 0.0338, 0.0362, 0.0433, 0.0361, 0.0318, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:12:07,830 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79193.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:12:26,659 INFO [train.py:903] (3/4) Epoch 12, batch 4100, loss[loss=0.201, simple_loss=0.2734, pruned_loss=0.06433, over 19752.00 frames. ], tot_loss[loss=0.2316, simple_loss=0.3058, pruned_loss=0.07871, over 3796162.13 frames. ], batch size: 46, lr: 6.91e-03, grad_scale: 8.0 +2023-04-01 22:12:39,328 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79218.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:13:03,857 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-01 22:13:27,008 INFO [train.py:903] (3/4) Epoch 12, batch 4150, loss[loss=0.2218, simple_loss=0.3055, pruned_loss=0.06903, over 19478.00 frames. ], tot_loss[loss=0.2318, simple_loss=0.3059, pruned_loss=0.07882, over 3769752.69 frames. ], batch size: 64, lr: 6.91e-03, grad_scale: 8.0 +2023-04-01 22:13:49,448 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.496e+02 5.552e+02 6.928e+02 9.304e+02 2.111e+03, threshold=1.386e+03, percent-clipped=6.0 +2023-04-01 22:14:06,893 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79289.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:14:23,430 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79303.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:14:28,771 INFO [train.py:903] (3/4) Epoch 12, batch 4200, loss[loss=0.2009, simple_loss=0.2789, pruned_loss=0.06139, over 19854.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3057, pruned_loss=0.07865, over 3779376.84 frames. ], batch size: 52, lr: 6.91e-03, grad_scale: 8.0 +2023-04-01 22:14:33,323 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-01 22:14:35,721 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79314.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:14:53,789 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79327.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:15:22,593 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79351.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:15:26,816 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2330, 4.3051, 4.7936, 4.7946, 2.7320, 4.4857, 4.0948, 4.5096], + device='cuda:3'), covar=tensor([0.1188, 0.2080, 0.0500, 0.0535, 0.3876, 0.0738, 0.0528, 0.0956], + device='cuda:3'), in_proj_covar=tensor([0.0689, 0.0619, 0.0816, 0.0695, 0.0741, 0.0573, 0.0497, 0.0753], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 22:15:31,208 INFO [train.py:903] (3/4) Epoch 12, batch 4250, loss[loss=0.2126, simple_loss=0.2929, pruned_loss=0.06618, over 19844.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3051, pruned_loss=0.07834, over 3777012.00 frames. ], batch size: 52, lr: 6.90e-03, grad_scale: 8.0 +2023-04-01 22:15:43,240 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-01 22:15:52,358 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.517e+02 4.925e+02 6.557e+02 8.003e+02 1.515e+03, threshold=1.311e+03, percent-clipped=3.0 +2023-04-01 22:15:54,777 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-01 22:16:00,656 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79382.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:16:13,159 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79393.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 22:16:13,617 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.47 vs. limit=5.0 +2023-04-01 22:16:33,041 INFO [train.py:903] (3/4) Epoch 12, batch 4300, loss[loss=0.2398, simple_loss=0.3144, pruned_loss=0.08265, over 19673.00 frames. ], tot_loss[loss=0.2314, simple_loss=0.3055, pruned_loss=0.07866, over 3779038.25 frames. ], batch size: 55, lr: 6.90e-03, grad_scale: 8.0 +2023-04-01 22:16:43,555 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9031, 1.3492, 1.0506, 0.9704, 1.1901, 0.9569, 0.8060, 1.2646], + device='cuda:3'), covar=tensor([0.0518, 0.0699, 0.1002, 0.0553, 0.0476, 0.1096, 0.0578, 0.0410], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0299, 0.0324, 0.0245, 0.0235, 0.0318, 0.0287, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:17:07,850 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3079, 1.3190, 1.4055, 1.4350, 1.7326, 1.8506, 1.7300, 0.5117], + device='cuda:3'), covar=tensor([0.2027, 0.3537, 0.2213, 0.1640, 0.1316, 0.1799, 0.1199, 0.3747], + device='cuda:3'), in_proj_covar=tensor([0.0493, 0.0580, 0.0611, 0.0439, 0.0595, 0.0495, 0.0648, 0.0495], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 22:17:14,353 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79442.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:17:24,321 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-01 22:17:32,590 INFO [train.py:903] (3/4) Epoch 12, batch 4350, loss[loss=0.23, simple_loss=0.3144, pruned_loss=0.07286, over 19672.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3059, pruned_loss=0.07851, over 3796500.12 frames. ], batch size: 59, lr: 6.90e-03, grad_scale: 8.0 +2023-04-01 22:17:54,382 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.719e+02 5.544e+02 6.866e+02 8.754e+02 2.036e+03, threshold=1.373e+03, percent-clipped=4.0 +2023-04-01 22:17:59,265 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-01 22:18:34,726 INFO [train.py:903] (3/4) Epoch 12, batch 4400, loss[loss=0.1874, simple_loss=0.2687, pruned_loss=0.05307, over 19622.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3056, pruned_loss=0.07867, over 3803665.09 frames. ], batch size: 50, lr: 6.90e-03, grad_scale: 8.0 +2023-04-01 22:18:45,484 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3844, 2.1860, 1.6395, 1.3287, 2.0694, 1.1472, 1.2648, 1.8013], + device='cuda:3'), covar=tensor([0.1059, 0.0741, 0.1044, 0.0860, 0.0541, 0.1186, 0.0731, 0.0444], + device='cuda:3'), in_proj_covar=tensor([0.0290, 0.0298, 0.0322, 0.0244, 0.0234, 0.0315, 0.0286, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:18:58,953 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-01 22:19:07,294 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-01 22:19:38,010 INFO [train.py:903] (3/4) Epoch 12, batch 4450, loss[loss=0.2594, simple_loss=0.3315, pruned_loss=0.09366, over 18179.00 frames. ], tot_loss[loss=0.2313, simple_loss=0.3053, pruned_loss=0.07865, over 3797042.49 frames. ], batch size: 83, lr: 6.90e-03, grad_scale: 8.0 +2023-04-01 22:20:00,031 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.129e+02 5.261e+02 6.909e+02 8.531e+02 1.990e+03, threshold=1.382e+03, percent-clipped=5.0 +2023-04-01 22:20:33,620 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79603.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:20:41,077 INFO [train.py:903] (3/4) Epoch 12, batch 4500, loss[loss=0.2145, simple_loss=0.2894, pruned_loss=0.0698, over 19515.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.3042, pruned_loss=0.07757, over 3810283.22 frames. ], batch size: 49, lr: 6.89e-03, grad_scale: 8.0 +2023-04-01 22:20:51,332 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.71 vs. limit=5.0 +2023-04-01 22:21:29,751 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79647.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 22:21:43,339 INFO [train.py:903] (3/4) Epoch 12, batch 4550, loss[loss=0.2333, simple_loss=0.3109, pruned_loss=0.07783, over 19344.00 frames. ], tot_loss[loss=0.2288, simple_loss=0.3033, pruned_loss=0.07719, over 3821794.16 frames. ], batch size: 66, lr: 6.89e-03, grad_scale: 8.0 +2023-04-01 22:21:49,802 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.64 vs. limit=5.0 +2023-04-01 22:21:52,477 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-01 22:22:04,139 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79675.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:22:04,980 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.248e+02 5.458e+02 6.277e+02 7.572e+02 1.495e+03, threshold=1.255e+03, percent-clipped=2.0 +2023-04-01 22:22:11,625 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0476, 4.4487, 4.8028, 4.7441, 1.7414, 4.3738, 3.8655, 4.4472], + device='cuda:3'), covar=tensor([0.1363, 0.0624, 0.0458, 0.0517, 0.4903, 0.0568, 0.0545, 0.0887], + device='cuda:3'), in_proj_covar=tensor([0.0686, 0.0617, 0.0812, 0.0691, 0.0732, 0.0565, 0.0495, 0.0745], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 22:22:15,958 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-01 22:22:29,773 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79695.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:22:29,915 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9895, 4.3728, 4.7023, 4.6790, 1.6735, 4.3538, 3.8108, 4.3146], + device='cuda:3'), covar=tensor([0.1355, 0.0829, 0.0487, 0.0500, 0.5325, 0.0688, 0.0597, 0.0984], + device='cuda:3'), in_proj_covar=tensor([0.0686, 0.0616, 0.0812, 0.0691, 0.0732, 0.0564, 0.0494, 0.0745], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 22:22:33,619 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79698.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:22:44,775 INFO [train.py:903] (3/4) Epoch 12, batch 4600, loss[loss=0.2231, simple_loss=0.2877, pruned_loss=0.07926, over 19752.00 frames. ], tot_loss[loss=0.2305, simple_loss=0.3048, pruned_loss=0.07804, over 3816960.01 frames. ], batch size: 45, lr: 6.89e-03, grad_scale: 8.0 +2023-04-01 22:23:04,296 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79723.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:23:07,650 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79726.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:23:21,793 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79737.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:23:46,651 INFO [train.py:903] (3/4) Epoch 12, batch 4650, loss[loss=0.2151, simple_loss=0.2855, pruned_loss=0.07228, over 19408.00 frames. ], tot_loss[loss=0.2319, simple_loss=0.3055, pruned_loss=0.07915, over 3809192.33 frames. ], batch size: 48, lr: 6.89e-03, grad_scale: 8.0 +2023-04-01 22:23:49,918 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.36 vs. limit=5.0 +2023-04-01 22:23:52,622 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79762.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:24:06,322 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-01 22:24:09,608 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.145e+02 5.614e+02 7.010e+02 8.934e+02 1.991e+03, threshold=1.402e+03, percent-clipped=7.0 +2023-04-01 22:24:14,690 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79780.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:24:15,619 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-01 22:24:48,914 INFO [train.py:903] (3/4) Epoch 12, batch 4700, loss[loss=0.2363, simple_loss=0.3066, pruned_loss=0.08304, over 19727.00 frames. ], tot_loss[loss=0.232, simple_loss=0.3055, pruned_loss=0.07923, over 3816793.24 frames. ], batch size: 51, lr: 6.88e-03, grad_scale: 8.0 +2023-04-01 22:24:52,488 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79810.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:25:12,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-01 22:25:30,342 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79841.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:25:45,568 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79852.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:25:52,118 INFO [train.py:903] (3/4) Epoch 12, batch 4750, loss[loss=0.2435, simple_loss=0.3052, pruned_loss=0.09094, over 19590.00 frames. ], tot_loss[loss=0.2333, simple_loss=0.3071, pruned_loss=0.07979, over 3825679.52 frames. ], batch size: 52, lr: 6.88e-03, grad_scale: 8.0 +2023-04-01 22:26:14,213 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.848e+02 5.173e+02 6.366e+02 7.623e+02 1.625e+03, threshold=1.273e+03, percent-clipped=2.0 +2023-04-01 22:26:54,248 INFO [train.py:903] (3/4) Epoch 12, batch 4800, loss[loss=0.1983, simple_loss=0.2701, pruned_loss=0.06332, over 19747.00 frames. ], tot_loss[loss=0.2329, simple_loss=0.3066, pruned_loss=0.07963, over 3832430.10 frames. ], batch size: 46, lr: 6.88e-03, grad_scale: 8.0 +2023-04-01 22:26:58,977 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4230, 1.1289, 1.2525, 2.0653, 1.5427, 1.6150, 1.7988, 1.4320], + device='cuda:3'), covar=tensor([0.0926, 0.1259, 0.1200, 0.0682, 0.0900, 0.0913, 0.0889, 0.0852], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0226, 0.0224, 0.0247, 0.0237, 0.0213, 0.0194, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:27:42,679 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=79947.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:27:55,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8376, 4.3341, 4.5983, 4.5959, 1.7379, 4.2962, 3.7221, 4.2196], + device='cuda:3'), covar=tensor([0.1496, 0.0657, 0.0498, 0.0555, 0.5301, 0.0558, 0.0611, 0.1044], + device='cuda:3'), in_proj_covar=tensor([0.0679, 0.0615, 0.0805, 0.0688, 0.0726, 0.0558, 0.0487, 0.0739], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 22:27:56,279 INFO [train.py:903] (3/4) Epoch 12, batch 4850, loss[loss=0.1982, simple_loss=0.2815, pruned_loss=0.05751, over 19484.00 frames. ], tot_loss[loss=0.2318, simple_loss=0.3058, pruned_loss=0.0789, over 3830125.95 frames. ], batch size: 49, lr: 6.88e-03, grad_scale: 16.0 +2023-04-01 22:28:19,174 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.933e+02 5.432e+02 6.685e+02 9.186e+02 1.976e+03, threshold=1.337e+03, percent-clipped=11.0 +2023-04-01 22:28:21,572 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0193, 2.4011, 2.4697, 2.9310, 2.9446, 2.4249, 2.3619, 2.8100], + device='cuda:3'), covar=tensor([0.0638, 0.1523, 0.1184, 0.0824, 0.1065, 0.0426, 0.1077, 0.0520], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0351, 0.0293, 0.0240, 0.0298, 0.0242, 0.0281, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:28:23,589 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-01 22:28:43,800 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-01 22:28:48,489 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-01 22:28:51,011 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-01 22:28:59,976 INFO [train.py:903] (3/4) Epoch 12, batch 4900, loss[loss=0.2529, simple_loss=0.3314, pruned_loss=0.0872, over 19291.00 frames. ], tot_loss[loss=0.2317, simple_loss=0.3057, pruned_loss=0.07883, over 3820295.22 frames. ], batch size: 66, lr: 6.88e-03, grad_scale: 8.0 +2023-04-01 22:29:02,317 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-01 22:29:13,772 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80018.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:29:14,545 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80019.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:29:21,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-01 22:29:43,741 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80043.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 22:30:03,743 INFO [train.py:903] (3/4) Epoch 12, batch 4950, loss[loss=0.2328, simple_loss=0.31, pruned_loss=0.07778, over 19769.00 frames. ], tot_loss[loss=0.2312, simple_loss=0.3056, pruned_loss=0.07844, over 3820281.25 frames. ], batch size: 54, lr: 6.87e-03, grad_scale: 8.0 +2023-04-01 22:30:08,655 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80062.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:30:13,660 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80066.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:30:20,436 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-01 22:30:26,813 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.980e+02 5.263e+02 6.788e+02 8.958e+02 2.034e+03, threshold=1.358e+03, percent-clipped=5.0 +2023-04-01 22:30:38,272 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9679, 2.7287, 1.9869, 1.9034, 1.5505, 2.1825, 0.8690, 1.8686], + device='cuda:3'), covar=tensor([0.0716, 0.0649, 0.0659, 0.1169, 0.1268, 0.1140, 0.1198, 0.1082], + device='cuda:3'), in_proj_covar=tensor([0.0342, 0.0328, 0.0331, 0.0357, 0.0430, 0.0356, 0.0312, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:30:44,589 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80091.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:30:45,317 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-01 22:30:52,419 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80097.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:31:04,646 INFO [train.py:903] (3/4) Epoch 12, batch 5000, loss[loss=0.2053, simple_loss=0.2937, pruned_loss=0.05844, over 19528.00 frames. ], tot_loss[loss=0.2311, simple_loss=0.3052, pruned_loss=0.07848, over 3816089.81 frames. ], batch size: 54, lr: 6.87e-03, grad_scale: 8.0 +2023-04-01 22:31:05,128 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80108.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 22:31:13,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-01 22:31:22,111 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80122.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:31:24,071 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80124.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:31:25,169 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-01 22:31:37,121 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80133.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 22:31:38,229 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80134.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:31:46,984 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9341, 1.1647, 1.6016, 0.5274, 2.1233, 2.4805, 2.1184, 2.5897], + device='cuda:3'), covar=tensor([0.1492, 0.3386, 0.2869, 0.2284, 0.0475, 0.0228, 0.0337, 0.0296], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0299, 0.0330, 0.0252, 0.0217, 0.0161, 0.0207, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:32:06,836 INFO [train.py:903] (3/4) Epoch 12, batch 5050, loss[loss=0.2151, simple_loss=0.2757, pruned_loss=0.07728, over 16855.00 frames. ], tot_loss[loss=0.2311, simple_loss=0.3053, pruned_loss=0.07848, over 3814913.89 frames. ], batch size: 37, lr: 6.87e-03, grad_scale: 8.0 +2023-04-01 22:32:30,897 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.181e+02 5.736e+02 7.301e+02 9.465e+02 2.500e+03, threshold=1.460e+03, percent-clipped=5.0 +2023-04-01 22:32:41,347 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-01 22:33:08,748 INFO [train.py:903] (3/4) Epoch 12, batch 5100, loss[loss=0.1822, simple_loss=0.2575, pruned_loss=0.05347, over 19158.00 frames. ], tot_loss[loss=0.2306, simple_loss=0.3047, pruned_loss=0.07829, over 3805735.56 frames. ], batch size: 42, lr: 6.87e-03, grad_scale: 8.0 +2023-04-01 22:33:21,051 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-01 22:33:23,173 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-01 22:33:26,511 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-01 22:33:36,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5797, 1.3250, 1.3422, 1.8754, 1.4387, 1.8656, 1.8859, 1.6124], + device='cuda:3'), covar=tensor([0.0849, 0.1033, 0.1097, 0.0839, 0.0903, 0.0677, 0.0836, 0.0715], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0226, 0.0224, 0.0245, 0.0235, 0.0212, 0.0195, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:33:47,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80239.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:34:11,775 INFO [train.py:903] (3/4) Epoch 12, batch 5150, loss[loss=0.2416, simple_loss=0.307, pruned_loss=0.08813, over 19608.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3049, pruned_loss=0.07841, over 3810263.74 frames. ], batch size: 50, lr: 6.87e-03, grad_scale: 8.0 +2023-04-01 22:34:23,651 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-01 22:34:34,579 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.377e+02 5.205e+02 6.062e+02 7.794e+02 1.645e+03, threshold=1.212e+03, percent-clipped=2.0 +2023-04-01 22:34:51,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8613, 4.3567, 2.6673, 3.7434, 0.9766, 4.2372, 4.1677, 4.3623], + device='cuda:3'), covar=tensor([0.0576, 0.1114, 0.1912, 0.0771, 0.3969, 0.0655, 0.0781, 0.0928], + device='cuda:3'), in_proj_covar=tensor([0.0438, 0.0370, 0.0439, 0.0315, 0.0377, 0.0371, 0.0362, 0.0393], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:34:59,513 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 22:35:00,953 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5200, 2.2965, 1.6660, 1.5339, 2.1135, 1.2511, 1.3526, 1.8244], + device='cuda:3'), covar=tensor([0.1033, 0.0659, 0.0976, 0.0735, 0.0471, 0.1145, 0.0722, 0.0462], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0302, 0.0328, 0.0247, 0.0239, 0.0319, 0.0288, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:35:05,768 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80301.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:35:13,723 INFO [train.py:903] (3/4) Epoch 12, batch 5200, loss[loss=0.1978, simple_loss=0.2743, pruned_loss=0.06062, over 19749.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3057, pruned_loss=0.07869, over 3809746.36 frames. ], batch size: 51, lr: 6.86e-03, grad_scale: 8.0 +2023-04-01 22:35:24,160 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3067, 2.9773, 2.0926, 2.7137, 0.7301, 2.8529, 2.7897, 2.8801], + device='cuda:3'), covar=tensor([0.1052, 0.1281, 0.2158, 0.1002, 0.3883, 0.1065, 0.1102, 0.1299], + device='cuda:3'), in_proj_covar=tensor([0.0438, 0.0369, 0.0438, 0.0315, 0.0376, 0.0371, 0.0362, 0.0392], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:35:26,523 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80318.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:35:27,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-01 22:35:41,991 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.56 vs. limit=5.0 +2023-04-01 22:35:43,944 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2751, 1.3567, 1.6107, 1.4860, 2.2045, 1.8517, 2.1843, 0.9643], + device='cuda:3'), covar=tensor([0.2412, 0.3972, 0.2345, 0.2002, 0.1447, 0.2235, 0.1433, 0.3876], + device='cuda:3'), in_proj_covar=tensor([0.0492, 0.0583, 0.0609, 0.0440, 0.0594, 0.0497, 0.0647, 0.0497], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 22:35:58,873 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80343.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:36:03,605 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6820, 1.5678, 1.5088, 2.1751, 1.8060, 2.0171, 2.0625, 1.8451], + device='cuda:3'), covar=tensor([0.0769, 0.0887, 0.0945, 0.0760, 0.0778, 0.0681, 0.0831, 0.0632], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0224, 0.0222, 0.0243, 0.0233, 0.0211, 0.0193, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:36:13,579 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-01 22:36:16,616 INFO [train.py:903] (3/4) Epoch 12, batch 5250, loss[loss=0.2755, simple_loss=0.3371, pruned_loss=0.107, over 18166.00 frames. ], tot_loss[loss=0.2305, simple_loss=0.305, pruned_loss=0.07802, over 3819485.48 frames. ], batch size: 83, lr: 6.86e-03, grad_scale: 8.0 +2023-04-01 22:36:41,380 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.664e+02 5.603e+02 6.465e+02 8.351e+02 1.434e+03, threshold=1.293e+03, percent-clipped=3.0 +2023-04-01 22:36:57,028 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80390.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:37:18,497 INFO [train.py:903] (3/4) Epoch 12, batch 5300, loss[loss=0.1746, simple_loss=0.2473, pruned_loss=0.05095, over 19722.00 frames. ], tot_loss[loss=0.229, simple_loss=0.3038, pruned_loss=0.07708, over 3832772.78 frames. ], batch size: 45, lr: 6.86e-03, grad_scale: 8.0 +2023-04-01 22:37:22,973 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80411.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:37:28,765 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80415.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:37:42,163 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-01 22:38:12,125 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9732, 1.9301, 1.9367, 1.7893, 4.4667, 0.9806, 2.5609, 4.8215], + device='cuda:3'), covar=tensor([0.0380, 0.2409, 0.2520, 0.1671, 0.0688, 0.2654, 0.1260, 0.0188], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0342, 0.0357, 0.0321, 0.0349, 0.0332, 0.0339, 0.0361], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:38:23,298 INFO [train.py:903] (3/4) Epoch 12, batch 5350, loss[loss=0.2383, simple_loss=0.3176, pruned_loss=0.07954, over 19546.00 frames. ], tot_loss[loss=0.2298, simple_loss=0.3044, pruned_loss=0.0776, over 3806207.08 frames. ], batch size: 56, lr: 6.86e-03, grad_scale: 8.0 +2023-04-01 22:38:44,527 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.162e+02 5.085e+02 6.728e+02 8.660e+02 2.071e+03, threshold=1.346e+03, percent-clipped=4.0 +2023-04-01 22:38:50,329 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80481.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:38:52,618 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3527, 1.3733, 1.8522, 1.7501, 3.1116, 4.6183, 4.4950, 5.0495], + device='cuda:3'), covar=tensor([0.1613, 0.3579, 0.3085, 0.1863, 0.0495, 0.0137, 0.0152, 0.0146], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0297, 0.0329, 0.0252, 0.0218, 0.0161, 0.0207, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:38:59,007 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-01 22:39:08,265 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80495.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:39:23,540 INFO [train.py:903] (3/4) Epoch 12, batch 5400, loss[loss=0.262, simple_loss=0.3335, pruned_loss=0.09525, over 18763.00 frames. ], tot_loss[loss=0.2304, simple_loss=0.3049, pruned_loss=0.07799, over 3820265.81 frames. ], batch size: 74, lr: 6.85e-03, grad_scale: 8.0 +2023-04-01 22:39:28,318 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80512.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:39:38,457 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80520.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:40:24,797 INFO [train.py:903] (3/4) Epoch 12, batch 5450, loss[loss=0.1699, simple_loss=0.2472, pruned_loss=0.04628, over 19007.00 frames. ], tot_loss[loss=0.2294, simple_loss=0.3037, pruned_loss=0.07757, over 3808494.92 frames. ], batch size: 42, lr: 6.85e-03, grad_scale: 8.0 +2023-04-01 22:40:49,073 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.130e+02 5.165e+02 6.319e+02 8.444e+02 1.726e+03, threshold=1.264e+03, percent-clipped=5.0 +2023-04-01 22:41:18,911 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 22:41:26,464 INFO [train.py:903] (3/4) Epoch 12, batch 5500, loss[loss=0.1916, simple_loss=0.268, pruned_loss=0.05758, over 19773.00 frames. ], tot_loss[loss=0.2305, simple_loss=0.3047, pruned_loss=0.07816, over 3805408.51 frames. ], batch size: 47, lr: 6.85e-03, grad_scale: 8.0 +2023-04-01 22:41:53,700 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-01 22:42:12,382 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80645.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:42:29,081 INFO [train.py:903] (3/4) Epoch 12, batch 5550, loss[loss=0.2041, simple_loss=0.2995, pruned_loss=0.05441, over 19675.00 frames. ], tot_loss[loss=0.2302, simple_loss=0.3047, pruned_loss=0.07791, over 3803252.21 frames. ], batch size: 55, lr: 6.85e-03, grad_scale: 8.0 +2023-04-01 22:42:38,039 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-01 22:42:51,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.910e+02 5.291e+02 6.725e+02 8.423e+02 1.958e+03, threshold=1.345e+03, percent-clipped=4.0 +2023-04-01 22:43:28,251 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-01 22:43:31,864 INFO [train.py:903] (3/4) Epoch 12, batch 5600, loss[loss=0.2206, simple_loss=0.2873, pruned_loss=0.07699, over 19735.00 frames. ], tot_loss[loss=0.23, simple_loss=0.3045, pruned_loss=0.07771, over 3812379.96 frames. ], batch size: 45, lr: 6.85e-03, grad_scale: 8.0 +2023-04-01 22:43:32,234 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4871, 1.2734, 1.2557, 1.8979, 1.5861, 1.7673, 1.9653, 1.6637], + device='cuda:3'), covar=tensor([0.0882, 0.1043, 0.1111, 0.0866, 0.0886, 0.0799, 0.0828, 0.0692], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0225, 0.0221, 0.0244, 0.0234, 0.0210, 0.0193, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-01 22:44:23,295 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5436, 2.1384, 2.1898, 2.7368, 2.2859, 2.0815, 2.0171, 2.6235], + device='cuda:3'), covar=tensor([0.0821, 0.1645, 0.1267, 0.0848, 0.1291, 0.0464, 0.1188, 0.0611], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0351, 0.0293, 0.0238, 0.0298, 0.0240, 0.0281, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:44:29,105 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80754.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:44:30,113 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80755.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:44:33,519 INFO [train.py:903] (3/4) Epoch 12, batch 5650, loss[loss=0.232, simple_loss=0.3087, pruned_loss=0.07764, over 19472.00 frames. ], tot_loss[loss=0.23, simple_loss=0.3047, pruned_loss=0.0777, over 3815287.31 frames. ], batch size: 64, lr: 6.84e-03, grad_scale: 8.0 +2023-04-01 22:44:36,186 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80760.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:44:42,834 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2053, 1.2676, 1.7940, 1.2080, 2.5538, 3.6068, 3.4027, 3.8809], + device='cuda:3'), covar=tensor([0.1529, 0.3481, 0.2866, 0.2092, 0.0578, 0.0157, 0.0204, 0.0188], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0301, 0.0332, 0.0254, 0.0221, 0.0163, 0.0209, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:44:57,742 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.312e+02 5.237e+02 6.303e+02 7.862e+02 2.175e+03, threshold=1.261e+03, percent-clipped=3.0 +2023-04-01 22:45:00,185 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7069, 1.8678, 2.2552, 1.9214, 2.8073, 3.2839, 3.2397, 3.4860], + device='cuda:3'), covar=tensor([0.1330, 0.2728, 0.2390, 0.1784, 0.1253, 0.0305, 0.0179, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0300, 0.0331, 0.0253, 0.0220, 0.0162, 0.0208, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:45:15,141 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80791.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:45:20,712 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-01 22:45:35,186 INFO [train.py:903] (3/4) Epoch 12, batch 5700, loss[loss=0.2244, simple_loss=0.3082, pruned_loss=0.07036, over 19349.00 frames. ], tot_loss[loss=0.2319, simple_loss=0.3062, pruned_loss=0.07882, over 3797037.89 frames. ], batch size: 66, lr: 6.84e-03, grad_scale: 8.0 +2023-04-01 22:45:57,693 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80825.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:46:12,918 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2408, 1.3100, 1.2470, 1.0796, 1.0685, 1.0667, 0.0267, 0.3119], + device='cuda:3'), covar=tensor([0.0403, 0.0429, 0.0280, 0.0388, 0.0820, 0.0433, 0.0838, 0.0766], + device='cuda:3'), in_proj_covar=tensor([0.0340, 0.0329, 0.0331, 0.0355, 0.0427, 0.0357, 0.0311, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 22:46:35,835 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=80856.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:46:38,758 INFO [train.py:903] (3/4) Epoch 12, batch 5750, loss[loss=0.2573, simple_loss=0.3253, pruned_loss=0.09464, over 13712.00 frames. ], tot_loss[loss=0.2317, simple_loss=0.3058, pruned_loss=0.07875, over 3799457.33 frames. ], batch size: 137, lr: 6.84e-03, grad_scale: 8.0 +2023-04-01 22:46:39,968 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-01 22:46:47,872 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-01 22:46:52,475 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-01 22:46:52,811 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80870.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:47:00,404 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.450e+02 5.415e+02 6.686e+02 8.336e+02 1.819e+03, threshold=1.337e+03, percent-clipped=1.0 +2023-04-01 22:47:40,332 INFO [train.py:903] (3/4) Epoch 12, batch 5800, loss[loss=0.2532, simple_loss=0.3179, pruned_loss=0.09422, over 13263.00 frames. ], tot_loss[loss=0.2328, simple_loss=0.3071, pruned_loss=0.07925, over 3797536.39 frames. ], batch size: 136, lr: 6.84e-03, grad_scale: 8.0 +2023-04-01 22:48:21,634 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80940.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:48:41,888 INFO [train.py:903] (3/4) Epoch 12, batch 5850, loss[loss=0.2449, simple_loss=0.3219, pruned_loss=0.08392, over 18115.00 frames. ], tot_loss[loss=0.2333, simple_loss=0.3074, pruned_loss=0.0796, over 3804801.99 frames. ], batch size: 83, lr: 6.84e-03, grad_scale: 8.0 +2023-04-01 22:48:58,818 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80971.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:49:06,076 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.947e+02 5.387e+02 6.409e+02 7.183e+02 1.679e+03, threshold=1.282e+03, percent-clipped=1.0 +2023-04-01 22:49:17,135 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-01 22:49:43,638 INFO [train.py:903] (3/4) Epoch 12, batch 5900, loss[loss=0.1919, simple_loss=0.2634, pruned_loss=0.0602, over 19223.00 frames. ], tot_loss[loss=0.2331, simple_loss=0.3068, pruned_loss=0.0797, over 3805538.40 frames. ], batch size: 42, lr: 6.83e-03, grad_scale: 8.0 +2023-04-01 22:49:47,140 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-01 22:49:55,127 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81016.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:49:57,126 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3359, 1.2776, 1.4213, 1.6068, 2.8963, 1.1238, 2.1084, 3.2698], + device='cuda:3'), covar=tensor([0.0515, 0.2634, 0.2707, 0.1683, 0.0726, 0.2330, 0.1301, 0.0301], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0341, 0.0355, 0.0323, 0.0347, 0.0332, 0.0337, 0.0361], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:50:09,763 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-01 22:50:25,016 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81041.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:50:47,171 INFO [train.py:903] (3/4) Epoch 12, batch 5950, loss[loss=0.2097, simple_loss=0.2754, pruned_loss=0.07199, over 19723.00 frames. ], tot_loss[loss=0.2317, simple_loss=0.3056, pruned_loss=0.07892, over 3803773.80 frames. ], batch size: 46, lr: 6.83e-03, grad_scale: 8.0 +2023-04-01 22:51:10,059 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.575e+02 5.377e+02 6.760e+02 8.757e+02 1.989e+03, threshold=1.352e+03, percent-clipped=8.0 +2023-04-01 22:51:37,910 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=81098.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:51:49,672 INFO [train.py:903] (3/4) Epoch 12, batch 6000, loss[loss=0.2063, simple_loss=0.2921, pruned_loss=0.0602, over 19669.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.304, pruned_loss=0.07771, over 3821241.12 frames. ], batch size: 58, lr: 6.83e-03, grad_scale: 8.0 +2023-04-01 22:51:49,673 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 22:52:03,355 INFO [train.py:937] (3/4) Epoch 12, validation: loss=0.1765, simple_loss=0.2774, pruned_loss=0.03779, over 944034.00 frames. +2023-04-01 22:52:03,357 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 22:52:25,548 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81126.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:52:34,799 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81134.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:52:35,862 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=81135.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:52:57,839 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81151.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:53:05,463 INFO [train.py:903] (3/4) Epoch 12, batch 6050, loss[loss=0.1876, simple_loss=0.2604, pruned_loss=0.0574, over 19751.00 frames. ], tot_loss[loss=0.231, simple_loss=0.3052, pruned_loss=0.0784, over 3803413.38 frames. ], batch size: 46, lr: 6.83e-03, grad_scale: 8.0 +2023-04-01 22:53:27,715 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.001e+02 5.041e+02 6.677e+02 8.260e+02 1.738e+03, threshold=1.335e+03, percent-clipped=2.0 +2023-04-01 22:53:33,661 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0447, 3.6757, 2.1080, 2.2768, 3.2345, 1.7077, 1.4095, 2.0138], + device='cuda:3'), covar=tensor([0.1209, 0.0467, 0.0990, 0.0703, 0.0433, 0.1140, 0.0911, 0.0716], + device='cuda:3'), in_proj_covar=tensor([0.0290, 0.0303, 0.0325, 0.0246, 0.0240, 0.0320, 0.0282, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:53:52,478 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81196.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:54:05,933 INFO [train.py:903] (3/4) Epoch 12, batch 6100, loss[loss=0.2469, simple_loss=0.3213, pruned_loss=0.08627, over 19548.00 frames. ], tot_loss[loss=0.2321, simple_loss=0.3063, pruned_loss=0.07893, over 3801005.24 frames. ], batch size: 61, lr: 6.83e-03, grad_scale: 8.0 +2023-04-01 22:54:11,841 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81213.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:54:21,001 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81221.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:54:29,621 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81227.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:54:57,841 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81250.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:55:00,291 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81252.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:55:06,760 INFO [train.py:903] (3/4) Epoch 12, batch 6150, loss[loss=0.2421, simple_loss=0.3156, pruned_loss=0.08433, over 18911.00 frames. ], tot_loss[loss=0.2331, simple_loss=0.307, pruned_loss=0.07955, over 3790405.70 frames. ], batch size: 74, lr: 6.82e-03, grad_scale: 8.0 +2023-04-01 22:55:33,022 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.393e+02 5.279e+02 6.402e+02 8.020e+02 2.167e+03, threshold=1.280e+03, percent-clipped=2.0 +2023-04-01 22:55:39,189 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-01 22:56:12,038 INFO [train.py:903] (3/4) Epoch 12, batch 6200, loss[loss=0.2289, simple_loss=0.3092, pruned_loss=0.07423, over 19663.00 frames. ], tot_loss[loss=0.2315, simple_loss=0.3056, pruned_loss=0.07875, over 3797326.99 frames. ], batch size: 55, lr: 6.82e-03, grad_scale: 8.0 +2023-04-01 22:56:22,891 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81317.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:56:25,179 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81319.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:57:11,779 INFO [train.py:903] (3/4) Epoch 12, batch 6250, loss[loss=0.297, simple_loss=0.351, pruned_loss=0.1215, over 13380.00 frames. ], tot_loss[loss=0.2302, simple_loss=0.3045, pruned_loss=0.07798, over 3807243.31 frames. ], batch size: 136, lr: 6.82e-03, grad_scale: 8.0 +2023-04-01 22:57:13,551 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-01 22:57:33,897 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.420e+02 5.051e+02 6.163e+02 7.297e+02 1.401e+03, threshold=1.233e+03, percent-clipped=3.0 +2023-04-01 22:57:43,147 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-01 22:57:53,407 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9624, 3.3370, 1.9250, 2.0716, 3.0084, 1.6345, 1.3215, 1.9404], + device='cuda:3'), covar=tensor([0.1345, 0.0567, 0.1024, 0.0718, 0.0503, 0.1131, 0.0977, 0.0739], + device='cuda:3'), in_proj_covar=tensor([0.0288, 0.0301, 0.0323, 0.0246, 0.0239, 0.0317, 0.0282, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 22:58:13,185 INFO [train.py:903] (3/4) Epoch 12, batch 6300, loss[loss=0.2114, simple_loss=0.2903, pruned_loss=0.06621, over 18830.00 frames. ], tot_loss[loss=0.2296, simple_loss=0.3039, pruned_loss=0.07759, over 3804356.45 frames. ], batch size: 74, lr: 6.82e-03, grad_scale: 8.0 +2023-04-01 22:59:14,518 INFO [train.py:903] (3/4) Epoch 12, batch 6350, loss[loss=0.2089, simple_loss=0.2906, pruned_loss=0.06362, over 19686.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3037, pruned_loss=0.07744, over 3808718.77 frames. ], batch size: 53, lr: 6.81e-03, grad_scale: 8.0 +2023-04-01 22:59:29,283 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81469.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:59:39,340 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.788e+02 5.412e+02 6.997e+02 8.497e+02 1.750e+03, threshold=1.399e+03, percent-clipped=2.0 +2023-04-01 22:59:40,737 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=81478.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 22:59:59,525 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81494.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:00:14,190 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81506.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:00:16,028 INFO [train.py:903] (3/4) Epoch 12, batch 6400, loss[loss=0.2321, simple_loss=0.3098, pruned_loss=0.07723, over 19737.00 frames. ], tot_loss[loss=0.2281, simple_loss=0.3031, pruned_loss=0.07662, over 3825603.15 frames. ], batch size: 63, lr: 6.81e-03, grad_scale: 8.0 +2023-04-01 23:00:44,239 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8596, 1.9808, 2.1825, 2.7141, 1.8655, 2.5152, 2.3796, 1.9224], + device='cuda:3'), covar=tensor([0.3660, 0.3253, 0.1507, 0.1841, 0.3553, 0.1614, 0.3482, 0.2846], + device='cuda:3'), in_proj_covar=tensor([0.0807, 0.0830, 0.0652, 0.0892, 0.0792, 0.0719, 0.0790, 0.0718], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 23:00:45,264 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81531.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:01:05,922 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2283, 2.0999, 1.8419, 1.6164, 1.5210, 1.5580, 0.4117, 1.0640], + device='cuda:3'), covar=tensor([0.0646, 0.0607, 0.0433, 0.0788, 0.1157, 0.0963, 0.1122, 0.0976], + device='cuda:3'), in_proj_covar=tensor([0.0341, 0.0330, 0.0331, 0.0357, 0.0428, 0.0358, 0.0313, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:01:19,148 INFO [train.py:903] (3/4) Epoch 12, batch 6450, loss[loss=0.2282, simple_loss=0.3082, pruned_loss=0.07406, over 19689.00 frames. ], tot_loss[loss=0.2282, simple_loss=0.3033, pruned_loss=0.07653, over 3831651.40 frames. ], batch size: 60, lr: 6.81e-03, grad_scale: 8.0 +2023-04-01 23:01:41,092 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-04-01 23:01:41,363 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.161e+02 5.839e+02 6.972e+02 8.326e+02 2.886e+03, threshold=1.394e+03, percent-clipped=3.0 +2023-04-01 23:01:48,762 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9094, 2.0321, 2.1646, 2.7588, 1.8639, 2.5599, 2.3246, 1.9660], + device='cuda:3'), covar=tensor([0.3794, 0.3284, 0.1534, 0.1927, 0.3739, 0.1630, 0.3767, 0.2879], + device='cuda:3'), in_proj_covar=tensor([0.0805, 0.0828, 0.0652, 0.0890, 0.0792, 0.0718, 0.0790, 0.0718], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 23:02:03,372 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81593.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:02:06,520 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-01 23:02:20,376 INFO [train.py:903] (3/4) Epoch 12, batch 6500, loss[loss=0.2151, simple_loss=0.294, pruned_loss=0.06809, over 19680.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3045, pruned_loss=0.07708, over 3824390.47 frames. ], batch size: 53, lr: 6.81e-03, grad_scale: 8.0 +2023-04-01 23:02:27,386 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-01 23:03:22,583 INFO [train.py:903] (3/4) Epoch 12, batch 6550, loss[loss=0.2223, simple_loss=0.289, pruned_loss=0.07777, over 19754.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.304, pruned_loss=0.0771, over 3810874.72 frames. ], batch size: 47, lr: 6.81e-03, grad_scale: 8.0 +2023-04-01 23:03:26,202 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=81661.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:03:28,614 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=81663.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:03:47,187 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.342e+02 5.338e+02 6.617e+02 7.892e+02 1.534e+03, threshold=1.323e+03, percent-clipped=1.0 +2023-04-01 23:04:00,866 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81688.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:04:24,219 INFO [train.py:903] (3/4) Epoch 12, batch 6600, loss[loss=0.2338, simple_loss=0.3125, pruned_loss=0.07749, over 19657.00 frames. ], tot_loss[loss=0.2294, simple_loss=0.3046, pruned_loss=0.07714, over 3813170.93 frames. ], batch size: 58, lr: 6.80e-03, grad_scale: 8.0 +2023-04-01 23:05:26,098 INFO [train.py:903] (3/4) Epoch 12, batch 6650, loss[loss=0.2424, simple_loss=0.3174, pruned_loss=0.0837, over 17314.00 frames. ], tot_loss[loss=0.2292, simple_loss=0.304, pruned_loss=0.07717, over 3807415.06 frames. ], batch size: 102, lr: 6.80e-03, grad_scale: 8.0 +2023-04-01 23:05:46,867 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81776.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:05:47,580 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.542e+02 5.145e+02 6.461e+02 8.134e+02 1.737e+03, threshold=1.292e+03, percent-clipped=3.0 +2023-04-01 23:05:49,147 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81778.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:06:26,737 INFO [train.py:903] (3/4) Epoch 12, batch 6700, loss[loss=0.236, simple_loss=0.3144, pruned_loss=0.07876, over 19531.00 frames. ], tot_loss[loss=0.2284, simple_loss=0.3032, pruned_loss=0.07673, over 3817300.59 frames. ], batch size: 54, lr: 6.80e-03, grad_scale: 8.0 +2023-04-01 23:07:03,324 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-01 23:07:16,990 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81849.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:07:26,530 INFO [train.py:903] (3/4) Epoch 12, batch 6750, loss[loss=0.2121, simple_loss=0.2879, pruned_loss=0.06815, over 19476.00 frames. ], tot_loss[loss=0.2299, simple_loss=0.3043, pruned_loss=0.07777, over 3826633.79 frames. ], batch size: 49, lr: 6.80e-03, grad_scale: 4.0 +2023-04-01 23:07:45,313 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81874.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:07:49,447 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.155e+02 6.276e+02 7.333e+02 1.082e+03 2.540e+03, threshold=1.467e+03, percent-clipped=11.0 +2023-04-01 23:08:23,233 INFO [train.py:903] (3/4) Epoch 12, batch 6800, loss[loss=0.1942, simple_loss=0.2734, pruned_loss=0.05744, over 19746.00 frames. ], tot_loss[loss=0.2289, simple_loss=0.3033, pruned_loss=0.07722, over 3832904.86 frames. ], batch size: 47, lr: 6.80e-03, grad_scale: 8.0 +2023-04-01 23:08:40,542 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5702, 1.5626, 1.4440, 1.3454, 1.2405, 1.3866, 0.5558, 0.9446], + device='cuda:3'), covar=tensor([0.0392, 0.0413, 0.0255, 0.0409, 0.0637, 0.0487, 0.0804, 0.0614], + device='cuda:3'), in_proj_covar=tensor([0.0343, 0.0332, 0.0330, 0.0358, 0.0429, 0.0357, 0.0313, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:09:08,242 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-01 23:09:09,381 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-01 23:09:12,676 INFO [train.py:903] (3/4) Epoch 13, batch 0, loss[loss=0.2206, simple_loss=0.3002, pruned_loss=0.07048, over 19526.00 frames. ], tot_loss[loss=0.2206, simple_loss=0.3002, pruned_loss=0.07048, over 19526.00 frames. ], batch size: 54, lr: 6.53e-03, grad_scale: 8.0 +2023-04-01 23:09:12,677 INFO [train.py:928] (3/4) Computing validation loss +2023-04-01 23:09:23,574 INFO [train.py:937] (3/4) Epoch 13, validation: loss=0.176, simple_loss=0.2772, pruned_loss=0.03738, over 944034.00 frames. +2023-04-01 23:09:23,576 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-01 23:09:35,422 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-01 23:10:14,567 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.117e+02 5.222e+02 6.740e+02 8.452e+02 3.268e+03, threshold=1.348e+03, percent-clipped=4.0 +2023-04-01 23:10:23,826 INFO [train.py:903] (3/4) Epoch 13, batch 50, loss[loss=0.2226, simple_loss=0.2886, pruned_loss=0.07831, over 19762.00 frames. ], tot_loss[loss=0.2317, simple_loss=0.3075, pruned_loss=0.07794, over 864392.78 frames. ], batch size: 47, lr: 6.53e-03, grad_scale: 8.0 +2023-04-01 23:10:46,566 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82005.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:10:59,198 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-01 23:11:20,790 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82032.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:11:21,027 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82032.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:11:23,327 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82034.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:11:25,070 INFO [train.py:903] (3/4) Epoch 13, batch 100, loss[loss=0.2464, simple_loss=0.3197, pruned_loss=0.08658, over 19669.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.3017, pruned_loss=0.07509, over 1532905.27 frames. ], batch size: 55, lr: 6.53e-03, grad_scale: 8.0 +2023-04-01 23:11:36,613 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-01 23:11:52,562 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82057.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:11:55,601 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82059.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:12:11,459 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82073.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:12:16,817 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.549e+02 4.805e+02 6.218e+02 7.513e+02 1.266e+03, threshold=1.244e+03, percent-clipped=0.0 +2023-04-01 23:12:25,769 INFO [train.py:903] (3/4) Epoch 13, batch 150, loss[loss=0.1887, simple_loss=0.2592, pruned_loss=0.05907, over 19775.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3011, pruned_loss=0.07562, over 2032357.92 frames. ], batch size: 45, lr: 6.52e-03, grad_scale: 8.0 +2023-04-01 23:13:23,609 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-01 23:13:24,755 INFO [train.py:903] (3/4) Epoch 13, batch 200, loss[loss=0.2318, simple_loss=0.3132, pruned_loss=0.07521, over 18217.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.303, pruned_loss=0.07632, over 2438582.06 frames. ], batch size: 83, lr: 6.52e-03, grad_scale: 8.0 +2023-04-01 23:13:25,537 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.73 vs. limit=5.0 +2023-04-01 23:13:37,355 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-01 23:13:40,477 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82147.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:14:14,403 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.039e+02 5.002e+02 5.972e+02 7.403e+02 2.257e+03, threshold=1.194e+03, percent-clipped=4.0 +2023-04-01 23:14:26,891 INFO [train.py:903] (3/4) Epoch 13, batch 250, loss[loss=0.1796, simple_loss=0.2649, pruned_loss=0.04714, over 19477.00 frames. ], tot_loss[loss=0.2288, simple_loss=0.3039, pruned_loss=0.07683, over 2747718.97 frames. ], batch size: 49, lr: 6.52e-03, grad_scale: 8.0 +2023-04-01 23:15:26,892 INFO [train.py:903] (3/4) Epoch 13, batch 300, loss[loss=0.289, simple_loss=0.3458, pruned_loss=0.1161, over 13257.00 frames. ], tot_loss[loss=0.2295, simple_loss=0.3044, pruned_loss=0.07734, over 2984402.68 frames. ], batch size: 135, lr: 6.52e-03, grad_scale: 8.0 +2023-04-01 23:15:48,493 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-04-01 23:16:18,769 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.276e+02 5.753e+02 6.815e+02 9.164e+02 1.837e+03, threshold=1.363e+03, percent-clipped=5.0 +2023-04-01 23:16:28,135 INFO [train.py:903] (3/4) Epoch 13, batch 350, loss[loss=0.2243, simple_loss=0.3057, pruned_loss=0.07143, over 19471.00 frames. ], tot_loss[loss=0.2282, simple_loss=0.3032, pruned_loss=0.07658, over 3152904.97 frames. ], batch size: 64, lr: 6.52e-03, grad_scale: 8.0 +2023-04-01 23:16:30,461 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-01 23:17:28,520 INFO [train.py:903] (3/4) Epoch 13, batch 400, loss[loss=0.2252, simple_loss=0.2962, pruned_loss=0.07714, over 19466.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3036, pruned_loss=0.07748, over 3312721.57 frames. ], batch size: 49, lr: 6.51e-03, grad_scale: 8.0 +2023-04-01 23:17:47,110 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82349.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:18:04,879 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82364.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:18:21,974 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.254e+02 5.325e+02 6.166e+02 7.720e+02 2.046e+03, threshold=1.233e+03, percent-clipped=4.0 +2023-04-01 23:18:26,264 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.66 vs. limit=5.0 +2023-04-01 23:18:31,259 INFO [train.py:903] (3/4) Epoch 13, batch 450, loss[loss=0.2463, simple_loss=0.3335, pruned_loss=0.07959, over 19316.00 frames. ], tot_loss[loss=0.2299, simple_loss=0.3038, pruned_loss=0.07795, over 3424431.00 frames. ], batch size: 70, lr: 6.51e-03, grad_scale: 8.0 +2023-04-01 23:18:53,570 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82403.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:19:01,701 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8286, 4.3559, 2.8024, 3.8200, 1.0889, 4.2134, 4.1897, 4.3078], + device='cuda:3'), covar=tensor([0.0621, 0.1116, 0.1912, 0.0807, 0.4075, 0.0739, 0.0734, 0.0988], + device='cuda:3'), in_proj_covar=tensor([0.0444, 0.0374, 0.0445, 0.0321, 0.0381, 0.0375, 0.0369, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:19:04,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-01 23:19:04,702 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-01 23:19:09,597 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82417.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:19:23,702 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82428.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:19:33,812 INFO [train.py:903] (3/4) Epoch 13, batch 500, loss[loss=0.184, simple_loss=0.2561, pruned_loss=0.05598, over 19749.00 frames. ], tot_loss[loss=0.229, simple_loss=0.3034, pruned_loss=0.07726, over 3514690.11 frames. ], batch size: 45, lr: 6.51e-03, grad_scale: 8.0 +2023-04-01 23:20:06,493 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82464.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:20:24,365 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8403, 4.2852, 4.5664, 4.5560, 1.5117, 4.2401, 3.7676, 4.2327], + device='cuda:3'), covar=tensor([0.1538, 0.0817, 0.0555, 0.0635, 0.5840, 0.0716, 0.0600, 0.1036], + device='cuda:3'), in_proj_covar=tensor([0.0701, 0.0621, 0.0821, 0.0706, 0.0750, 0.0573, 0.0501, 0.0755], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-01 23:20:27,377 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.531e+02 5.153e+02 6.569e+02 8.401e+02 1.477e+03, threshold=1.314e+03, percent-clipped=3.0 +2023-04-01 23:20:35,263 INFO [train.py:903] (3/4) Epoch 13, batch 550, loss[loss=0.2289, simple_loss=0.3097, pruned_loss=0.07409, over 19535.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.3037, pruned_loss=0.07731, over 3591254.93 frames. ], batch size: 56, lr: 6.51e-03, grad_scale: 8.0 +2023-04-01 23:21:05,922 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82511.0, num_to_drop=1, layers_to_drop={0} +2023-04-01 23:21:30,816 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82532.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:21:35,044 INFO [train.py:903] (3/4) Epoch 13, batch 600, loss[loss=0.2216, simple_loss=0.3054, pruned_loss=0.06892, over 19585.00 frames. ], tot_loss[loss=0.231, simple_loss=0.3057, pruned_loss=0.0781, over 3638909.73 frames. ], batch size: 61, lr: 6.51e-03, grad_scale: 8.0 +2023-04-01 23:22:17,379 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-01 23:22:28,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.675e+02 5.313e+02 6.751e+02 8.249e+02 1.619e+03, threshold=1.350e+03, percent-clipped=3.0 +2023-04-01 23:22:36,943 INFO [train.py:903] (3/4) Epoch 13, batch 650, loss[loss=0.2432, simple_loss=0.3137, pruned_loss=0.08631, over 19600.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3052, pruned_loss=0.07816, over 3680524.53 frames. ], batch size: 61, lr: 6.50e-03, grad_scale: 8.0 +2023-04-01 23:22:38,468 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82587.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:23:40,983 INFO [train.py:903] (3/4) Epoch 13, batch 700, loss[loss=0.2415, simple_loss=0.3131, pruned_loss=0.08497, over 19592.00 frames. ], tot_loss[loss=0.2304, simple_loss=0.3051, pruned_loss=0.07785, over 3713870.49 frames. ], batch size: 52, lr: 6.50e-03, grad_scale: 8.0 +2023-04-01 23:24:01,395 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2505, 1.2775, 1.6730, 1.3326, 2.7801, 3.6436, 3.3828, 3.8311], + device='cuda:3'), covar=tensor([0.1500, 0.3514, 0.3102, 0.2084, 0.0513, 0.0156, 0.0198, 0.0194], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0302, 0.0331, 0.0255, 0.0222, 0.0163, 0.0207, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:24:36,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.507e+02 5.358e+02 6.603e+02 8.553e+02 2.977e+03, threshold=1.321e+03, percent-clipped=4.0 +2023-04-01 23:24:44,574 INFO [train.py:903] (3/4) Epoch 13, batch 750, loss[loss=0.2422, simple_loss=0.3185, pruned_loss=0.08292, over 19667.00 frames. ], tot_loss[loss=0.2287, simple_loss=0.3035, pruned_loss=0.07692, over 3744971.31 frames. ], batch size: 55, lr: 6.50e-03, grad_scale: 8.0 +2023-04-01 23:25:10,151 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82708.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:25:28,840 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82720.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:25:47,331 INFO [train.py:903] (3/4) Epoch 13, batch 800, loss[loss=0.2117, simple_loss=0.2849, pruned_loss=0.06927, over 19377.00 frames. ], tot_loss[loss=0.2287, simple_loss=0.3035, pruned_loss=0.07695, over 3768739.37 frames. ], batch size: 47, lr: 6.50e-03, grad_scale: 8.0 +2023-04-01 23:25:58,347 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82745.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:26:01,433 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-01 23:26:10,312 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7152, 1.8623, 1.7095, 2.7036, 1.9154, 2.5206, 1.9609, 1.3273], + device='cuda:3'), covar=tensor([0.4527, 0.3527, 0.2305, 0.2414, 0.3834, 0.1851, 0.5153, 0.4508], + device='cuda:3'), in_proj_covar=tensor([0.0812, 0.0837, 0.0661, 0.0899, 0.0800, 0.0727, 0.0800, 0.0727], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 23:26:24,324 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0006, 1.9066, 1.6401, 1.4735, 1.3774, 1.5288, 0.4128, 0.8277], + device='cuda:3'), covar=tensor([0.0512, 0.0487, 0.0333, 0.0538, 0.1016, 0.0605, 0.0915, 0.0822], + device='cuda:3'), in_proj_covar=tensor([0.0340, 0.0333, 0.0334, 0.0359, 0.0430, 0.0357, 0.0317, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:26:42,347 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.821e+02 5.442e+02 6.436e+02 7.821e+02 1.140e+03, threshold=1.287e+03, percent-clipped=0.0 +2023-04-01 23:26:50,562 INFO [train.py:903] (3/4) Epoch 13, batch 850, loss[loss=0.1963, simple_loss=0.2688, pruned_loss=0.06188, over 19721.00 frames. ], tot_loss[loss=0.2289, simple_loss=0.3035, pruned_loss=0.07715, over 3779604.13 frames. ], batch size: 51, lr: 6.50e-03, grad_scale: 8.0 +2023-04-01 23:26:53,227 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82788.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:27:27,005 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82813.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:27:38,385 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82823.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:27:44,746 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-01 23:27:52,718 INFO [train.py:903] (3/4) Epoch 13, batch 900, loss[loss=0.2501, simple_loss=0.3214, pruned_loss=0.08936, over 19380.00 frames. ], tot_loss[loss=0.2288, simple_loss=0.3035, pruned_loss=0.07707, over 3790070.59 frames. ], batch size: 70, lr: 6.49e-03, grad_scale: 8.0 +2023-04-01 23:28:19,292 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82855.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 23:28:47,654 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.460e+02 5.906e+02 6.958e+02 9.103e+02 2.196e+03, threshold=1.392e+03, percent-clipped=5.0 +2023-04-01 23:28:59,846 INFO [train.py:903] (3/4) Epoch 13, batch 950, loss[loss=0.2129, simple_loss=0.2932, pruned_loss=0.06629, over 19758.00 frames. ], tot_loss[loss=0.228, simple_loss=0.3027, pruned_loss=0.07662, over 3798049.10 frames. ], batch size: 54, lr: 6.49e-03, grad_scale: 8.0 +2023-04-01 23:29:04,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-01 23:29:55,940 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=82931.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:30:01,529 INFO [train.py:903] (3/4) Epoch 13, batch 1000, loss[loss=0.2379, simple_loss=0.312, pruned_loss=0.08192, over 19738.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3028, pruned_loss=0.07599, over 3796533.18 frames. ], batch size: 51, lr: 6.49e-03, grad_scale: 8.0 +2023-04-01 23:30:44,823 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82970.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 23:30:52,626 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-01 23:30:54,574 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.083e+02 5.161e+02 6.395e+02 8.326e+02 2.115e+03, threshold=1.279e+03, percent-clipped=2.0 +2023-04-01 23:30:57,031 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82981.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:31:02,726 INFO [train.py:903] (3/4) Epoch 13, batch 1050, loss[loss=0.2362, simple_loss=0.3133, pruned_loss=0.07951, over 18215.00 frames. ], tot_loss[loss=0.2286, simple_loss=0.3037, pruned_loss=0.07672, over 3786537.56 frames. ], batch size: 83, lr: 6.49e-03, grad_scale: 8.0 +2023-04-01 23:31:05,481 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9463, 1.2795, 1.6394, 1.0151, 2.4992, 3.2736, 2.9610, 3.4857], + device='cuda:3'), covar=tensor([0.1712, 0.3467, 0.3108, 0.2362, 0.0604, 0.0195, 0.0235, 0.0217], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0301, 0.0330, 0.0254, 0.0222, 0.0163, 0.0207, 0.0212], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:31:19,502 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0315, 1.3444, 1.6226, 0.5891, 2.0918, 2.4620, 2.1431, 2.6253], + device='cuda:3'), covar=tensor([0.1382, 0.3168, 0.2796, 0.2267, 0.0514, 0.0250, 0.0332, 0.0301], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0300, 0.0330, 0.0254, 0.0222, 0.0163, 0.0206, 0.0212], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:31:34,328 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-01 23:31:58,555 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-01 23:32:04,655 INFO [train.py:903] (3/4) Epoch 13, batch 1100, loss[loss=0.2459, simple_loss=0.32, pruned_loss=0.08589, over 19606.00 frames. ], tot_loss[loss=0.2294, simple_loss=0.3041, pruned_loss=0.07734, over 3802580.74 frames. ], batch size: 57, lr: 6.49e-03, grad_scale: 8.0 +2023-04-01 23:32:19,498 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83046.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:32:57,834 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.212e+02 5.014e+02 6.117e+02 7.878e+02 1.226e+03, threshold=1.223e+03, percent-clipped=0.0 +2023-04-01 23:32:58,250 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83079.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:33:08,488 INFO [train.py:903] (3/4) Epoch 13, batch 1150, loss[loss=0.2442, simple_loss=0.3232, pruned_loss=0.08255, over 19526.00 frames. ], tot_loss[loss=0.2292, simple_loss=0.3039, pruned_loss=0.07719, over 3786531.00 frames. ], batch size: 54, lr: 6.48e-03, grad_scale: 8.0 +2023-04-01 23:33:30,577 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83104.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:33:41,804 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1287, 1.7026, 1.6820, 2.0390, 1.9004, 1.8915, 1.7106, 1.9583], + device='cuda:3'), covar=tensor([0.0821, 0.1395, 0.1321, 0.0872, 0.1117, 0.0449, 0.1128, 0.0599], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0348, 0.0295, 0.0237, 0.0293, 0.0240, 0.0279, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:34:10,017 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7788, 4.3575, 2.6358, 3.8761, 0.8935, 4.2159, 4.1285, 4.1983], + device='cuda:3'), covar=tensor([0.0547, 0.0935, 0.2033, 0.0756, 0.4196, 0.0583, 0.0796, 0.1069], + device='cuda:3'), in_proj_covar=tensor([0.0447, 0.0377, 0.0448, 0.0323, 0.0387, 0.0379, 0.0374, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:34:10,932 INFO [train.py:903] (3/4) Epoch 13, batch 1200, loss[loss=0.2042, simple_loss=0.2883, pruned_loss=0.06009, over 19598.00 frames. ], tot_loss[loss=0.2284, simple_loss=0.3033, pruned_loss=0.07679, over 3790659.34 frames. ], batch size: 52, lr: 6.48e-03, grad_scale: 8.0 +2023-04-01 23:34:18,569 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2596, 1.3791, 1.6701, 1.4801, 2.4409, 2.0615, 2.5104, 0.9682], + device='cuda:3'), covar=tensor([0.2424, 0.4063, 0.2508, 0.2019, 0.1603, 0.2079, 0.1560, 0.4167], + device='cuda:3'), in_proj_covar=tensor([0.0489, 0.0577, 0.0612, 0.0438, 0.0594, 0.0495, 0.0641, 0.0493], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:34:22,355 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-01 23:34:40,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-01 23:34:40,970 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.47 vs. limit=5.0 +2023-04-01 23:35:06,466 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.490e+02 6.101e+02 7.395e+02 1.032e+03 1.939e+03, threshold=1.479e+03, percent-clipped=13.0 +2023-04-01 23:35:12,369 INFO [train.py:903] (3/4) Epoch 13, batch 1250, loss[loss=0.2521, simple_loss=0.3418, pruned_loss=0.08125, over 19614.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.3038, pruned_loss=0.07719, over 3798294.70 frames. ], batch size: 57, lr: 6.48e-03, grad_scale: 4.0 +2023-04-01 23:35:48,393 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83214.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:36:02,509 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83226.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 23:36:13,268 INFO [train.py:903] (3/4) Epoch 13, batch 1300, loss[loss=0.2545, simple_loss=0.3276, pruned_loss=0.09068, over 19554.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3052, pruned_loss=0.07837, over 3790792.53 frames. ], batch size: 61, lr: 6.48e-03, grad_scale: 4.0 +2023-04-01 23:36:33,765 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83251.0, num_to_drop=1, layers_to_drop={1} +2023-04-01 23:37:08,172 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.019e+02 5.042e+02 6.195e+02 7.677e+02 1.204e+03, threshold=1.239e+03, percent-clipped=0.0 +2023-04-01 23:37:17,150 INFO [train.py:903] (3/4) Epoch 13, batch 1350, loss[loss=0.2777, simple_loss=0.3467, pruned_loss=0.1044, over 19569.00 frames. ], tot_loss[loss=0.2298, simple_loss=0.3039, pruned_loss=0.07786, over 3794093.59 frames. ], batch size: 61, lr: 6.48e-03, grad_scale: 4.0 +2023-04-01 23:37:37,933 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:37:38,050 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83302.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:38:07,359 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=83325.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:38:09,911 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83327.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:38:21,068 INFO [train.py:903] (3/4) Epoch 13, batch 1400, loss[loss=0.1968, simple_loss=0.2739, pruned_loss=0.05988, over 18293.00 frames. ], tot_loss[loss=0.2292, simple_loss=0.3033, pruned_loss=0.07758, over 3798217.47 frames. ], batch size: 40, lr: 6.47e-03, grad_scale: 4.0 +2023-04-01 23:38:49,431 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1380, 2.2061, 2.4211, 3.1902, 2.1573, 3.1616, 2.6599, 2.2391], + device='cuda:3'), covar=tensor([0.3732, 0.3189, 0.1420, 0.2016, 0.3741, 0.1512, 0.3406, 0.2640], + device='cuda:3'), in_proj_covar=tensor([0.0809, 0.0832, 0.0659, 0.0892, 0.0796, 0.0723, 0.0797, 0.0720], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 23:39:10,620 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6206, 1.6679, 1.9469, 1.9604, 1.3519, 1.8603, 2.0635, 1.8399], + device='cuda:3'), covar=tensor([0.3696, 0.3095, 0.1570, 0.1912, 0.3496, 0.1708, 0.3980, 0.2779], + device='cuda:3'), in_proj_covar=tensor([0.0811, 0.0834, 0.0659, 0.0894, 0.0797, 0.0724, 0.0798, 0.0722], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-01 23:39:16,728 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.257e+02 5.533e+02 6.494e+02 7.603e+02 1.656e+03, threshold=1.299e+03, percent-clipped=2.0 +2023-04-01 23:39:20,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-01 23:39:22,623 INFO [train.py:903] (3/4) Epoch 13, batch 1450, loss[loss=0.2361, simple_loss=0.3093, pruned_loss=0.08143, over 19667.00 frames. ], tot_loss[loss=0.2308, simple_loss=0.3048, pruned_loss=0.07839, over 3807699.03 frames. ], batch size: 53, lr: 6.47e-03, grad_scale: 4.0 +2023-04-01 23:40:24,411 INFO [train.py:903] (3/4) Epoch 13, batch 1500, loss[loss=0.2531, simple_loss=0.3281, pruned_loss=0.08906, over 19779.00 frames. ], tot_loss[loss=0.2301, simple_loss=0.3044, pruned_loss=0.07796, over 3807270.81 frames. ], batch size: 56, lr: 6.47e-03, grad_scale: 4.0 +2023-04-01 23:40:29,243 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83440.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:41:19,835 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.552e+02 5.496e+02 6.437e+02 7.955e+02 2.023e+03, threshold=1.287e+03, percent-clipped=5.0 +2023-04-01 23:41:26,543 INFO [train.py:903] (3/4) Epoch 13, batch 1550, loss[loss=0.236, simple_loss=0.309, pruned_loss=0.0815, over 17630.00 frames. ], tot_loss[loss=0.2295, simple_loss=0.3041, pruned_loss=0.07749, over 3822058.25 frames. ], batch size: 101, lr: 6.47e-03, grad_scale: 4.0 +2023-04-01 23:42:30,077 INFO [train.py:903] (3/4) Epoch 13, batch 1600, loss[loss=0.2683, simple_loss=0.3373, pruned_loss=0.09959, over 19710.00 frames. ], tot_loss[loss=0.2289, simple_loss=0.3036, pruned_loss=0.07709, over 3830519.27 frames. ], batch size: 59, lr: 6.47e-03, grad_scale: 8.0 +2023-04-01 23:42:53,336 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-01 23:42:55,762 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=83558.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:43:25,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.823e+02 5.445e+02 6.386e+02 7.908e+02 1.256e+03, threshold=1.277e+03, percent-clipped=0.0 +2023-04-01 23:43:31,085 INFO [train.py:903] (3/4) Epoch 13, batch 1650, loss[loss=0.223, simple_loss=0.3044, pruned_loss=0.07083, over 19755.00 frames. ], tot_loss[loss=0.2272, simple_loss=0.302, pruned_loss=0.07625, over 3840790.31 frames. ], batch size: 54, lr: 6.47e-03, grad_scale: 8.0 +2023-04-01 23:43:53,242 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8624, 1.3401, 1.0570, 0.9073, 1.1846, 0.9543, 0.9042, 1.2587], + device='cuda:3'), covar=tensor([0.0590, 0.0755, 0.1081, 0.0681, 0.0522, 0.1205, 0.0541, 0.0448], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0305, 0.0324, 0.0247, 0.0237, 0.0320, 0.0284, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:44:33,594 INFO [train.py:903] (3/4) Epoch 13, batch 1700, loss[loss=0.2236, simple_loss=0.3116, pruned_loss=0.06787, over 19680.00 frames. ], tot_loss[loss=0.228, simple_loss=0.303, pruned_loss=0.07656, over 3839554.50 frames. ], batch size: 60, lr: 6.46e-03, grad_scale: 8.0 +2023-04-01 23:44:45,676 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=83646.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:45:15,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-01 23:45:19,391 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83673.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:45:27,804 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.234e+02 5.386e+02 6.785e+02 9.131e+02 1.645e+03, threshold=1.357e+03, percent-clipped=5.0 +2023-04-01 23:45:33,655 INFO [train.py:903] (3/4) Epoch 13, batch 1750, loss[loss=0.2024, simple_loss=0.271, pruned_loss=0.06692, over 19058.00 frames. ], tot_loss[loss=0.2277, simple_loss=0.3026, pruned_loss=0.07644, over 3827505.17 frames. ], batch size: 42, lr: 6.46e-03, grad_scale: 8.0 +2023-04-01 23:45:48,950 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83696.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:45:53,567 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5187, 1.6850, 2.0084, 1.7234, 2.5791, 2.9729, 2.8937, 3.1034], + device='cuda:3'), covar=tensor([0.1361, 0.2717, 0.2403, 0.1926, 0.0940, 0.0274, 0.0211, 0.0304], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0298, 0.0326, 0.0251, 0.0218, 0.0161, 0.0205, 0.0210], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:46:19,396 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83721.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:46:37,927 INFO [train.py:903] (3/4) Epoch 13, batch 1800, loss[loss=0.2847, simple_loss=0.3396, pruned_loss=0.1149, over 13479.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3014, pruned_loss=0.07637, over 3804313.75 frames. ], batch size: 136, lr: 6.46e-03, grad_scale: 8.0 +2023-04-01 23:47:08,620 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83761.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:47:29,808 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83778.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:47:32,955 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.146e+02 5.542e+02 6.939e+02 8.095e+02 2.139e+03, threshold=1.388e+03, percent-clipped=3.0 +2023-04-01 23:47:35,283 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-01 23:47:39,813 INFO [train.py:903] (3/4) Epoch 13, batch 1850, loss[loss=0.2759, simple_loss=0.3337, pruned_loss=0.1091, over 13020.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3016, pruned_loss=0.07632, over 3803676.87 frames. ], batch size: 135, lr: 6.46e-03, grad_scale: 8.0 +2023-04-01 23:48:11,509 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-01 23:48:38,899 INFO [train.py:903] (3/4) Epoch 13, batch 1900, loss[loss=0.2468, simple_loss=0.3155, pruned_loss=0.08899, over 19605.00 frames. ], tot_loss[loss=0.229, simple_loss=0.3034, pruned_loss=0.07735, over 3813446.50 frames. ], batch size: 57, lr: 6.46e-03, grad_scale: 8.0 +2023-04-01 23:48:56,220 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-01 23:49:00,754 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-01 23:49:22,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9501, 2.5680, 1.8140, 1.9338, 2.4309, 1.6905, 1.6567, 2.0131], + device='cuda:3'), covar=tensor([0.0895, 0.0672, 0.0755, 0.0596, 0.0404, 0.0878, 0.0568, 0.0543], + device='cuda:3'), in_proj_covar=tensor([0.0291, 0.0303, 0.0325, 0.0248, 0.0237, 0.0317, 0.0284, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:49:23,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-01 23:49:32,948 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.547e+02 5.453e+02 6.640e+02 7.751e+02 1.927e+03, threshold=1.328e+03, percent-clipped=4.0 +2023-04-01 23:49:38,629 INFO [train.py:903] (3/4) Epoch 13, batch 1950, loss[loss=0.2107, simple_loss=0.2814, pruned_loss=0.06996, over 19571.00 frames. ], tot_loss[loss=0.2283, simple_loss=0.3027, pruned_loss=0.07699, over 3814890.45 frames. ], batch size: 52, lr: 6.45e-03, grad_scale: 8.0 +2023-04-01 23:50:31,106 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83929.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:50:40,968 INFO [train.py:903] (3/4) Epoch 13, batch 2000, loss[loss=0.2271, simple_loss=0.3053, pruned_loss=0.0744, over 19564.00 frames. ], tot_loss[loss=0.2283, simple_loss=0.3027, pruned_loss=0.07693, over 3806363.85 frames. ], batch size: 61, lr: 6.45e-03, grad_scale: 8.0 +2023-04-01 23:51:02,420 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83954.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:51:17,807 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83967.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:51:36,105 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.295e+02 5.067e+02 6.527e+02 8.467e+02 1.955e+03, threshold=1.305e+03, percent-clipped=7.0 +2023-04-01 23:51:38,312 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-01 23:51:42,661 INFO [train.py:903] (3/4) Epoch 13, batch 2050, loss[loss=0.28, simple_loss=0.347, pruned_loss=0.1065, over 17548.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3039, pruned_loss=0.07733, over 3812712.85 frames. ], batch size: 101, lr: 6.45e-03, grad_scale: 8.0 +2023-04-01 23:51:48,289 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.10 vs. limit=2.0 +2023-04-01 23:51:56,832 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-01 23:51:57,786 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-01 23:52:03,662 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84003.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:52:16,701 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-04-01 23:52:21,311 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-01 23:52:22,839 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84017.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:52:43,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2224, 1.2654, 1.1784, 0.9979, 1.0311, 0.9849, 0.0374, 0.2991], + device='cuda:3'), covar=tensor([0.0497, 0.0489, 0.0331, 0.0411, 0.1029, 0.0455, 0.0952, 0.0842], + device='cuda:3'), in_proj_covar=tensor([0.0340, 0.0333, 0.0334, 0.0356, 0.0425, 0.0357, 0.0313, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-01 23:52:44,647 INFO [train.py:903] (3/4) Epoch 13, batch 2100, loss[loss=0.276, simple_loss=0.3455, pruned_loss=0.1033, over 19651.00 frames. ], tot_loss[loss=0.2302, simple_loss=0.3046, pruned_loss=0.07786, over 3825947.40 frames. ], batch size: 58, lr: 6.45e-03, grad_scale: 8.0 +2023-04-01 23:52:52,275 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84042.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:53:14,864 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-01 23:53:36,201 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-01 23:53:39,576 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.864e+02 5.545e+02 6.946e+02 9.457e+02 3.064e+03, threshold=1.389e+03, percent-clipped=12.0 +2023-04-01 23:53:45,274 INFO [train.py:903] (3/4) Epoch 13, batch 2150, loss[loss=0.2248, simple_loss=0.3074, pruned_loss=0.07105, over 18699.00 frames. ], tot_loss[loss=0.2306, simple_loss=0.3047, pruned_loss=0.07824, over 3836359.31 frames. ], batch size: 74, lr: 6.45e-03, grad_scale: 8.0 +2023-04-01 23:54:23,402 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84115.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:54:31,390 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84122.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:54:49,428 INFO [train.py:903] (3/4) Epoch 13, batch 2200, loss[loss=0.2666, simple_loss=0.3427, pruned_loss=0.09518, over 19587.00 frames. ], tot_loss[loss=0.228, simple_loss=0.3024, pruned_loss=0.07682, over 3851528.32 frames. ], batch size: 61, lr: 6.44e-03, grad_scale: 8.0 +2023-04-01 23:55:44,487 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.166e+02 5.936e+02 7.647e+02 9.699e+02 2.302e+03, threshold=1.529e+03, percent-clipped=8.0 +2023-04-01 23:55:50,242 INFO [train.py:903] (3/4) Epoch 13, batch 2250, loss[loss=0.2442, simple_loss=0.3168, pruned_loss=0.08584, over 19378.00 frames. ], tot_loss[loss=0.2286, simple_loss=0.3028, pruned_loss=0.07718, over 3835650.35 frames. ], batch size: 66, lr: 6.44e-03, grad_scale: 8.0 +2023-04-01 23:56:29,751 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84217.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:56:51,979 INFO [train.py:903] (3/4) Epoch 13, batch 2300, loss[loss=0.1918, simple_loss=0.2726, pruned_loss=0.05551, over 19572.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.3018, pruned_loss=0.07693, over 3831970.28 frames. ], batch size: 52, lr: 6.44e-03, grad_scale: 8.0 +2023-04-01 23:56:53,522 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84237.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:57:05,857 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-01 23:57:15,129 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84254.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:57:47,149 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.862e+02 5.192e+02 6.483e+02 8.696e+02 2.103e+03, threshold=1.297e+03, percent-clipped=4.0 +2023-04-01 23:57:49,826 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5713, 1.1610, 1.3837, 1.2777, 2.2420, 0.9874, 2.0339, 2.4420], + device='cuda:3'), covar=tensor([0.0579, 0.2585, 0.2498, 0.1459, 0.0795, 0.1860, 0.0875, 0.0428], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0348, 0.0357, 0.0326, 0.0351, 0.0335, 0.0346, 0.0368], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-01 23:57:52,885 INFO [train.py:903] (3/4) Epoch 13, batch 2350, loss[loss=0.2306, simple_loss=0.3048, pruned_loss=0.0782, over 18740.00 frames. ], tot_loss[loss=0.2283, simple_loss=0.3028, pruned_loss=0.07687, over 3828414.08 frames. ], batch size: 74, lr: 6.44e-03, grad_scale: 8.0 +2023-04-01 23:58:25,969 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84311.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:58:37,171 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-01 23:58:54,358 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-01 23:58:57,898 INFO [train.py:903] (3/4) Epoch 13, batch 2400, loss[loss=0.1828, simple_loss=0.2679, pruned_loss=0.04889, over 19474.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3022, pruned_loss=0.07631, over 3827103.76 frames. ], batch size: 49, lr: 6.44e-03, grad_scale: 8.0 +2023-04-01 23:59:11,491 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84347.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:59:20,759 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84354.0, num_to_drop=0, layers_to_drop=set() +2023-04-01 23:59:54,316 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.264e+02 5.138e+02 6.932e+02 8.383e+02 1.660e+03, threshold=1.386e+03, percent-clipped=4.0 +2023-04-01 23:59:59,973 INFO [train.py:903] (3/4) Epoch 13, batch 2450, loss[loss=0.2739, simple_loss=0.3492, pruned_loss=0.09933, over 19690.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.3047, pruned_loss=0.07738, over 3827055.28 frames. ], batch size: 59, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:00:22,351 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2880, 2.2519, 2.4796, 3.2689, 2.2373, 3.2027, 2.9028, 2.3799], + device='cuda:3'), covar=tensor([0.3832, 0.3503, 0.1455, 0.1996, 0.3865, 0.1585, 0.3371, 0.2857], + device='cuda:3'), in_proj_covar=tensor([0.0811, 0.0841, 0.0662, 0.0898, 0.0796, 0.0724, 0.0798, 0.0726], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:00:31,662 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5540, 2.0133, 2.1815, 2.6861, 2.5164, 2.3570, 2.1217, 2.6662], + device='cuda:3'), covar=tensor([0.0820, 0.1744, 0.1267, 0.0941, 0.1207, 0.0453, 0.1080, 0.0552], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0351, 0.0294, 0.0239, 0.0295, 0.0243, 0.0280, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:00:50,511 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5471, 1.6245, 2.0317, 1.8345, 3.6058, 2.8566, 3.7268, 1.7895], + device='cuda:3'), covar=tensor([0.2166, 0.3893, 0.2408, 0.1660, 0.1166, 0.1723, 0.1386, 0.3365], + device='cuda:3'), in_proj_covar=tensor([0.0498, 0.0591, 0.0624, 0.0443, 0.0601, 0.0503, 0.0650, 0.0504], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:00:51,597 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84426.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:01:03,732 INFO [train.py:903] (3/4) Epoch 13, batch 2500, loss[loss=0.2244, simple_loss=0.3013, pruned_loss=0.07369, over 19758.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.3054, pruned_loss=0.07796, over 3830621.63 frames. ], batch size: 51, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:01:34,297 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84459.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:01:38,081 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84462.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:01:46,358 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5190, 3.7493, 4.0356, 4.0250, 2.2412, 3.7281, 3.4749, 3.8049], + device='cuda:3'), covar=tensor([0.1153, 0.1961, 0.0522, 0.0563, 0.3818, 0.1048, 0.0489, 0.0851], + device='cuda:3'), in_proj_covar=tensor([0.0693, 0.0621, 0.0827, 0.0705, 0.0748, 0.0572, 0.0498, 0.0757], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 00:01:47,594 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9779, 3.6112, 2.5007, 3.3065, 0.9878, 3.4638, 3.4168, 3.5142], + device='cuda:3'), covar=tensor([0.0784, 0.1177, 0.1966, 0.0835, 0.3764, 0.0803, 0.0887, 0.1078], + device='cuda:3'), in_proj_covar=tensor([0.0444, 0.0371, 0.0445, 0.0319, 0.0382, 0.0377, 0.0370, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:01:47,714 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1215, 1.2832, 1.4827, 1.3652, 2.6893, 1.0454, 2.0509, 2.9758], + device='cuda:3'), covar=tensor([0.0467, 0.2498, 0.2580, 0.1760, 0.0756, 0.2165, 0.1098, 0.0333], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0345, 0.0356, 0.0325, 0.0350, 0.0334, 0.0342, 0.0367], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:01:55,290 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1488, 2.0126, 2.0327, 2.3700, 2.1924, 1.9303, 2.0137, 2.2202], + device='cuda:3'), covar=tensor([0.0717, 0.1213, 0.1002, 0.0610, 0.0904, 0.0422, 0.0918, 0.0472], + device='cuda:3'), in_proj_covar=tensor([0.0254, 0.0350, 0.0294, 0.0238, 0.0295, 0.0242, 0.0279, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:02:00,517 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.656e+02 5.326e+02 7.100e+02 9.098e+02 1.657e+03, threshold=1.420e+03, percent-clipped=3.0 +2023-04-02 00:02:06,422 INFO [train.py:903] (3/4) Epoch 13, batch 2550, loss[loss=0.2391, simple_loss=0.3111, pruned_loss=0.08359, over 19587.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3051, pruned_loss=0.07828, over 3820292.51 frames. ], batch size: 52, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:02:16,070 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84493.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:02:17,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6355, 1.6036, 1.8127, 1.7077, 4.1697, 0.9748, 2.4739, 4.3502], + device='cuda:3'), covar=tensor([0.0397, 0.2590, 0.2592, 0.1915, 0.0701, 0.2771, 0.1419, 0.0234], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0345, 0.0355, 0.0324, 0.0348, 0.0333, 0.0341, 0.0367], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:02:48,834 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84518.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:03:04,537 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 00:03:10,382 INFO [train.py:903] (3/4) Epoch 13, batch 2600, loss[loss=0.254, simple_loss=0.3369, pruned_loss=0.0855, over 19661.00 frames. ], tot_loss[loss=0.2304, simple_loss=0.3047, pruned_loss=0.07804, over 3822421.11 frames. ], batch size: 55, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:03:21,100 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84543.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:03:42,743 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84561.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:03:59,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84574.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:04:09,372 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.575e+02 5.097e+02 6.359e+02 8.045e+02 2.004e+03, threshold=1.272e+03, percent-clipped=4.0 +2023-04-02 00:04:15,152 INFO [train.py:903] (3/4) Epoch 13, batch 2650, loss[loss=0.2783, simple_loss=0.3416, pruned_loss=0.1075, over 19767.00 frames. ], tot_loss[loss=0.2306, simple_loss=0.3048, pruned_loss=0.07817, over 3823888.05 frames. ], batch size: 63, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:04:26,074 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9125, 2.0443, 2.1270, 2.6958, 1.9002, 2.5730, 2.3700, 2.0747], + device='cuda:3'), covar=tensor([0.3503, 0.3004, 0.1546, 0.1757, 0.3411, 0.1541, 0.3535, 0.2625], + device='cuda:3'), in_proj_covar=tensor([0.0817, 0.0845, 0.0667, 0.0902, 0.0801, 0.0731, 0.0803, 0.0727], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:04:30,351 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84598.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:04:34,949 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 00:05:17,047 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-04-02 00:05:17,486 INFO [train.py:903] (3/4) Epoch 13, batch 2700, loss[loss=0.2251, simple_loss=0.3003, pruned_loss=0.07498, over 19540.00 frames. ], tot_loss[loss=0.23, simple_loss=0.3046, pruned_loss=0.07769, over 3830997.17 frames. ], batch size: 54, lr: 6.43e-03, grad_scale: 8.0 +2023-04-02 00:05:36,190 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84651.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:06:08,610 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5196, 1.2631, 1.6901, 1.3766, 2.6961, 3.7883, 3.4966, 4.0233], + device='cuda:3'), covar=tensor([0.1341, 0.3514, 0.3090, 0.2158, 0.0551, 0.0159, 0.0190, 0.0174], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0298, 0.0324, 0.0250, 0.0217, 0.0162, 0.0205, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:06:08,623 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84676.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:06:14,002 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.791e+02 5.394e+02 6.395e+02 8.456e+02 1.799e+03, threshold=1.279e+03, percent-clipped=4.0 +2023-04-02 00:06:16,795 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84682.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:06:21,011 INFO [train.py:903] (3/4) Epoch 13, batch 2750, loss[loss=0.2266, simple_loss=0.31, pruned_loss=0.07162, over 19783.00 frames. ], tot_loss[loss=0.2289, simple_loss=0.3035, pruned_loss=0.07718, over 3841478.27 frames. ], batch size: 56, lr: 6.42e-03, grad_scale: 8.0 +2023-04-02 00:06:37,475 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84698.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:06:44,851 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-02 00:06:49,105 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84707.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:06:56,179 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84713.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:07:02,029 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84718.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:07:22,612 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3613, 1.4228, 1.7438, 1.5742, 2.6400, 2.3022, 2.7981, 1.1950], + device='cuda:3'), covar=tensor([0.2306, 0.3965, 0.2523, 0.1770, 0.1369, 0.1861, 0.1397, 0.3736], + device='cuda:3'), in_proj_covar=tensor([0.0498, 0.0588, 0.0624, 0.0442, 0.0600, 0.0500, 0.0649, 0.0502], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:07:24,546 INFO [train.py:903] (3/4) Epoch 13, batch 2800, loss[loss=0.2129, simple_loss=0.2977, pruned_loss=0.06407, over 19338.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3037, pruned_loss=0.07744, over 3824179.46 frames. ], batch size: 66, lr: 6.42e-03, grad_scale: 8.0 +2023-04-02 00:07:35,151 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84743.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:08:22,666 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.267e+02 5.423e+02 6.884e+02 8.957e+02 1.568e+03, threshold=1.377e+03, percent-clipped=4.0 +2023-04-02 00:08:27,029 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.54 vs. limit=2.0 +2023-04-02 00:08:29,874 INFO [train.py:903] (3/4) Epoch 13, batch 2850, loss[loss=0.2664, simple_loss=0.3363, pruned_loss=0.09821, over 19768.00 frames. ], tot_loss[loss=0.2309, simple_loss=0.3048, pruned_loss=0.07854, over 3807882.34 frames. ], batch size: 63, lr: 6.42e-03, grad_scale: 8.0 +2023-04-02 00:09:04,392 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84813.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:09:11,246 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84818.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:09:26,676 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84830.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:09:33,386 INFO [train.py:903] (3/4) Epoch 13, batch 2900, loss[loss=0.1896, simple_loss=0.2776, pruned_loss=0.05084, over 19760.00 frames. ], tot_loss[loss=0.2301, simple_loss=0.3039, pruned_loss=0.07814, over 3806613.19 frames. ], batch size: 54, lr: 6.42e-03, grad_scale: 8.0 +2023-04-02 00:09:33,421 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 00:09:43,376 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1885, 1.2678, 1.1754, 0.9851, 1.0574, 1.0773, 0.0621, 0.3643], + device='cuda:3'), covar=tensor([0.0520, 0.0535, 0.0339, 0.0457, 0.1036, 0.0477, 0.0984, 0.0828], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0333, 0.0335, 0.0356, 0.0429, 0.0356, 0.0314, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:09:58,778 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84855.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:10:31,987 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.833e+02 5.191e+02 6.710e+02 8.572e+02 2.238e+03, threshold=1.342e+03, percent-clipped=4.0 +2023-04-02 00:10:38,049 INFO [train.py:903] (3/4) Epoch 13, batch 2950, loss[loss=0.2287, simple_loss=0.3056, pruned_loss=0.07588, over 19577.00 frames. ], tot_loss[loss=0.2296, simple_loss=0.3038, pruned_loss=0.07772, over 3811931.11 frames. ], batch size: 61, lr: 6.42e-03, grad_scale: 8.0 +2023-04-02 00:10:40,481 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84887.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:11:39,659 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84932.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:11:45,164 INFO [train.py:903] (3/4) Epoch 13, batch 3000, loss[loss=0.2016, simple_loss=0.2853, pruned_loss=0.05893, over 19674.00 frames. ], tot_loss[loss=0.2279, simple_loss=0.3025, pruned_loss=0.07668, over 3807401.38 frames. ], batch size: 53, lr: 6.41e-03, grad_scale: 8.0 +2023-04-02 00:11:45,164 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 00:12:00,828 INFO [train.py:937] (3/4) Epoch 13, validation: loss=0.1754, simple_loss=0.276, pruned_loss=0.03742, over 944034.00 frames. +2023-04-02 00:12:00,829 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 00:12:05,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 00:12:29,848 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84957.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:12:43,850 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84969.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:12:43,984 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84969.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:12:59,475 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.728e+02 4.901e+02 6.331e+02 8.447e+02 1.208e+03, threshold=1.266e+03, percent-clipped=0.0 +2023-04-02 00:13:05,452 INFO [train.py:903] (3/4) Epoch 13, batch 3050, loss[loss=0.2378, simple_loss=0.3122, pruned_loss=0.08173, over 19722.00 frames. ], tot_loss[loss=0.2286, simple_loss=0.3034, pruned_loss=0.07691, over 3816171.98 frames. ], batch size: 63, lr: 6.41e-03, grad_scale: 8.0 +2023-04-02 00:13:17,195 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84994.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:13:18,157 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=84995.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:13:26,442 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85002.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:14:09,057 INFO [train.py:903] (3/4) Epoch 13, batch 3100, loss[loss=0.2172, simple_loss=0.2895, pruned_loss=0.07246, over 19738.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.3035, pruned_loss=0.07739, over 3813975.46 frames. ], batch size: 51, lr: 6.41e-03, grad_scale: 8.0 +2023-04-02 00:14:50,660 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85069.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:15:05,350 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.206e+02 5.118e+02 6.727e+02 8.323e+02 1.616e+03, threshold=1.345e+03, percent-clipped=6.0 +2023-04-02 00:15:11,220 INFO [train.py:903] (3/4) Epoch 13, batch 3150, loss[loss=0.2144, simple_loss=0.2947, pruned_loss=0.06706, over 19849.00 frames. ], tot_loss[loss=0.2281, simple_loss=0.3028, pruned_loss=0.07671, over 3820402.25 frames. ], batch size: 52, lr: 6.41e-03, grad_scale: 8.0 +2023-04-02 00:15:20,649 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:15:37,410 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 00:15:40,757 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3642, 1.4574, 1.7475, 1.4262, 2.3266, 2.7780, 2.5414, 2.8696], + device='cuda:3'), covar=tensor([0.1331, 0.2773, 0.2433, 0.2093, 0.1023, 0.0245, 0.0276, 0.0333], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0300, 0.0329, 0.0251, 0.0218, 0.0164, 0.0206, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:15:41,868 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85110.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:16:00,162 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85125.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:16:09,398 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85133.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:16:12,575 INFO [train.py:903] (3/4) Epoch 13, batch 3200, loss[loss=0.2165, simple_loss=0.299, pruned_loss=0.06705, over 18024.00 frames. ], tot_loss[loss=0.2297, simple_loss=0.3038, pruned_loss=0.07779, over 3804594.85 frames. ], batch size: 83, lr: 6.41e-03, grad_scale: 8.0 +2023-04-02 00:16:25,412 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1701, 2.2490, 2.5120, 3.2114, 2.2179, 3.1491, 2.7129, 2.3575], + device='cuda:3'), covar=tensor([0.3992, 0.3607, 0.1466, 0.2012, 0.3982, 0.1572, 0.3644, 0.2707], + device='cuda:3'), in_proj_covar=tensor([0.0812, 0.0842, 0.0660, 0.0895, 0.0797, 0.0724, 0.0795, 0.0721], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:16:46,262 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85162.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:17:10,848 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.817e+02 5.513e+02 6.647e+02 8.266e+02 1.326e+03, threshold=1.329e+03, percent-clipped=0.0 +2023-04-02 00:17:16,637 INFO [train.py:903] (3/4) Epoch 13, batch 3250, loss[loss=0.236, simple_loss=0.3127, pruned_loss=0.07965, over 19653.00 frames. ], tot_loss[loss=0.2283, simple_loss=0.3026, pruned_loss=0.07697, over 3809508.54 frames. ], batch size: 60, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:17:58,815 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5162, 1.1394, 1.1594, 1.4089, 1.0869, 1.2931, 1.0718, 1.3395], + device='cuda:3'), covar=tensor([0.1124, 0.1385, 0.1529, 0.0975, 0.1275, 0.0640, 0.1462, 0.0818], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0353, 0.0295, 0.0241, 0.0296, 0.0246, 0.0285, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:18:20,657 INFO [train.py:903] (3/4) Epoch 13, batch 3300, loss[loss=0.1907, simple_loss=0.2744, pruned_loss=0.05355, over 19461.00 frames. ], tot_loss[loss=0.2275, simple_loss=0.3017, pruned_loss=0.0766, over 3798968.01 frames. ], batch size: 49, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:18:21,909 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 00:18:34,734 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3047, 1.4360, 1.8076, 1.5709, 3.0883, 2.3978, 3.3276, 1.5222], + device='cuda:3'), covar=tensor([0.2388, 0.4073, 0.2468, 0.1930, 0.1470, 0.2002, 0.1576, 0.3764], + device='cuda:3'), in_proj_covar=tensor([0.0493, 0.0584, 0.0619, 0.0438, 0.0597, 0.0499, 0.0644, 0.0502], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:18:39,428 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5265, 1.2194, 1.4909, 1.4947, 3.0742, 0.9544, 2.1799, 3.5219], + device='cuda:3'), covar=tensor([0.0452, 0.2907, 0.2741, 0.1878, 0.0703, 0.2726, 0.1419, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0350, 0.0357, 0.0325, 0.0348, 0.0334, 0.0343, 0.0365], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:18:47,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85258.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:18:53,204 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4803, 2.2937, 1.6451, 1.6359, 2.1032, 1.2934, 1.2446, 1.8053], + device='cuda:3'), covar=tensor([0.0971, 0.0672, 0.0999, 0.0709, 0.0491, 0.1139, 0.0734, 0.0490], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0304, 0.0328, 0.0250, 0.0238, 0.0321, 0.0287, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:19:00,110 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-02 00:19:12,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85277.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:19:17,835 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.152e+02 5.269e+02 6.447e+02 8.165e+02 1.494e+03, threshold=1.289e+03, percent-clipped=3.0 +2023-04-02 00:19:19,510 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85283.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:19:22,467 INFO [train.py:903] (3/4) Epoch 13, batch 3350, loss[loss=0.2512, simple_loss=0.3315, pruned_loss=0.08544, over 19674.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3019, pruned_loss=0.07642, over 3817575.40 frames. ], batch size: 59, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:19:57,334 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85313.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:20:21,525 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3363, 1.4871, 1.8468, 1.4873, 2.7666, 3.5341, 3.3736, 3.7211], + device='cuda:3'), covar=tensor([0.1622, 0.3325, 0.2980, 0.2182, 0.0602, 0.0196, 0.0192, 0.0242], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0301, 0.0329, 0.0252, 0.0220, 0.0165, 0.0206, 0.0213], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:20:24,674 INFO [train.py:903] (3/4) Epoch 13, batch 3400, loss[loss=0.1983, simple_loss=0.2748, pruned_loss=0.0609, over 19623.00 frames. ], tot_loss[loss=0.2275, simple_loss=0.3019, pruned_loss=0.07656, over 3810478.79 frames. ], batch size: 50, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:20:57,121 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2250, 1.2759, 1.1704, 0.9777, 0.9561, 1.0362, 0.1194, 0.3209], + device='cuda:3'), covar=tensor([0.0680, 0.0652, 0.0421, 0.0516, 0.1437, 0.0650, 0.1137, 0.1158], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0334, 0.0335, 0.0357, 0.0428, 0.0357, 0.0313, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:21:03,105 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85366.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:21:22,538 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.171e+02 5.371e+02 6.826e+02 8.500e+02 2.504e+03, threshold=1.365e+03, percent-clipped=9.0 +2023-04-02 00:21:27,160 INFO [train.py:903] (3/4) Epoch 13, batch 3450, loss[loss=0.2334, simple_loss=0.3079, pruned_loss=0.07947, over 17620.00 frames. ], tot_loss[loss=0.228, simple_loss=0.3023, pruned_loss=0.07688, over 3811450.74 frames. ], batch size: 101, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:21:30,654 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 00:21:34,227 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85391.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:22:19,381 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85428.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:22:28,802 INFO [train.py:903] (3/4) Epoch 13, batch 3500, loss[loss=0.2056, simple_loss=0.2737, pruned_loss=0.06876, over 19383.00 frames. ], tot_loss[loss=0.2282, simple_loss=0.3025, pruned_loss=0.07697, over 3813240.66 frames. ], batch size: 47, lr: 6.40e-03, grad_scale: 8.0 +2023-04-02 00:22:42,065 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85446.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:23:10,410 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85469.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:23:20,670 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:23:27,193 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.349e+02 5.339e+02 6.675e+02 8.042e+02 1.644e+03, threshold=1.335e+03, percent-clipped=3.0 +2023-04-02 00:23:31,539 INFO [train.py:903] (3/4) Epoch 13, batch 3550, loss[loss=0.2565, simple_loss=0.3299, pruned_loss=0.09152, over 19542.00 frames. ], tot_loss[loss=0.2281, simple_loss=0.3027, pruned_loss=0.07677, over 3823980.20 frames. ], batch size: 56, lr: 6.39e-03, grad_scale: 8.0 +2023-04-02 00:23:31,994 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0048, 1.6943, 1.5950, 1.8670, 1.6826, 1.7973, 1.4561, 1.8770], + device='cuda:3'), covar=tensor([0.0925, 0.1393, 0.1436, 0.1038, 0.1251, 0.0478, 0.1350, 0.0696], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0348, 0.0294, 0.0238, 0.0294, 0.0241, 0.0281, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:23:50,612 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85501.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:24:04,313 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3159, 1.3414, 1.5745, 1.5215, 2.1173, 2.0334, 2.1737, 0.7401], + device='cuda:3'), covar=tensor([0.2442, 0.4149, 0.2561, 0.1927, 0.1541, 0.2096, 0.1474, 0.4383], + device='cuda:3'), in_proj_covar=tensor([0.0495, 0.0588, 0.0623, 0.0439, 0.0598, 0.0501, 0.0649, 0.0504], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:24:31,256 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85533.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:24:34,224 INFO [train.py:903] (3/4) Epoch 13, batch 3600, loss[loss=0.2578, simple_loss=0.3286, pruned_loss=0.09347, over 17664.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.3027, pruned_loss=0.07644, over 3829463.67 frames. ], batch size: 101, lr: 6.39e-03, grad_scale: 8.0 +2023-04-02 00:24:50,374 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85548.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:25:04,582 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85558.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:25:25,618 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8901, 1.1619, 1.4621, 0.5913, 2.1217, 2.4516, 2.1509, 2.5838], + device='cuda:3'), covar=tensor([0.1616, 0.3509, 0.3229, 0.2483, 0.0519, 0.0227, 0.0331, 0.0292], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0300, 0.0329, 0.0251, 0.0220, 0.0164, 0.0205, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:25:33,193 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.573e+02 5.969e+02 7.367e+02 9.197e+02 1.857e+03, threshold=1.473e+03, percent-clipped=2.0 +2023-04-02 00:25:36,809 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85584.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:25:38,625 INFO [train.py:903] (3/4) Epoch 13, batch 3650, loss[loss=0.2186, simple_loss=0.2982, pruned_loss=0.06947, over 19594.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.3027, pruned_loss=0.07651, over 3837707.71 frames. ], batch size: 52, lr: 6.39e-03, grad_scale: 8.0 +2023-04-02 00:25:46,147 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85592.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:26:39,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2462, 1.3530, 2.0184, 1.6219, 2.9154, 4.6407, 4.5529, 4.9461], + device='cuda:3'), covar=tensor([0.1601, 0.3492, 0.3004, 0.1937, 0.0516, 0.0134, 0.0139, 0.0141], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0300, 0.0331, 0.0251, 0.0220, 0.0164, 0.0206, 0.0211], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:26:41,858 INFO [train.py:903] (3/4) Epoch 13, batch 3700, loss[loss=0.2789, simple_loss=0.3543, pruned_loss=0.1017, over 19501.00 frames. ], tot_loss[loss=0.2286, simple_loss=0.3031, pruned_loss=0.07703, over 3828938.29 frames. ], batch size: 64, lr: 6.39e-03, grad_scale: 8.0 +2023-04-02 00:26:51,261 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85643.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:27:41,800 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.698e+02 5.337e+02 6.385e+02 1.025e+03 1.989e+03, threshold=1.277e+03, percent-clipped=4.0 +2023-04-02 00:27:44,677 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85684.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:27:46,550 INFO [train.py:903] (3/4) Epoch 13, batch 3750, loss[loss=0.2998, simple_loss=0.3607, pruned_loss=0.1195, over 17086.00 frames. ], tot_loss[loss=0.2272, simple_loss=0.302, pruned_loss=0.07615, over 3837311.71 frames. ], batch size: 101, lr: 6.39e-03, grad_scale: 8.0 +2023-04-02 00:28:01,531 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4554, 1.5405, 2.1680, 1.7981, 3.0394, 2.4070, 3.2725, 1.6572], + device='cuda:3'), covar=tensor([0.2564, 0.4368, 0.2516, 0.1973, 0.1690, 0.2312, 0.1880, 0.3913], + device='cuda:3'), in_proj_covar=tensor([0.0497, 0.0590, 0.0625, 0.0439, 0.0598, 0.0502, 0.0648, 0.0505], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:28:14,067 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 00:28:14,943 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85709.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:28:34,773 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5046, 1.5421, 1.8513, 1.6996, 3.3584, 2.7017, 3.5751, 1.4467], + device='cuda:3'), covar=tensor([0.2206, 0.3961, 0.2526, 0.1701, 0.1276, 0.1772, 0.1369, 0.3736], + device='cuda:3'), in_proj_covar=tensor([0.0496, 0.0589, 0.0625, 0.0439, 0.0597, 0.0501, 0.0648, 0.0505], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:28:48,156 INFO [train.py:903] (3/4) Epoch 13, batch 3800, loss[loss=0.2215, simple_loss=0.2988, pruned_loss=0.07216, over 19668.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.3009, pruned_loss=0.07539, over 3832853.56 frames. ], batch size: 53, lr: 6.38e-03, grad_scale: 8.0 +2023-04-02 00:29:20,372 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-02 00:29:20,848 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 00:29:44,888 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.310e+02 5.587e+02 7.205e+02 9.194e+02 2.888e+03, threshold=1.441e+03, percent-clipped=8.0 +2023-04-02 00:29:50,626 INFO [train.py:903] (3/4) Epoch 13, batch 3850, loss[loss=0.1896, simple_loss=0.2661, pruned_loss=0.05652, over 19477.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.302, pruned_loss=0.0764, over 3823675.58 frames. ], batch size: 49, lr: 6.38e-03, grad_scale: 8.0 +2023-04-02 00:29:56,236 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85790.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:30:53,504 INFO [train.py:903] (3/4) Epoch 13, batch 3900, loss[loss=0.2147, simple_loss=0.2912, pruned_loss=0.06911, over 19573.00 frames. ], tot_loss[loss=0.2284, simple_loss=0.3029, pruned_loss=0.07694, over 3831473.62 frames. ], batch size: 52, lr: 6.38e-03, grad_scale: 8.0 +2023-04-02 00:31:00,076 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85840.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:31:06,825 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85845.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:31:10,509 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:31:30,227 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85865.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:31:40,504 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85873.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:31:51,442 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.894e+02 5.450e+02 6.738e+02 8.252e+02 2.044e+03, threshold=1.348e+03, percent-clipped=4.0 +2023-04-02 00:31:57,035 INFO [train.py:903] (3/4) Epoch 13, batch 3950, loss[loss=0.2292, simple_loss=0.2944, pruned_loss=0.08205, over 19801.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.3034, pruned_loss=0.07736, over 3840668.46 frames. ], batch size: 49, lr: 6.38e-03, grad_scale: 8.0 +2023-04-02 00:32:00,560 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 00:32:04,161 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85892.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:32:05,564 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85893.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:32:20,763 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85905.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:32:59,797 INFO [train.py:903] (3/4) Epoch 13, batch 4000, loss[loss=0.2105, simple_loss=0.2836, pruned_loss=0.06868, over 19472.00 frames. ], tot_loss[loss=0.2287, simple_loss=0.3032, pruned_loss=0.07711, over 3827667.74 frames. ], batch size: 49, lr: 6.38e-03, grad_scale: 8.0 +2023-04-02 00:33:14,767 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85948.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 00:33:30,762 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85960.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:33:45,317 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9167, 1.3532, 1.5834, 1.5541, 3.4359, 1.0831, 2.4112, 3.8359], + device='cuda:3'), covar=tensor([0.0383, 0.2469, 0.2404, 0.1750, 0.0667, 0.2350, 0.1182, 0.0231], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0341, 0.0352, 0.0321, 0.0344, 0.0329, 0.0341, 0.0362], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:33:46,159 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 00:33:48,582 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85975.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:33:56,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.510e+02 5.280e+02 6.258e+02 8.022e+02 1.377e+03, threshold=1.252e+03, percent-clipped=2.0 +2023-04-02 00:34:02,169 INFO [train.py:903] (3/4) Epoch 13, batch 4050, loss[loss=0.2639, simple_loss=0.3392, pruned_loss=0.09424, over 18777.00 frames. ], tot_loss[loss=0.2293, simple_loss=0.3038, pruned_loss=0.07743, over 3829095.04 frames. ], batch size: 74, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:34:03,499 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=85987.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:34:30,862 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86007.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:34:46,101 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86020.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:35:02,921 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9669, 2.0668, 2.1557, 2.6961, 1.9416, 2.5865, 2.2733, 1.9314], + device='cuda:3'), covar=tensor([0.3696, 0.3163, 0.1585, 0.2102, 0.3609, 0.1662, 0.4226, 0.2973], + device='cuda:3'), in_proj_covar=tensor([0.0813, 0.0845, 0.0660, 0.0898, 0.0797, 0.0724, 0.0798, 0.0723], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:35:05,744 INFO [train.py:903] (3/4) Epoch 13, batch 4100, loss[loss=0.2244, simple_loss=0.2903, pruned_loss=0.0792, over 19414.00 frames. ], tot_loss[loss=0.2289, simple_loss=0.3033, pruned_loss=0.07729, over 3819107.58 frames. ], batch size: 48, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:35:33,084 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-02 00:35:41,505 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 00:35:48,406 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86070.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:36:04,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.212e+02 4.943e+02 6.094e+02 7.890e+02 1.986e+03, threshold=1.219e+03, percent-clipped=5.0 +2023-04-02 00:36:08,782 INFO [train.py:903] (3/4) Epoch 13, batch 4150, loss[loss=0.2, simple_loss=0.266, pruned_loss=0.06703, over 19302.00 frames. ], tot_loss[loss=0.2284, simple_loss=0.3034, pruned_loss=0.07669, over 3818790.37 frames. ], batch size: 44, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:36:11,025 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86087.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:36:28,541 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86102.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:36:36,343 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7581, 1.4450, 1.5172, 2.1352, 1.7386, 2.0425, 1.9839, 1.7637], + device='cuda:3'), covar=tensor([0.0741, 0.0949, 0.0953, 0.0711, 0.0777, 0.0689, 0.0826, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0223, 0.0223, 0.0242, 0.0231, 0.0209, 0.0191, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 00:37:11,323 INFO [train.py:903] (3/4) Epoch 13, batch 4200, loss[loss=0.1903, simple_loss=0.2712, pruned_loss=0.05464, over 19700.00 frames. ], tot_loss[loss=0.2276, simple_loss=0.3026, pruned_loss=0.07627, over 3813239.54 frames. ], batch size: 53, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:37:14,929 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 00:37:42,913 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86161.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:38:03,679 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5952, 1.3484, 1.4953, 1.6060, 3.1189, 1.1365, 2.2305, 3.5101], + device='cuda:3'), covar=tensor([0.0413, 0.2530, 0.2556, 0.1619, 0.0675, 0.2278, 0.1315, 0.0260], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0342, 0.0354, 0.0321, 0.0345, 0.0329, 0.0342, 0.0362], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:38:08,163 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.547e+02 5.506e+02 6.651e+02 9.015e+02 1.898e+03, threshold=1.330e+03, percent-clipped=8.0 +2023-04-02 00:38:12,742 INFO [train.py:903] (3/4) Epoch 13, batch 4250, loss[loss=0.2696, simple_loss=0.3311, pruned_loss=0.104, over 19782.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3014, pruned_loss=0.0754, over 3820701.41 frames. ], batch size: 56, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:38:13,197 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86186.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:38:31,046 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 00:38:31,968 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-02 00:38:43,576 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 00:38:47,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1393, 5.5073, 2.6198, 4.7780, 1.3085, 5.5005, 5.4151, 5.6065], + device='cuda:3'), covar=tensor([0.0429, 0.0837, 0.2155, 0.0658, 0.3510, 0.0536, 0.0737, 0.1004], + device='cuda:3'), in_proj_covar=tensor([0.0444, 0.0370, 0.0444, 0.0320, 0.0386, 0.0378, 0.0367, 0.0404], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:38:52,293 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86216.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:39:15,640 INFO [train.py:903] (3/4) Epoch 13, batch 4300, loss[loss=0.2353, simple_loss=0.3136, pruned_loss=0.07852, over 19654.00 frames. ], tot_loss[loss=0.2275, simple_loss=0.3029, pruned_loss=0.07604, over 3804947.57 frames. ], batch size: 55, lr: 6.37e-03, grad_scale: 8.0 +2023-04-02 00:39:17,009 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86237.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:39:22,953 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86241.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:39:50,489 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86263.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:39:51,564 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86264.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:40:13,251 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.647e+02 5.244e+02 6.314e+02 8.271e+02 2.210e+03, threshold=1.263e+03, percent-clipped=7.0 +2023-04-02 00:40:15,231 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 00:40:19,670 INFO [train.py:903] (3/4) Epoch 13, batch 4350, loss[loss=0.2571, simple_loss=0.3338, pruned_loss=0.09017, over 19835.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3024, pruned_loss=0.07584, over 3808377.07 frames. ], batch size: 52, lr: 6.36e-03, grad_scale: 8.0 +2023-04-02 00:40:23,414 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86288.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:40:27,970 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86292.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 00:41:00,081 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86319.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:41:22,566 INFO [train.py:903] (3/4) Epoch 13, batch 4400, loss[loss=0.2166, simple_loss=0.2845, pruned_loss=0.07438, over 16024.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.301, pruned_loss=0.07513, over 3802549.82 frames. ], batch size: 35, lr: 6.36e-03, grad_scale: 8.0 +2023-04-02 00:41:42,037 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86352.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:41:49,360 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86358.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:41:50,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 00:41:58,614 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86364.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:42:00,825 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 00:42:20,371 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.626e+02 5.303e+02 6.614e+02 7.903e+02 1.976e+03, threshold=1.323e+03, percent-clipped=7.0 +2023-04-02 00:42:21,992 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86383.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:42:25,156 INFO [train.py:903] (3/4) Epoch 13, batch 4450, loss[loss=0.2151, simple_loss=0.2796, pruned_loss=0.07535, over 19721.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.3008, pruned_loss=0.0751, over 3817402.17 frames. ], batch size: 46, lr: 6.36e-03, grad_scale: 8.0 +2023-04-02 00:42:49,909 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9913, 1.6267, 1.5706, 1.8850, 1.6437, 1.6735, 1.4416, 1.9139], + device='cuda:3'), covar=tensor([0.0894, 0.1379, 0.1344, 0.0914, 0.1200, 0.0523, 0.1289, 0.0645], + device='cuda:3'), in_proj_covar=tensor([0.0253, 0.0343, 0.0290, 0.0238, 0.0290, 0.0237, 0.0279, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:42:52,180 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86407.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 00:43:00,999 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86414.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:43:21,646 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86431.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:43:25,404 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86434.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:43:27,355 INFO [train.py:903] (3/4) Epoch 13, batch 4500, loss[loss=0.2275, simple_loss=0.3118, pruned_loss=0.07157, over 19592.00 frames. ], tot_loss[loss=0.2265, simple_loss=0.3018, pruned_loss=0.07558, over 3822479.98 frames. ], batch size: 57, lr: 6.36e-03, grad_scale: 8.0 +2023-04-02 00:43:48,840 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2111, 1.1874, 1.1838, 1.3362, 1.0720, 1.3339, 1.4104, 1.2743], + device='cuda:3'), covar=tensor([0.0870, 0.0983, 0.1096, 0.0701, 0.0860, 0.0822, 0.0785, 0.0755], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0225, 0.0223, 0.0243, 0.0231, 0.0210, 0.0192, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 00:43:55,649 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86458.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:44:01,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86463.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:44:21,266 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:44:24,464 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.347e+02 5.430e+02 6.459e+02 7.870e+02 1.871e+03, threshold=1.292e+03, percent-clipped=3.0 +2023-04-02 00:44:29,962 INFO [train.py:903] (3/4) Epoch 13, batch 4550, loss[loss=0.2616, simple_loss=0.3381, pruned_loss=0.09259, over 19611.00 frames. ], tot_loss[loss=0.2279, simple_loss=0.3028, pruned_loss=0.07656, over 3805408.80 frames. ], batch size: 57, lr: 6.36e-03, grad_scale: 8.0 +2023-04-02 00:44:39,221 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 00:45:02,119 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 00:45:25,853 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86529.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:45:33,583 INFO [train.py:903] (3/4) Epoch 13, batch 4600, loss[loss=0.2565, simple_loss=0.3293, pruned_loss=0.09184, over 19687.00 frames. ], tot_loss[loss=0.2277, simple_loss=0.3028, pruned_loss=0.07625, over 3812324.91 frames. ], batch size: 60, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:45:46,439 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.3526, 5.3462, 6.2226, 6.1100, 2.2375, 5.8116, 5.0176, 5.7980], + device='cuda:3'), covar=tensor([0.1436, 0.0632, 0.0451, 0.0501, 0.5292, 0.0524, 0.0513, 0.0894], + device='cuda:3'), in_proj_covar=tensor([0.0707, 0.0625, 0.0835, 0.0706, 0.0751, 0.0576, 0.0503, 0.0766], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 00:45:46,527 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86546.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:46:31,573 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.083e+02 5.452e+02 6.634e+02 8.020e+02 1.693e+03, threshold=1.327e+03, percent-clipped=1.0 +2023-04-02 00:46:35,936 INFO [train.py:903] (3/4) Epoch 13, batch 4650, loss[loss=0.2004, simple_loss=0.2781, pruned_loss=0.06135, over 19484.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3024, pruned_loss=0.07584, over 3805166.58 frames. ], batch size: 49, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:46:51,371 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86598.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:46:52,257 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 00:47:03,875 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 00:47:03,998 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86608.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:47:04,256 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86608.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:47:35,039 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86633.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:47:35,324 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 00:47:38,175 INFO [train.py:903] (3/4) Epoch 13, batch 4700, loss[loss=0.2247, simple_loss=0.3061, pruned_loss=0.07168, over 19315.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3012, pruned_loss=0.07511, over 3818710.21 frames. ], batch size: 66, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:47:53,756 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 00:48:04,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 00:48:06,042 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2266, 1.3273, 1.2626, 1.0704, 1.0422, 1.0901, 0.0420, 0.3429], + device='cuda:3'), covar=tensor([0.0511, 0.0487, 0.0336, 0.0431, 0.1067, 0.0447, 0.0968, 0.0859], + device='cuda:3'), in_proj_covar=tensor([0.0341, 0.0335, 0.0332, 0.0357, 0.0432, 0.0357, 0.0314, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 00:48:13,320 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86663.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 00:48:35,953 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.257e+02 5.083e+02 5.953e+02 7.054e+02 1.649e+03, threshold=1.191e+03, percent-clipped=1.0 +2023-04-02 00:48:42,059 INFO [train.py:903] (3/4) Epoch 13, batch 4750, loss[loss=0.1982, simple_loss=0.2814, pruned_loss=0.05746, over 19599.00 frames. ], tot_loss[loss=0.2251, simple_loss=0.3007, pruned_loss=0.07476, over 3819447.14 frames. ], batch size: 52, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:48:45,718 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86688.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 00:48:48,062 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86690.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:48:58,152 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4759, 1.4143, 1.5094, 1.5262, 3.0573, 1.2446, 2.3220, 3.4090], + device='cuda:3'), covar=tensor([0.0465, 0.2578, 0.2653, 0.1858, 0.0679, 0.2364, 0.1310, 0.0300], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0345, 0.0356, 0.0325, 0.0352, 0.0332, 0.0346, 0.0365], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:49:07,677 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-02 00:49:10,076 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-04-02 00:49:15,251 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.10 vs. limit=2.0 +2023-04-02 00:49:18,565 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86715.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:49:27,264 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86723.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:49:31,061 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-04-02 00:49:43,063 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86735.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:49:43,762 INFO [train.py:903] (3/4) Epoch 13, batch 4800, loss[loss=0.243, simple_loss=0.311, pruned_loss=0.08749, over 19681.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3016, pruned_loss=0.07534, over 3826201.52 frames. ], batch size: 53, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:50:12,867 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5040, 2.1801, 2.1860, 2.5800, 2.6048, 2.1178, 2.1753, 2.4495], + device='cuda:3'), covar=tensor([0.0878, 0.1712, 0.1301, 0.0937, 0.1138, 0.0503, 0.1134, 0.0643], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0348, 0.0293, 0.0241, 0.0294, 0.0240, 0.0282, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:50:12,875 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86760.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:50:41,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.218e+02 5.270e+02 6.552e+02 8.488e+02 1.715e+03, threshold=1.310e+03, percent-clipped=7.0 +2023-04-02 00:50:45,052 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86785.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:50:45,824 INFO [train.py:903] (3/4) Epoch 13, batch 4850, loss[loss=0.2085, simple_loss=0.2895, pruned_loss=0.06369, over 19617.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.3014, pruned_loss=0.07525, over 3821754.57 frames. ], batch size: 57, lr: 6.35e-03, grad_scale: 8.0 +2023-04-02 00:51:06,435 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86802.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:06,677 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86802.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:08,605 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 00:51:14,209 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86807.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:18,126 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86810.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:28,824 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 00:51:31,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 00:51:37,279 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 00:51:37,323 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 00:51:38,821 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86827.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:45,318 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86832.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:51:47,410 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 00:51:49,795 INFO [train.py:903] (3/4) Epoch 13, batch 4900, loss[loss=0.1935, simple_loss=0.2704, pruned_loss=0.05831, over 19354.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3018, pruned_loss=0.07542, over 3816010.23 frames. ], batch size: 48, lr: 6.34e-03, grad_scale: 8.0 +2023-04-02 00:51:59,426 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1259, 1.8091, 1.7067, 2.6174, 1.8475, 2.2675, 2.5440, 2.2243], + device='cuda:3'), covar=tensor([0.0773, 0.0961, 0.1117, 0.0949, 0.0989, 0.0776, 0.0913, 0.0642], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0224, 0.0224, 0.0244, 0.0230, 0.0211, 0.0192, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 00:52:07,082 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 00:52:46,044 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.083e+02 5.393e+02 6.845e+02 8.725e+02 1.892e+03, threshold=1.369e+03, percent-clipped=5.0 +2023-04-02 00:52:50,738 INFO [train.py:903] (3/4) Epoch 13, batch 4950, loss[loss=0.2741, simple_loss=0.3471, pruned_loss=0.1005, over 19356.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3023, pruned_loss=0.07587, over 3804914.09 frames. ], batch size: 70, lr: 6.34e-03, grad_scale: 8.0 +2023-04-02 00:53:05,783 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86896.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:53:08,070 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 00:53:31,827 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86917.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:53:32,649 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 00:53:37,700 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86922.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:53:56,223 INFO [train.py:903] (3/4) Epoch 13, batch 5000, loss[loss=0.2464, simple_loss=0.3178, pruned_loss=0.08746, over 19607.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3016, pruned_loss=0.07529, over 3806478.81 frames. ], batch size: 57, lr: 6.34e-03, grad_scale: 8.0 +2023-04-02 00:54:03,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 00:54:03,558 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=86942.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:54:14,008 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 00:54:49,343 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:54:52,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.998e+02 5.650e+02 6.679e+02 8.615e+02 1.943e+03, threshold=1.336e+03, percent-clipped=4.0 +2023-04-02 00:54:56,919 INFO [train.py:903] (3/4) Epoch 13, batch 5050, loss[loss=0.2357, simple_loss=0.3235, pruned_loss=0.07399, over 18638.00 frames. ], tot_loss[loss=0.2265, simple_loss=0.3021, pruned_loss=0.07544, over 3809929.26 frames. ], batch size: 74, lr: 6.34e-03, grad_scale: 8.0 +2023-04-02 00:55:19,921 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87004.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:55:33,341 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 00:55:38,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5296, 1.2407, 1.1707, 1.4540, 1.1682, 1.3355, 1.2177, 1.3777], + device='cuda:3'), covar=tensor([0.1051, 0.1205, 0.1471, 0.0925, 0.1137, 0.0575, 0.1313, 0.0775], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0349, 0.0292, 0.0239, 0.0295, 0.0241, 0.0281, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 00:55:41,790 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87020.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 00:56:01,009 INFO [train.py:903] (3/4) Epoch 13, batch 5100, loss[loss=0.2496, simple_loss=0.3175, pruned_loss=0.09087, over 18792.00 frames. ], tot_loss[loss=0.2265, simple_loss=0.302, pruned_loss=0.07543, over 3812157.64 frames. ], batch size: 74, lr: 6.34e-03, grad_scale: 8.0 +2023-04-02 00:56:07,909 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 00:56:11,370 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 00:56:18,087 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 00:56:28,907 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87057.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:56:58,519 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.198e+02 5.089e+02 6.519e+02 8.713e+02 1.677e+03, threshold=1.304e+03, percent-clipped=2.0 +2023-04-02 00:57:03,281 INFO [train.py:903] (3/4) Epoch 13, batch 5150, loss[loss=0.2191, simple_loss=0.306, pruned_loss=0.06608, over 19658.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3024, pruned_loss=0.07586, over 3818110.07 frames. ], batch size: 60, lr: 6.33e-03, grad_scale: 8.0 +2023-04-02 00:57:15,160 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 00:57:46,783 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9572, 2.0065, 2.3088, 2.6211, 1.8584, 2.4657, 2.5101, 2.1472], + device='cuda:3'), covar=tensor([0.3813, 0.3397, 0.1505, 0.1979, 0.3725, 0.1787, 0.3620, 0.2767], + device='cuda:3'), in_proj_covar=tensor([0.0815, 0.0844, 0.0661, 0.0894, 0.0795, 0.0728, 0.0796, 0.0728], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:57:48,534 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 00:58:07,739 INFO [train.py:903] (3/4) Epoch 13, batch 5200, loss[loss=0.238, simple_loss=0.3193, pruned_loss=0.07837, over 17568.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3022, pruned_loss=0.07628, over 3821783.57 frames. ], batch size: 101, lr: 6.33e-03, grad_scale: 8.0 +2023-04-02 00:58:18,406 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 00:58:36,891 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5772, 4.0081, 4.2137, 4.2036, 1.5809, 3.9363, 3.4481, 3.8912], + device='cuda:3'), covar=tensor([0.1448, 0.0869, 0.0571, 0.0618, 0.5490, 0.0831, 0.0667, 0.1054], + device='cuda:3'), in_proj_covar=tensor([0.0705, 0.0622, 0.0829, 0.0704, 0.0748, 0.0578, 0.0499, 0.0768], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 00:58:54,672 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87173.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:58:58,886 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87176.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:59:01,499 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87178.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:59:04,669 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 00:59:05,686 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.198e+02 5.751e+02 6.794e+02 9.141e+02 2.147e+03, threshold=1.359e+03, percent-clipped=10.0 +2023-04-02 00:59:10,443 INFO [train.py:903] (3/4) Epoch 13, batch 5250, loss[loss=0.2321, simple_loss=0.3199, pruned_loss=0.0722, over 19670.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3014, pruned_loss=0.07557, over 3831219.04 frames. ], batch size: 60, lr: 6.33e-03, grad_scale: 16.0 +2023-04-02 00:59:25,776 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87198.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:59:30,538 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8295, 1.9484, 2.1275, 2.5920, 1.8174, 2.4450, 2.3444, 1.9606], + device='cuda:3'), covar=tensor([0.3912, 0.3050, 0.1606, 0.1979, 0.3382, 0.1678, 0.3744, 0.2902], + device='cuda:3'), in_proj_covar=tensor([0.0814, 0.0843, 0.0661, 0.0896, 0.0797, 0.0728, 0.0796, 0.0726], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 00:59:31,629 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87203.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 00:59:39,976 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-02 01:00:08,204 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0062, 4.4511, 4.7493, 4.7530, 1.7882, 4.4626, 3.8818, 4.4016], + device='cuda:3'), covar=tensor([0.1396, 0.0757, 0.0553, 0.0558, 0.5329, 0.0681, 0.0649, 0.1097], + device='cuda:3'), in_proj_covar=tensor([0.0718, 0.0634, 0.0845, 0.0720, 0.0763, 0.0590, 0.0509, 0.0781], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 01:00:13,791 INFO [train.py:903] (3/4) Epoch 13, batch 5300, loss[loss=0.1865, simple_loss=0.2634, pruned_loss=0.05485, over 19414.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3022, pruned_loss=0.07584, over 3837121.87 frames. ], batch size: 48, lr: 6.33e-03, grad_scale: 16.0 +2023-04-02 01:00:18,367 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87240.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:00:27,871 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87248.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:00:28,732 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 01:01:10,975 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.140e+02 5.034e+02 6.226e+02 7.840e+02 1.929e+03, threshold=1.245e+03, percent-clipped=3.0 +2023-04-02 01:01:15,806 INFO [train.py:903] (3/4) Epoch 13, batch 5350, loss[loss=0.2187, simple_loss=0.2993, pruned_loss=0.06902, over 19438.00 frames. ], tot_loss[loss=0.2269, simple_loss=0.3021, pruned_loss=0.07589, over 3835853.95 frames. ], batch size: 70, lr: 6.33e-03, grad_scale: 16.0 +2023-04-02 01:01:24,120 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87291.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:01:49,519 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 01:01:51,106 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87313.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:02:08,494 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8769, 1.2896, 1.6131, 0.5165, 1.9433, 2.4331, 2.0665, 2.5670], + device='cuda:3'), covar=tensor([0.1536, 0.3347, 0.2972, 0.2436, 0.0555, 0.0238, 0.0343, 0.0305], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0303, 0.0333, 0.0253, 0.0224, 0.0164, 0.0209, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:02:21,294 INFO [train.py:903] (3/4) Epoch 13, batch 5400, loss[loss=0.2568, simple_loss=0.3225, pruned_loss=0.09552, over 19653.00 frames. ], tot_loss[loss=0.2275, simple_loss=0.3027, pruned_loss=0.07612, over 3840498.87 frames. ], batch size: 60, lr: 6.33e-03, grad_scale: 8.0 +2023-04-02 01:02:24,101 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87338.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:02:44,281 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87355.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:02:52,487 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-02 01:02:55,501 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87364.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:03:01,413 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9252, 5.3238, 2.8319, 4.5653, 1.0033, 5.3232, 5.3439, 5.4422], + device='cuda:3'), covar=tensor([0.0441, 0.0841, 0.1943, 0.0672, 0.4156, 0.0563, 0.0625, 0.0914], + device='cuda:3'), in_proj_covar=tensor([0.0447, 0.0372, 0.0450, 0.0322, 0.0384, 0.0383, 0.0372, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:03:16,676 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87380.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:03:19,989 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.794e+02 4.952e+02 6.479e+02 7.900e+02 1.578e+03, threshold=1.296e+03, percent-clipped=3.0 +2023-04-02 01:03:23,446 INFO [train.py:903] (3/4) Epoch 13, batch 5450, loss[loss=0.2395, simple_loss=0.3166, pruned_loss=0.08117, over 19431.00 frames. ], tot_loss[loss=0.2273, simple_loss=0.3026, pruned_loss=0.07598, over 3839502.82 frames. ], batch size: 64, lr: 6.32e-03, grad_scale: 8.0 +2023-04-02 01:04:25,744 INFO [train.py:903] (3/4) Epoch 13, batch 5500, loss[loss=0.21, simple_loss=0.2974, pruned_loss=0.06132, over 19587.00 frames. ], tot_loss[loss=0.2283, simple_loss=0.3035, pruned_loss=0.07656, over 3832884.97 frames. ], batch size: 61, lr: 6.32e-03, grad_scale: 8.0 +2023-04-02 01:04:47,842 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 01:05:21,386 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:05:21,453 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87479.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:05:25,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.829e+02 5.041e+02 6.443e+02 8.590e+02 2.643e+03, threshold=1.289e+03, percent-clipped=7.0 +2023-04-02 01:05:29,325 INFO [train.py:903] (3/4) Epoch 13, batch 5550, loss[loss=0.2193, simple_loss=0.2906, pruned_loss=0.07398, over 19739.00 frames. ], tot_loss[loss=0.2285, simple_loss=0.3038, pruned_loss=0.07659, over 3821056.31 frames. ], batch size: 51, lr: 6.32e-03, grad_scale: 8.0 +2023-04-02 01:05:31,797 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0076, 4.3929, 4.7162, 4.7094, 1.6430, 4.3857, 3.7785, 4.3402], + device='cuda:3'), covar=tensor([0.1484, 0.0832, 0.0589, 0.0629, 0.5773, 0.0765, 0.0678, 0.1195], + device='cuda:3'), in_proj_covar=tensor([0.0709, 0.0630, 0.0837, 0.0716, 0.0752, 0.0582, 0.0504, 0.0773], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 01:05:33,837 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 01:05:58,148 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4915, 1.2034, 1.5076, 1.7464, 3.7775, 1.1008, 2.4325, 4.0874], + device='cuda:3'), covar=tensor([0.0624, 0.3738, 0.3619, 0.2286, 0.1210, 0.3282, 0.1661, 0.0405], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0345, 0.0356, 0.0324, 0.0352, 0.0333, 0.0343, 0.0366], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:06:12,095 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0159, 1.6847, 1.9473, 1.8343, 4.5563, 0.9990, 2.5483, 4.9065], + device='cuda:3'), covar=tensor([0.0347, 0.2469, 0.2380, 0.1723, 0.0673, 0.2592, 0.1266, 0.0169], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0345, 0.0355, 0.0324, 0.0352, 0.0332, 0.0343, 0.0365], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:06:16,486 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87523.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:06:23,056 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 01:06:33,784 INFO [train.py:903] (3/4) Epoch 13, batch 5600, loss[loss=0.2208, simple_loss=0.3037, pruned_loss=0.06891, over 19505.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3029, pruned_loss=0.07598, over 3815594.09 frames. ], batch size: 64, lr: 6.32e-03, grad_scale: 8.0 +2023-04-02 01:06:47,659 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87547.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:06:50,180 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-04-02 01:07:17,493 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87572.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:07:32,968 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.526e+02 5.023e+02 6.222e+02 8.051e+02 1.328e+03, threshold=1.244e+03, percent-clipped=2.0 +2023-04-02 01:07:36,347 INFO [train.py:903] (3/4) Epoch 13, batch 5650, loss[loss=0.2003, simple_loss=0.2865, pruned_loss=0.05707, over 19596.00 frames. ], tot_loss[loss=0.226, simple_loss=0.3015, pruned_loss=0.07525, over 3826803.89 frames. ], batch size: 61, lr: 6.32e-03, grad_scale: 8.0 +2023-04-02 01:07:38,457 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-02 01:07:39,087 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2638, 1.9836, 1.5040, 1.3132, 1.8181, 1.1713, 1.2413, 1.7418], + device='cuda:3'), covar=tensor([0.0817, 0.0669, 0.0999, 0.0695, 0.0472, 0.1180, 0.0609, 0.0387], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0305, 0.0330, 0.0251, 0.0241, 0.0324, 0.0292, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:07:43,448 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87592.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:07:56,302 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87602.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:08:01,807 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87607.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:08:06,355 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87611.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:08:23,603 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87623.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:08:24,453 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 01:08:38,519 INFO [train.py:903] (3/4) Epoch 13, batch 5700, loss[loss=0.2476, simple_loss=0.321, pruned_loss=0.08709, over 19518.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.2996, pruned_loss=0.07415, over 3828995.24 frames. ], batch size: 54, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:08:38,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87636.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:08:53,290 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-02 01:09:36,750 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.413e+02 4.997e+02 6.690e+02 8.841e+02 1.921e+03, threshold=1.338e+03, percent-clipped=6.0 +2023-04-02 01:09:40,128 INFO [train.py:903] (3/4) Epoch 13, batch 5750, loss[loss=0.1918, simple_loss=0.2726, pruned_loss=0.05549, over 19676.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.301, pruned_loss=0.07501, over 3829284.07 frames. ], batch size: 53, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:09:41,281 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 01:09:49,518 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 01:09:55,222 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 01:10:07,764 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87707.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:10:27,882 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87724.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:10:37,353 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2481, 2.0502, 1.6017, 1.2668, 1.8981, 1.1120, 1.2244, 1.7987], + device='cuda:3'), covar=tensor([0.0817, 0.0691, 0.1005, 0.0785, 0.0434, 0.1309, 0.0634, 0.0373], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0307, 0.0332, 0.0252, 0.0241, 0.0326, 0.0294, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:10:41,880 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87735.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:10:42,502 INFO [train.py:903] (3/4) Epoch 13, batch 5800, loss[loss=0.2805, simple_loss=0.3378, pruned_loss=0.1115, over 19622.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.3005, pruned_loss=0.07521, over 3822444.54 frames. ], batch size: 50, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:10:46,879 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-02 01:11:07,373 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.67 vs. limit=5.0 +2023-04-02 01:11:13,897 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87760.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:11:27,838 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0396, 4.3885, 4.7439, 4.7489, 1.6650, 4.4559, 3.8487, 4.4292], + device='cuda:3'), covar=tensor([0.1463, 0.0771, 0.0548, 0.0543, 0.5184, 0.0640, 0.0598, 0.1049], + device='cuda:3'), in_proj_covar=tensor([0.0703, 0.0625, 0.0836, 0.0714, 0.0745, 0.0581, 0.0502, 0.0765], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 01:11:43,246 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.294e+02 5.041e+02 6.242e+02 7.928e+02 1.581e+03, threshold=1.248e+03, percent-clipped=4.0 +2023-04-02 01:11:46,964 INFO [train.py:903] (3/4) Epoch 13, batch 5850, loss[loss=0.2088, simple_loss=0.2793, pruned_loss=0.0691, over 19782.00 frames. ], tot_loss[loss=0.2264, simple_loss=0.3012, pruned_loss=0.07576, over 3807523.51 frames. ], batch size: 47, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:12:18,374 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2356, 2.0219, 1.8351, 2.5191, 1.8289, 2.6411, 2.5584, 2.2362], + device='cuda:3'), covar=tensor([0.0774, 0.0814, 0.1011, 0.0928, 0.0909, 0.0680, 0.0812, 0.0620], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0223, 0.0224, 0.0242, 0.0229, 0.0211, 0.0191, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 01:12:33,216 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87823.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:12:47,982 INFO [train.py:903] (3/4) Epoch 13, batch 5900, loss[loss=0.235, simple_loss=0.3042, pruned_loss=0.0829, over 19679.00 frames. ], tot_loss[loss=0.2249, simple_loss=0.3, pruned_loss=0.07493, over 3816859.12 frames. ], batch size: 53, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:12:49,039 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 01:12:51,599 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87839.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:13:03,845 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9166, 1.8277, 1.6962, 2.0546, 1.8314, 1.6848, 1.7090, 1.9086], + device='cuda:3'), covar=tensor([0.0843, 0.1270, 0.1257, 0.0814, 0.1061, 0.0555, 0.1108, 0.0654], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0353, 0.0296, 0.0239, 0.0295, 0.0243, 0.0284, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:13:10,521 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 01:13:27,552 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87867.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:13:47,009 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.002e+02 5.447e+02 6.662e+02 8.353e+02 2.279e+03, threshold=1.332e+03, percent-clipped=6.0 +2023-04-02 01:13:50,411 INFO [train.py:903] (3/4) Epoch 13, batch 5950, loss[loss=0.2507, simple_loss=0.3205, pruned_loss=0.09051, over 19585.00 frames. ], tot_loss[loss=0.2251, simple_loss=0.3001, pruned_loss=0.07499, over 3811465.18 frames. ], batch size: 52, lr: 6.31e-03, grad_scale: 8.0 +2023-04-02 01:14:30,803 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6903, 1.7699, 1.4011, 1.7904, 1.6791, 1.3953, 1.5892, 1.7091], + device='cuda:3'), covar=tensor([0.1147, 0.1585, 0.1809, 0.1186, 0.1471, 0.0965, 0.1565, 0.0983], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0352, 0.0295, 0.0239, 0.0295, 0.0242, 0.0284, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:14:53,069 INFO [train.py:903] (3/4) Epoch 13, batch 6000, loss[loss=0.2804, simple_loss=0.3449, pruned_loss=0.108, over 13357.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3007, pruned_loss=0.07533, over 3803280.67 frames. ], batch size: 135, lr: 6.30e-03, grad_scale: 8.0 +2023-04-02 01:14:53,069 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 01:15:06,406 INFO [train.py:937] (3/4) Epoch 13, validation: loss=0.175, simple_loss=0.2755, pruned_loss=0.03726, over 944034.00 frames. +2023-04-02 01:15:06,407 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 01:15:09,205 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87938.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:15:19,419 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87946.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:15:26,087 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87951.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:15:41,054 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87963.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:15:45,514 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=87967.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:16:04,017 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87982.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:16:04,765 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.636e+02 5.142e+02 6.649e+02 8.424e+02 2.027e+03, threshold=1.330e+03, percent-clipped=9.0 +2023-04-02 01:16:08,329 INFO [train.py:903] (3/4) Epoch 13, batch 6050, loss[loss=0.1928, simple_loss=0.2691, pruned_loss=0.05822, over 19468.00 frames. ], tot_loss[loss=0.2269, simple_loss=0.3017, pruned_loss=0.07606, over 3804371.45 frames. ], batch size: 49, lr: 6.30e-03, grad_scale: 8.0 +2023-04-02 01:16:12,361 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87988.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:17:12,303 INFO [train.py:903] (3/4) Epoch 13, batch 6100, loss[loss=0.2481, simple_loss=0.3311, pruned_loss=0.08257, over 19340.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3018, pruned_loss=0.07623, over 3808798.55 frames. ], batch size: 66, lr: 6.30e-03, grad_scale: 8.0 +2023-04-02 01:17:42,751 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88061.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:17:49,462 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88066.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:18:10,011 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88082.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:18:10,751 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.507e+02 5.311e+02 6.775e+02 8.712e+02 1.953e+03, threshold=1.355e+03, percent-clipped=3.0 +2023-04-02 01:18:14,183 INFO [train.py:903] (3/4) Epoch 13, batch 6150, loss[loss=0.26, simple_loss=0.3318, pruned_loss=0.09408, over 19569.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3019, pruned_loss=0.07642, over 3798487.97 frames. ], batch size: 61, lr: 6.30e-03, grad_scale: 8.0 +2023-04-02 01:18:21,018 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.68 vs. limit=2.0 +2023-04-02 01:18:24,932 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88095.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:18:38,392 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 01:18:54,547 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=88118.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:18:56,839 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88120.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:19:15,199 INFO [train.py:903] (3/4) Epoch 13, batch 6200, loss[loss=0.2414, simple_loss=0.3156, pruned_loss=0.08366, over 19660.00 frames. ], tot_loss[loss=0.2281, simple_loss=0.3025, pruned_loss=0.07687, over 3794215.50 frames. ], batch size: 58, lr: 6.30e-03, grad_scale: 8.0 +2023-04-02 01:19:45,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3415, 1.3937, 1.4573, 1.4709, 1.7400, 1.8665, 1.7074, 0.5517], + device='cuda:3'), covar=tensor([0.2108, 0.3784, 0.2324, 0.1768, 0.1519, 0.2036, 0.1375, 0.3981], + device='cuda:3'), in_proj_covar=tensor([0.0501, 0.0591, 0.0630, 0.0447, 0.0604, 0.0498, 0.0649, 0.0507], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 01:20:13,333 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.409e+02 5.304e+02 6.722e+02 8.627e+02 2.252e+03, threshold=1.344e+03, percent-clipped=3.0 +2023-04-02 01:20:16,891 INFO [train.py:903] (3/4) Epoch 13, batch 6250, loss[loss=0.2225, simple_loss=0.3054, pruned_loss=0.06981, over 19674.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3018, pruned_loss=0.0761, over 3811257.18 frames. ], batch size: 58, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:20:29,072 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88194.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:20:44,711 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 01:20:53,433 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4070, 1.3392, 1.4089, 1.4784, 2.9748, 1.0187, 2.3068, 3.3367], + device='cuda:3'), covar=tensor([0.0476, 0.2602, 0.2873, 0.1809, 0.0730, 0.2644, 0.1234, 0.0280], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0344, 0.0356, 0.0324, 0.0350, 0.0332, 0.0341, 0.0365], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:20:58,211 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88219.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:21:03,624 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4465, 1.4954, 1.7266, 1.6271, 2.6738, 2.2395, 2.7009, 1.2227], + device='cuda:3'), covar=tensor([0.2258, 0.4067, 0.2642, 0.1816, 0.1394, 0.1989, 0.1471, 0.3827], + device='cuda:3'), in_proj_covar=tensor([0.0503, 0.0595, 0.0633, 0.0450, 0.0606, 0.0501, 0.0652, 0.0508], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 01:21:20,739 INFO [train.py:903] (3/4) Epoch 13, batch 6300, loss[loss=0.2157, simple_loss=0.2838, pruned_loss=0.07376, over 19779.00 frames. ], tot_loss[loss=0.2268, simple_loss=0.3013, pruned_loss=0.07619, over 3814508.10 frames. ], batch size: 47, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:21:23,582 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88238.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:21:52,953 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88263.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:22:18,335 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.601e+02 5.052e+02 6.199e+02 7.712e+02 1.989e+03, threshold=1.240e+03, percent-clipped=5.0 +2023-04-02 01:22:21,899 INFO [train.py:903] (3/4) Epoch 13, batch 6350, loss[loss=0.2091, simple_loss=0.292, pruned_loss=0.06313, over 19615.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3006, pruned_loss=0.07582, over 3811222.21 frames. ], batch size: 57, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:22:50,135 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=88309.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 01:23:02,557 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88317.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:08,251 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88322.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:24,611 INFO [train.py:903] (3/4) Epoch 13, batch 6400, loss[loss=0.207, simple_loss=0.2983, pruned_loss=0.05783, over 19537.00 frames. ], tot_loss[loss=0.2253, simple_loss=0.3, pruned_loss=0.07528, over 3810030.45 frames. ], batch size: 56, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:23:27,210 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88338.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:32,584 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88342.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:38,245 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:58,508 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88363.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:23:59,590 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9006, 1.1802, 1.5935, 0.6160, 2.0872, 2.4790, 2.1589, 2.5606], + device='cuda:3'), covar=tensor([0.1509, 0.3529, 0.2920, 0.2373, 0.0527, 0.0239, 0.0334, 0.0314], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0302, 0.0331, 0.0252, 0.0220, 0.0165, 0.0207, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:24:22,192 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.590e+02 5.639e+02 6.967e+02 8.699e+02 1.659e+03, threshold=1.393e+03, percent-clipped=7.0 +2023-04-02 01:24:25,788 INFO [train.py:903] (3/4) Epoch 13, batch 6450, loss[loss=0.2121, simple_loss=0.2997, pruned_loss=0.06229, over 18713.00 frames. ], tot_loss[loss=0.2251, simple_loss=0.3001, pruned_loss=0.07507, over 3814728.75 frames. ], batch size: 74, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:24:57,403 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9035, 1.2596, 1.5294, 0.5882, 2.0448, 2.4274, 2.1385, 2.5942], + device='cuda:3'), covar=tensor([0.1574, 0.3510, 0.3188, 0.2489, 0.0556, 0.0282, 0.0341, 0.0314], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0302, 0.0331, 0.0251, 0.0220, 0.0164, 0.0207, 0.0216], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:25:05,105 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 01:25:28,931 INFO [train.py:903] (3/4) Epoch 13, batch 6500, loss[loss=0.2149, simple_loss=0.3004, pruned_loss=0.06466, over 19717.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3013, pruned_loss=0.07563, over 3814297.38 frames. ], batch size: 59, lr: 6.29e-03, grad_scale: 8.0 +2023-04-02 01:25:30,963 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 01:25:41,356 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-02 01:26:00,497 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=88462.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:26:26,785 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.125e+02 5.190e+02 6.145e+02 7.751e+02 1.614e+03, threshold=1.229e+03, percent-clipped=4.0 +2023-04-02 01:26:30,118 INFO [train.py:903] (3/4) Epoch 13, batch 6550, loss[loss=0.2222, simple_loss=0.2994, pruned_loss=0.07251, over 19303.00 frames. ], tot_loss[loss=0.2251, simple_loss=0.3002, pruned_loss=0.075, over 3826792.57 frames. ], batch size: 44, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:27:31,538 INFO [train.py:903] (3/4) Epoch 13, batch 6600, loss[loss=0.2272, simple_loss=0.3116, pruned_loss=0.07143, over 19104.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.2993, pruned_loss=0.07448, over 3818581.63 frames. ], batch size: 69, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:28:21,368 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88577.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:28:28,438 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.648e+02 5.175e+02 6.632e+02 9.239e+02 1.861e+03, threshold=1.326e+03, percent-clipped=7.0 +2023-04-02 01:28:32,110 INFO [train.py:903] (3/4) Epoch 13, batch 6650, loss[loss=0.1897, simple_loss=0.2797, pruned_loss=0.04987, over 19613.00 frames. ], tot_loss[loss=0.2266, simple_loss=0.3014, pruned_loss=0.07592, over 3827603.07 frames. ], batch size: 57, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:29:34,641 INFO [train.py:903] (3/4) Epoch 13, batch 6700, loss[loss=0.2114, simple_loss=0.2925, pruned_loss=0.06517, over 19667.00 frames. ], tot_loss[loss=0.2266, simple_loss=0.3016, pruned_loss=0.07583, over 3826097.85 frames. ], batch size: 60, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:29:47,546 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1991, 1.1167, 1.7212, 0.8970, 2.3404, 3.0476, 2.7888, 3.2202], + device='cuda:3'), covar=tensor([0.1547, 0.3715, 0.3097, 0.2438, 0.0570, 0.0203, 0.0241, 0.0263], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0300, 0.0330, 0.0251, 0.0220, 0.0164, 0.0207, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:29:55,535 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=88653.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 01:30:29,674 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.623e+02 5.222e+02 6.682e+02 8.627e+02 1.913e+03, threshold=1.336e+03, percent-clipped=6.0 +2023-04-02 01:30:33,148 INFO [train.py:903] (3/4) Epoch 13, batch 6750, loss[loss=0.2545, simple_loss=0.3277, pruned_loss=0.09059, over 19678.00 frames. ], tot_loss[loss=0.2269, simple_loss=0.3017, pruned_loss=0.07601, over 3834859.73 frames. ], batch size: 59, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:31:31,214 INFO [train.py:903] (3/4) Epoch 13, batch 6800, loss[loss=0.1736, simple_loss=0.2475, pruned_loss=0.04988, over 19772.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.3027, pruned_loss=0.07645, over 3836425.76 frames. ], batch size: 47, lr: 6.28e-03, grad_scale: 8.0 +2023-04-02 01:32:17,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 01:32:18,136 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 01:32:21,842 INFO [train.py:903] (3/4) Epoch 14, batch 0, loss[loss=0.2774, simple_loss=0.3277, pruned_loss=0.1135, over 18658.00 frames. ], tot_loss[loss=0.2774, simple_loss=0.3277, pruned_loss=0.1135, over 18658.00 frames. ], batch size: 41, lr: 6.05e-03, grad_scale: 8.0 +2023-04-02 01:32:21,843 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 01:32:33,649 INFO [train.py:937] (3/4) Epoch 14, validation: loss=0.1763, simple_loss=0.2772, pruned_loss=0.03773, over 944034.00 frames. +2023-04-02 01:32:33,651 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 01:32:41,833 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88768.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:32:49,771 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 01:32:59,169 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.227e+02 4.998e+02 6.377e+02 7.989e+02 1.719e+03, threshold=1.275e+03, percent-clipped=2.0 +2023-04-02 01:32:59,657 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2711, 2.1049, 1.8370, 1.7135, 1.4816, 1.6932, 0.4047, 1.0699], + device='cuda:3'), covar=tensor([0.0454, 0.0500, 0.0433, 0.0653, 0.1080, 0.0742, 0.1129, 0.0887], + device='cuda:3'), in_proj_covar=tensor([0.0340, 0.0336, 0.0331, 0.0357, 0.0432, 0.0357, 0.0315, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:33:40,052 INFO [train.py:903] (3/4) Epoch 14, batch 50, loss[loss=0.2404, simple_loss=0.3171, pruned_loss=0.08183, over 18188.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2951, pruned_loss=0.0704, over 875453.65 frames. ], batch size: 83, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:34:01,734 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88833.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:34:04,147 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3848, 2.0223, 1.6779, 1.1696, 2.0683, 1.0946, 1.3226, 1.9142], + device='cuda:3'), covar=tensor([0.0770, 0.0664, 0.0856, 0.0826, 0.0380, 0.1186, 0.0616, 0.0342], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0308, 0.0329, 0.0251, 0.0241, 0.0326, 0.0298, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:34:15,378 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 01:34:34,167 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88858.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:34:40,591 INFO [train.py:903] (3/4) Epoch 14, batch 100, loss[loss=0.215, simple_loss=0.2974, pruned_loss=0.06626, over 19607.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.3005, pruned_loss=0.07348, over 1530213.02 frames. ], batch size: 57, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:34:51,080 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 01:35:02,786 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.971e+02 5.271e+02 6.621e+02 8.700e+02 2.391e+03, threshold=1.324e+03, percent-clipped=10.0 +2023-04-02 01:35:41,059 INFO [train.py:903] (3/4) Epoch 14, batch 150, loss[loss=0.2444, simple_loss=0.3126, pruned_loss=0.08815, over 19508.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3024, pruned_loss=0.07586, over 2023137.08 frames. ], batch size: 64, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:35:53,914 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4107, 1.2847, 1.3196, 1.7251, 1.2916, 1.6152, 1.6425, 1.5160], + device='cuda:3'), covar=tensor([0.0844, 0.0981, 0.1078, 0.0750, 0.0855, 0.0782, 0.0888, 0.0693], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0223, 0.0225, 0.0244, 0.0228, 0.0210, 0.0193, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 01:36:00,305 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-02 01:36:29,599 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0504, 1.1490, 1.4539, 0.6481, 1.9154, 2.2190, 1.9238, 2.2639], + device='cuda:3'), covar=tensor([0.1410, 0.3320, 0.2984, 0.2460, 0.0665, 0.0365, 0.0396, 0.0389], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0301, 0.0331, 0.0251, 0.0220, 0.0165, 0.0207, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:36:38,762 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 01:36:39,937 INFO [train.py:903] (3/4) Epoch 14, batch 200, loss[loss=0.176, simple_loss=0.2527, pruned_loss=0.04966, over 18718.00 frames. ], tot_loss[loss=0.2279, simple_loss=0.3035, pruned_loss=0.07614, over 2423476.62 frames. ], batch size: 41, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:37:03,953 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.747e+02 5.170e+02 6.545e+02 8.695e+02 1.666e+03, threshold=1.309e+03, percent-clipped=4.0 +2023-04-02 01:37:37,778 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7925, 4.2099, 4.4289, 4.4274, 1.6315, 4.1193, 3.5190, 4.1017], + device='cuda:3'), covar=tensor([0.1441, 0.0743, 0.0641, 0.0655, 0.5269, 0.0706, 0.0679, 0.1266], + device='cuda:3'), in_proj_covar=tensor([0.0707, 0.0631, 0.0835, 0.0715, 0.0755, 0.0583, 0.0503, 0.0765], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 01:37:41,112 INFO [train.py:903] (3/4) Epoch 14, batch 250, loss[loss=0.2207, simple_loss=0.2963, pruned_loss=0.07261, over 19871.00 frames. ], tot_loss[loss=0.227, simple_loss=0.3025, pruned_loss=0.07577, over 2731796.67 frames. ], batch size: 52, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:37:53,623 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89024.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:38:17,672 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89045.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:38:22,455 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89049.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 01:38:43,482 INFO [train.py:903] (3/4) Epoch 14, batch 300, loss[loss=0.1923, simple_loss=0.2671, pruned_loss=0.05878, over 19417.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3022, pruned_loss=0.07521, over 2979031.44 frames. ], batch size: 48, lr: 6.04e-03, grad_scale: 8.0 +2023-04-02 01:38:52,798 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9366, 5.0344, 5.8755, 5.7853, 1.9393, 5.4604, 4.6230, 5.4714], + device='cuda:3'), covar=tensor([0.1388, 0.0766, 0.0461, 0.0539, 0.5252, 0.0629, 0.0550, 0.0949], + device='cuda:3'), in_proj_covar=tensor([0.0707, 0.0630, 0.0836, 0.0715, 0.0753, 0.0583, 0.0504, 0.0764], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 01:39:05,433 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.258e+02 4.830e+02 5.923e+02 7.544e+02 2.111e+03, threshold=1.185e+03, percent-clipped=1.0 +2023-04-02 01:39:45,150 INFO [train.py:903] (3/4) Epoch 14, batch 350, loss[loss=0.2134, simple_loss=0.2907, pruned_loss=0.06804, over 19761.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3019, pruned_loss=0.07524, over 3171945.82 frames. ], batch size: 54, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:39:47,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 01:40:29,935 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9312, 2.0029, 2.1735, 2.7897, 1.8727, 2.5494, 2.3721, 1.9720], + device='cuda:3'), covar=tensor([0.3867, 0.3549, 0.1649, 0.1888, 0.3863, 0.1691, 0.3867, 0.2936], + device='cuda:3'), in_proj_covar=tensor([0.0824, 0.0859, 0.0666, 0.0908, 0.0807, 0.0738, 0.0811, 0.0728], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 01:40:46,829 INFO [train.py:903] (3/4) Epoch 14, batch 400, loss[loss=0.2023, simple_loss=0.2908, pruned_loss=0.05685, over 18197.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3016, pruned_loss=0.07489, over 3314809.11 frames. ], batch size: 84, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:41:11,934 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.999e+02 4.893e+02 5.942e+02 7.582e+02 1.529e+03, threshold=1.188e+03, percent-clipped=4.0 +2023-04-02 01:41:16,767 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89187.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:41:36,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4515, 1.1882, 1.1181, 1.3305, 1.0362, 1.2629, 1.1313, 1.3214], + device='cuda:3'), covar=tensor([0.0971, 0.1225, 0.1439, 0.0870, 0.1175, 0.0576, 0.1302, 0.0731], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0353, 0.0297, 0.0240, 0.0298, 0.0243, 0.0285, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:41:47,619 INFO [train.py:903] (3/4) Epoch 14, batch 450, loss[loss=0.2724, simple_loss=0.3355, pruned_loss=0.1046, over 19583.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3017, pruned_loss=0.07481, over 3434119.54 frames. ], batch size: 61, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:42:19,964 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 01:42:20,930 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 01:42:51,706 INFO [train.py:903] (3/4) Epoch 14, batch 500, loss[loss=0.2133, simple_loss=0.2873, pruned_loss=0.06965, over 19406.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3019, pruned_loss=0.0753, over 3536322.78 frames. ], batch size: 48, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:43:04,782 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7432, 1.7110, 1.4460, 1.7640, 1.6969, 1.4240, 1.3939, 1.6727], + device='cuda:3'), covar=tensor([0.1037, 0.1326, 0.1508, 0.0947, 0.1136, 0.0701, 0.1431, 0.0765], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0350, 0.0295, 0.0239, 0.0295, 0.0241, 0.0285, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:43:13,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.040e+02 5.086e+02 6.501e+02 8.394e+02 1.936e+03, threshold=1.300e+03, percent-clipped=6.0 +2023-04-02 01:43:51,223 INFO [train.py:903] (3/4) Epoch 14, batch 550, loss[loss=0.2593, simple_loss=0.3214, pruned_loss=0.09867, over 13083.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3016, pruned_loss=0.07488, over 3601448.24 frames. ], batch size: 137, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:44:05,921 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 01:44:50,890 INFO [train.py:903] (3/4) Epoch 14, batch 600, loss[loss=0.2236, simple_loss=0.3056, pruned_loss=0.07079, over 19662.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3022, pruned_loss=0.07509, over 3650208.88 frames. ], batch size: 55, lr: 6.03e-03, grad_scale: 8.0 +2023-04-02 01:45:14,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.733e+02 5.437e+02 6.463e+02 8.394e+02 1.645e+03, threshold=1.293e+03, percent-clipped=4.0 +2023-04-02 01:45:23,674 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89389.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:45:27,406 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89392.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:45:32,884 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 01:45:52,144 INFO [train.py:903] (3/4) Epoch 14, batch 650, loss[loss=0.1974, simple_loss=0.27, pruned_loss=0.06236, over 19319.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.302, pruned_loss=0.07529, over 3683130.81 frames. ], batch size: 44, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:46:29,575 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89443.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:46:45,786 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89457.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:46:53,907 INFO [train.py:903] (3/4) Epoch 14, batch 700, loss[loss=0.2229, simple_loss=0.2883, pruned_loss=0.07875, over 19748.00 frames. ], tot_loss[loss=0.2257, simple_loss=0.3013, pruned_loss=0.07501, over 3718749.50 frames. ], batch size: 47, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:47:21,077 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.092e+02 5.132e+02 6.716e+02 8.076e+02 1.293e+03, threshold=1.343e+03, percent-clipped=1.0 +2023-04-02 01:47:44,510 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89504.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:47:59,822 INFO [train.py:903] (3/4) Epoch 14, batch 750, loss[loss=0.2245, simple_loss=0.3008, pruned_loss=0.07408, over 19532.00 frames. ], tot_loss[loss=0.225, simple_loss=0.3007, pruned_loss=0.07464, over 3744619.21 frames. ], batch size: 54, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:48:14,756 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89527.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:48:19,171 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89531.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:48:30,293 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4988, 2.2123, 2.1982, 2.6867, 2.3694, 2.3993, 2.0403, 2.7078], + device='cuda:3'), covar=tensor([0.0906, 0.1635, 0.1321, 0.1005, 0.1346, 0.0435, 0.1230, 0.0588], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0350, 0.0294, 0.0239, 0.0294, 0.0241, 0.0284, 0.0240], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:48:48,362 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-02 01:49:01,551 INFO [train.py:903] (3/4) Epoch 14, batch 800, loss[loss=0.2158, simple_loss=0.2999, pruned_loss=0.0659, over 19686.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3018, pruned_loss=0.07527, over 3757410.13 frames. ], batch size: 59, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:49:16,442 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 01:49:24,442 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.547e+02 5.343e+02 6.523e+02 8.164e+02 1.688e+03, threshold=1.305e+03, percent-clipped=4.0 +2023-04-02 01:50:01,792 INFO [train.py:903] (3/4) Epoch 14, batch 850, loss[loss=0.2712, simple_loss=0.3399, pruned_loss=0.1012, over 19672.00 frames. ], tot_loss[loss=0.2267, simple_loss=0.3024, pruned_loss=0.07547, over 3779765.35 frames. ], batch size: 55, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:50:38,079 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3013, 2.0444, 1.5360, 1.1662, 2.0154, 1.1244, 1.1257, 1.8062], + device='cuda:3'), covar=tensor([0.0968, 0.0738, 0.1016, 0.0956, 0.0447, 0.1274, 0.0769, 0.0446], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0307, 0.0328, 0.0250, 0.0240, 0.0326, 0.0297, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:50:42,937 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89646.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:50:55,560 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 01:51:03,554 INFO [train.py:903] (3/4) Epoch 14, batch 900, loss[loss=0.2209, simple_loss=0.2909, pruned_loss=0.07545, over 19115.00 frames. ], tot_loss[loss=0.2267, simple_loss=0.3022, pruned_loss=0.07557, over 3792416.44 frames. ], batch size: 42, lr: 6.02e-03, grad_scale: 8.0 +2023-04-02 01:51:29,677 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.240e+02 4.983e+02 6.131e+02 7.467e+02 1.791e+03, threshold=1.226e+03, percent-clipped=3.0 +2023-04-02 01:51:46,412 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89698.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:52:07,438 INFO [train.py:903] (3/4) Epoch 14, batch 950, loss[loss=0.2488, simple_loss=0.3269, pruned_loss=0.08535, over 19515.00 frames. ], tot_loss[loss=0.2277, simple_loss=0.3031, pruned_loss=0.07614, over 3799097.96 frames. ], batch size: 56, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:52:07,496 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 01:52:33,961 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89736.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:53:06,438 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89760.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:53:10,802 INFO [train.py:903] (3/4) Epoch 14, batch 1000, loss[loss=0.249, simple_loss=0.3186, pruned_loss=0.08966, over 19534.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.3009, pruned_loss=0.07514, over 3814504.45 frames. ], batch size: 54, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:53:34,652 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.523e+02 5.300e+02 6.720e+02 8.420e+02 1.635e+03, threshold=1.344e+03, percent-clipped=5.0 +2023-04-02 01:53:36,448 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89785.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:53:38,576 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89787.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:53:59,176 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89801.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:54:02,293 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 01:54:13,791 INFO [train.py:903] (3/4) Epoch 14, batch 1050, loss[loss=0.2525, simple_loss=0.3246, pruned_loss=0.09015, over 19543.00 frames. ], tot_loss[loss=0.2252, simple_loss=0.3003, pruned_loss=0.0751, over 3815290.15 frames. ], batch size: 64, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:54:46,493 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 01:55:02,240 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89851.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:55:16,491 INFO [train.py:903] (3/4) Epoch 14, batch 1100, loss[loss=0.2425, simple_loss=0.317, pruned_loss=0.084, over 19675.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.302, pruned_loss=0.0761, over 3812193.99 frames. ], batch size: 55, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:55:24,741 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=89871.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:55:42,988 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89883.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:55:43,840 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.219e+02 5.450e+02 6.738e+02 9.006e+02 2.173e+03, threshold=1.348e+03, percent-clipped=6.0 +2023-04-02 01:56:05,616 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89902.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:56:05,691 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89902.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:56:19,139 INFO [train.py:903] (3/4) Epoch 14, batch 1150, loss[loss=0.1898, simple_loss=0.2772, pruned_loss=0.05118, over 19635.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.3007, pruned_loss=0.07522, over 3809231.01 frames. ], batch size: 53, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:56:24,155 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89916.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 01:56:39,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:56:54,475 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8820, 4.3691, 2.7222, 3.8591, 0.9159, 4.2515, 4.2252, 4.3292], + device='cuda:3'), covar=tensor([0.0580, 0.1011, 0.1946, 0.0807, 0.4419, 0.0695, 0.0787, 0.0900], + device='cuda:3'), in_proj_covar=tensor([0.0443, 0.0371, 0.0446, 0.0324, 0.0390, 0.0381, 0.0374, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 01:57:25,197 INFO [train.py:903] (3/4) Epoch 14, batch 1200, loss[loss=0.2191, simple_loss=0.2948, pruned_loss=0.07167, over 19777.00 frames. ], tot_loss[loss=0.2245, simple_loss=0.3, pruned_loss=0.07447, over 3828373.52 frames. ], batch size: 49, lr: 6.01e-03, grad_scale: 8.0 +2023-04-02 01:57:49,395 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.071e+02 4.889e+02 5.798e+02 7.146e+02 1.027e+03, threshold=1.160e+03, percent-clipped=0.0 +2023-04-02 01:57:51,938 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89986.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:57:55,114 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 01:58:28,856 INFO [train.py:903] (3/4) Epoch 14, batch 1250, loss[loss=0.2354, simple_loss=0.3103, pruned_loss=0.08029, over 19657.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2993, pruned_loss=0.07408, over 3839939.11 frames. ], batch size: 58, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 01:59:04,881 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90042.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 01:59:31,258 INFO [train.py:903] (3/4) Epoch 14, batch 1300, loss[loss=0.2748, simple_loss=0.3334, pruned_loss=0.1081, over 13623.00 frames. ], tot_loss[loss=0.2248, simple_loss=0.3003, pruned_loss=0.07461, over 3815045.94 frames. ], batch size: 136, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 01:59:36,538 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1887, 2.7890, 2.0872, 2.0003, 1.9557, 2.3673, 0.9127, 2.0087], + device='cuda:3'), covar=tensor([0.0594, 0.0566, 0.0573, 0.1013, 0.0915, 0.0970, 0.1116, 0.0899], + device='cuda:3'), in_proj_covar=tensor([0.0344, 0.0337, 0.0335, 0.0364, 0.0438, 0.0361, 0.0319, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 01:59:57,898 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.456e+02 5.410e+02 6.769e+02 7.813e+02 2.183e+03, threshold=1.354e+03, percent-clipped=2.0 +2023-04-02 02:00:02,808 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0941, 1.7656, 1.3627, 1.1143, 1.6034, 1.0502, 1.1744, 1.5453], + device='cuda:3'), covar=tensor([0.0691, 0.0706, 0.0967, 0.0712, 0.0435, 0.1166, 0.0553, 0.0398], + device='cuda:3'), in_proj_covar=tensor([0.0298, 0.0307, 0.0327, 0.0250, 0.0238, 0.0326, 0.0295, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:00:25,738 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90107.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:00:33,588 INFO [train.py:903] (3/4) Epoch 14, batch 1350, loss[loss=0.2144, simple_loss=0.2966, pruned_loss=0.06604, over 19574.00 frames. ], tot_loss[loss=0.2244, simple_loss=0.3, pruned_loss=0.07441, over 3817998.09 frames. ], batch size: 61, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 02:00:58,168 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90132.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:01:16,459 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90148.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:01:24,049 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-02 02:01:28,069 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90157.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:01:29,295 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90158.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:01:35,839 INFO [train.py:903] (3/4) Epoch 14, batch 1400, loss[loss=0.2139, simple_loss=0.29, pruned_loss=0.06889, over 19375.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2989, pruned_loss=0.07396, over 3810523.35 frames. ], batch size: 70, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 02:01:46,557 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90172.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:01:59,350 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90183.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:02:00,108 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.442e+02 5.450e+02 7.234e+02 1.006e+03 2.886e+03, threshold=1.447e+03, percent-clipped=11.0 +2023-04-02 02:02:17,962 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90197.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:02:34,957 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 02:02:38,345 INFO [train.py:903] (3/4) Epoch 14, batch 1450, loss[loss=0.2141, simple_loss=0.2987, pruned_loss=0.06477, over 19781.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.2995, pruned_loss=0.07438, over 3820478.68 frames. ], batch size: 56, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 02:02:53,329 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90227.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:03:12,044 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90242.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:03:22,213 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90250.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:03:38,201 INFO [train.py:903] (3/4) Epoch 14, batch 1500, loss[loss=0.2048, simple_loss=0.2688, pruned_loss=0.07042, over 19711.00 frames. ], tot_loss[loss=0.2243, simple_loss=0.3, pruned_loss=0.07431, over 3830181.98 frames. ], batch size: 45, lr: 6.00e-03, grad_scale: 8.0 +2023-04-02 02:03:42,164 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90267.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:04:03,122 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.947e+02 5.072e+02 6.110e+02 7.921e+02 2.000e+03, threshold=1.222e+03, percent-clipped=2.0 +2023-04-02 02:04:09,181 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2414, 3.7323, 3.8575, 3.8507, 1.6509, 3.6180, 3.1594, 3.5887], + device='cuda:3'), covar=tensor([0.1440, 0.0980, 0.0570, 0.0668, 0.4805, 0.0814, 0.0643, 0.1043], + device='cuda:3'), in_proj_covar=tensor([0.0700, 0.0624, 0.0827, 0.0710, 0.0746, 0.0576, 0.0499, 0.0764], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:04:23,088 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4769, 2.3882, 1.6192, 1.3780, 2.2250, 1.1860, 1.1129, 1.9211], + device='cuda:3'), covar=tensor([0.1050, 0.0622, 0.0974, 0.0799, 0.0389, 0.1251, 0.0870, 0.0490], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0308, 0.0327, 0.0251, 0.0238, 0.0327, 0.0298, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:04:39,093 INFO [train.py:903] (3/4) Epoch 14, batch 1550, loss[loss=0.2885, simple_loss=0.3472, pruned_loss=0.1149, over 13381.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.3009, pruned_loss=0.0755, over 3800520.85 frames. ], batch size: 136, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:05:16,349 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90342.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:05:42,816 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9317, 1.3454, 1.0433, 0.9282, 1.1961, 0.9508, 0.9398, 1.2613], + device='cuda:3'), covar=tensor([0.0529, 0.0722, 0.1040, 0.0645, 0.0488, 0.1166, 0.0537, 0.0420], + device='cuda:3'), in_proj_covar=tensor([0.0298, 0.0308, 0.0327, 0.0250, 0.0239, 0.0327, 0.0297, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:05:44,734 INFO [train.py:903] (3/4) Epoch 14, batch 1600, loss[loss=0.2446, simple_loss=0.3178, pruned_loss=0.08565, over 18197.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3011, pruned_loss=0.0758, over 3809621.35 frames. ], batch size: 83, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:05:51,741 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90369.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:06:02,642 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 02:06:08,371 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.055e+02 5.288e+02 6.387e+02 7.705e+02 1.564e+03, threshold=1.277e+03, percent-clipped=2.0 +2023-04-02 02:06:46,521 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90413.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:06:47,288 INFO [train.py:903] (3/4) Epoch 14, batch 1650, loss[loss=0.2128, simple_loss=0.2992, pruned_loss=0.06324, over 18624.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.301, pruned_loss=0.07542, over 3813640.03 frames. ], batch size: 74, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:07:15,928 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90438.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:07:47,909 INFO [train.py:903] (3/4) Epoch 14, batch 1700, loss[loss=0.2485, simple_loss=0.3224, pruned_loss=0.08729, over 19663.00 frames. ], tot_loss[loss=0.2264, simple_loss=0.3013, pruned_loss=0.07571, over 3808521.56 frames. ], batch size: 59, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:07:50,700 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7815, 1.8832, 2.0256, 2.3352, 1.6276, 2.1785, 2.1792, 1.9391], + device='cuda:3'), covar=tensor([0.3493, 0.2895, 0.1572, 0.1841, 0.3090, 0.1612, 0.3841, 0.2762], + device='cuda:3'), in_proj_covar=tensor([0.0817, 0.0852, 0.0665, 0.0893, 0.0797, 0.0737, 0.0800, 0.0729], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 02:08:03,253 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:08:13,896 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.443e+02 5.354e+02 6.441e+02 8.114e+02 1.316e+03, threshold=1.288e+03, percent-clipped=1.0 +2023-04-02 02:08:19,178 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-02 02:08:23,199 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90492.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:08:25,347 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 02:08:49,500 INFO [train.py:903] (3/4) Epoch 14, batch 1750, loss[loss=0.1908, simple_loss=0.2798, pruned_loss=0.05092, over 19673.00 frames. ], tot_loss[loss=0.2272, simple_loss=0.302, pruned_loss=0.07624, over 3798265.05 frames. ], batch size: 53, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:08:49,818 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90514.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:09:54,000 INFO [train.py:903] (3/4) Epoch 14, batch 1800, loss[loss=0.252, simple_loss=0.3373, pruned_loss=0.08333, over 19677.00 frames. ], tot_loss[loss=0.2273, simple_loss=0.3022, pruned_loss=0.07618, over 3797754.30 frames. ], batch size: 60, lr: 5.99e-03, grad_scale: 8.0 +2023-04-02 02:10:17,895 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.87 vs. limit=5.0 +2023-04-02 02:10:18,178 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.749e+02 5.076e+02 6.157e+02 8.007e+02 1.318e+03, threshold=1.231e+03, percent-clipped=1.0 +2023-04-02 02:10:29,854 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90594.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:10:34,922 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90598.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:10:40,544 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-02 02:10:47,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90607.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:10:48,804 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 02:10:55,678 INFO [train.py:903] (3/4) Epoch 14, batch 1850, loss[loss=0.2491, simple_loss=0.3216, pruned_loss=0.08834, over 19621.00 frames. ], tot_loss[loss=0.2269, simple_loss=0.302, pruned_loss=0.07585, over 3793934.28 frames. ], batch size: 57, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:11:06,608 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 02:11:07,265 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90623.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:11:24,428 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 02:11:30,936 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9702, 3.5560, 2.4300, 3.2004, 1.1117, 3.4890, 3.3941, 3.4861], + device='cuda:3'), covar=tensor([0.0770, 0.1099, 0.1995, 0.0919, 0.3539, 0.0761, 0.0848, 0.1130], + device='cuda:3'), in_proj_covar=tensor([0.0448, 0.0374, 0.0451, 0.0326, 0.0391, 0.0384, 0.0379, 0.0412], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:11:38,566 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-04-02 02:11:58,529 INFO [train.py:903] (3/4) Epoch 14, batch 1900, loss[loss=0.1817, simple_loss=0.2593, pruned_loss=0.05206, over 19792.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3015, pruned_loss=0.07552, over 3795465.80 frames. ], batch size: 48, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:12:08,016 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5501, 4.1126, 4.2520, 4.2387, 1.6181, 3.9707, 3.4783, 3.9976], + device='cuda:3'), covar=tensor([0.1550, 0.0702, 0.0571, 0.0635, 0.5273, 0.0719, 0.0655, 0.1011], + device='cuda:3'), in_proj_covar=tensor([0.0713, 0.0635, 0.0844, 0.0730, 0.0758, 0.0590, 0.0508, 0.0778], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:12:12,362 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 02:12:13,986 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8703, 1.3612, 1.0709, 0.9680, 1.1888, 0.9935, 0.9305, 1.2849], + device='cuda:3'), covar=tensor([0.0585, 0.0756, 0.0996, 0.0595, 0.0473, 0.1167, 0.0559, 0.0431], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0309, 0.0329, 0.0252, 0.0241, 0.0328, 0.0297, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:12:15,962 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 02:12:18,385 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2716, 3.8194, 3.9063, 3.9035, 1.4436, 3.6866, 3.2562, 3.6270], + device='cuda:3'), covar=tensor([0.1513, 0.0773, 0.0572, 0.0654, 0.5195, 0.0811, 0.0639, 0.1065], + device='cuda:3'), in_proj_covar=tensor([0.0713, 0.0634, 0.0842, 0.0729, 0.0756, 0.0590, 0.0507, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:12:24,251 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.000e+02 5.065e+02 6.349e+02 7.558e+02 1.663e+03, threshold=1.270e+03, percent-clipped=1.0 +2023-04-02 02:12:29,232 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0385, 4.4040, 4.6952, 4.7194, 1.6459, 4.3831, 3.8694, 4.3422], + device='cuda:3'), covar=tensor([0.1408, 0.0776, 0.0578, 0.0549, 0.5447, 0.0728, 0.0609, 0.1144], + device='cuda:3'), in_proj_covar=tensor([0.0713, 0.0633, 0.0842, 0.0729, 0.0756, 0.0590, 0.0507, 0.0778], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:12:43,840 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 02:12:48,695 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2694, 1.2141, 1.2425, 1.2849, 1.0018, 1.3836, 1.3560, 1.3348], + device='cuda:3'), covar=tensor([0.0846, 0.0929, 0.1035, 0.0720, 0.0834, 0.0764, 0.0796, 0.0729], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0222, 0.0223, 0.0243, 0.0230, 0.0209, 0.0192, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 02:12:54,580 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90709.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:12:59,232 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90713.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:13:00,238 INFO [train.py:903] (3/4) Epoch 14, batch 1950, loss[loss=0.1883, simple_loss=0.2815, pruned_loss=0.04755, over 19677.00 frames. ], tot_loss[loss=0.2265, simple_loss=0.3014, pruned_loss=0.07576, over 3789128.54 frames. ], batch size: 58, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:13:40,058 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2100, 2.8558, 2.2161, 2.1986, 2.0375, 2.5419, 1.0609, 2.0324], + device='cuda:3'), covar=tensor([0.0503, 0.0474, 0.0512, 0.0883, 0.0873, 0.0786, 0.1034, 0.0845], + device='cuda:3'), in_proj_covar=tensor([0.0339, 0.0336, 0.0331, 0.0359, 0.0434, 0.0358, 0.0315, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 02:14:03,111 INFO [train.py:903] (3/4) Epoch 14, batch 2000, loss[loss=0.2079, simple_loss=0.2961, pruned_loss=0.05984, over 19308.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3025, pruned_loss=0.07586, over 3796497.23 frames. ], batch size: 66, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:14:22,427 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6392, 1.2651, 1.3957, 1.6434, 3.1662, 1.0617, 2.3069, 3.5847], + device='cuda:3'), covar=tensor([0.0440, 0.2724, 0.2769, 0.1671, 0.0723, 0.2510, 0.1151, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0347, 0.0360, 0.0329, 0.0351, 0.0335, 0.0346, 0.0368], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:14:27,795 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.086e+02 5.207e+02 5.999e+02 7.179e+02 1.269e+03, threshold=1.200e+03, percent-clipped=0.0 +2023-04-02 02:14:59,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 02:15:05,779 INFO [train.py:903] (3/4) Epoch 14, batch 2050, loss[loss=0.2087, simple_loss=0.2894, pruned_loss=0.06399, over 19507.00 frames. ], tot_loss[loss=0.2268, simple_loss=0.3024, pruned_loss=0.07559, over 3798769.55 frames. ], batch size: 54, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:15:14,228 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90821.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:15:18,963 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 02:15:20,142 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 02:15:24,013 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90828.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:15:42,057 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 02:16:01,449 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=90858.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:16:07,309 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90863.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:16:08,091 INFO [train.py:903] (3/4) Epoch 14, batch 2100, loss[loss=0.2416, simple_loss=0.3229, pruned_loss=0.0802, over 18252.00 frames. ], tot_loss[loss=0.2269, simple_loss=0.3025, pruned_loss=0.07563, over 3803335.62 frames. ], batch size: 83, lr: 5.98e-03, grad_scale: 8.0 +2023-04-02 02:16:12,852 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90867.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:16:32,690 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.435e+02 5.132e+02 6.276e+02 7.348e+02 1.970e+03, threshold=1.255e+03, percent-clipped=2.0 +2023-04-02 02:16:37,022 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 02:16:39,478 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90888.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:16:46,421 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4857, 1.2030, 1.1399, 1.4015, 1.1114, 1.2662, 1.1577, 1.3333], + device='cuda:3'), covar=tensor([0.1117, 0.1237, 0.1525, 0.1012, 0.1174, 0.0596, 0.1343, 0.0815], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0354, 0.0296, 0.0240, 0.0295, 0.0240, 0.0288, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:16:58,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 02:17:09,974 INFO [train.py:903] (3/4) Epoch 14, batch 2150, loss[loss=0.2014, simple_loss=0.2872, pruned_loss=0.05777, over 19772.00 frames. ], tot_loss[loss=0.2252, simple_loss=0.301, pruned_loss=0.07468, over 3819415.31 frames. ], batch size: 54, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:17:38,852 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90936.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:17:53,046 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-02 02:18:11,004 INFO [train.py:903] (3/4) Epoch 14, batch 2200, loss[loss=0.1977, simple_loss=0.2666, pruned_loss=0.06445, over 19744.00 frames. ], tot_loss[loss=0.2249, simple_loss=0.3005, pruned_loss=0.07464, over 3820187.37 frames. ], batch size: 46, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:18:12,477 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90965.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:18:19,878 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2293, 1.5172, 2.0412, 1.4816, 2.9607, 4.6689, 4.5738, 5.0182], + device='cuda:3'), covar=tensor([0.1622, 0.3406, 0.2914, 0.2049, 0.0589, 0.0172, 0.0143, 0.0117], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0302, 0.0332, 0.0254, 0.0225, 0.0166, 0.0206, 0.0218], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 02:18:23,373 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90973.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:18:35,015 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.76 vs. limit=5.0 +2023-04-02 02:18:35,325 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.217e+02 5.132e+02 5.971e+02 7.684e+02 1.888e+03, threshold=1.194e+03, percent-clipped=2.0 +2023-04-02 02:18:43,460 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90990.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:19:13,346 INFO [train.py:903] (3/4) Epoch 14, batch 2250, loss[loss=0.1995, simple_loss=0.2785, pruned_loss=0.06024, over 19528.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2992, pruned_loss=0.07397, over 3810020.02 frames. ], batch size: 54, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:20:14,197 INFO [train.py:903] (3/4) Epoch 14, batch 2300, loss[loss=0.2327, simple_loss=0.3173, pruned_loss=0.07406, over 19494.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3001, pruned_loss=0.07459, over 3807787.08 frames. ], batch size: 64, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:20:26,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 02:20:38,185 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.297e+02 5.417e+02 6.829e+02 8.730e+02 1.535e+03, threshold=1.366e+03, percent-clipped=12.0 +2023-04-02 02:20:38,721 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91084.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:20:47,622 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4193, 2.1204, 2.4029, 2.7157, 2.3991, 2.2726, 2.2001, 2.6803], + device='cuda:3'), covar=tensor([0.0836, 0.1553, 0.1084, 0.0771, 0.1128, 0.0428, 0.0999, 0.0492], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0350, 0.0294, 0.0239, 0.0293, 0.0239, 0.0285, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:21:11,036 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91109.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:21:16,437 INFO [train.py:903] (3/4) Epoch 14, batch 2350, loss[loss=0.2094, simple_loss=0.2992, pruned_loss=0.05976, over 19610.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3003, pruned_loss=0.07454, over 3815282.98 frames. ], batch size: 57, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:21:57,692 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 02:22:13,116 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 02:22:14,984 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.91 vs. limit=5.0 +2023-04-02 02:22:17,788 INFO [train.py:903] (3/4) Epoch 14, batch 2400, loss[loss=0.2252, simple_loss=0.3007, pruned_loss=0.07486, over 19683.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.2998, pruned_loss=0.07417, over 3815161.90 frames. ], batch size: 55, lr: 5.97e-03, grad_scale: 8.0 +2023-04-02 02:22:42,190 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.550e+02 5.466e+02 6.599e+02 8.174e+02 1.804e+03, threshold=1.320e+03, percent-clipped=5.0 +2023-04-02 02:22:52,592 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6399, 4.1030, 4.2701, 4.2395, 1.6619, 3.9623, 3.5160, 3.9754], + device='cuda:3'), covar=tensor([0.1367, 0.0743, 0.0567, 0.0627, 0.5155, 0.0762, 0.0653, 0.1143], + device='cuda:3'), in_proj_covar=tensor([0.0708, 0.0636, 0.0839, 0.0724, 0.0756, 0.0586, 0.0504, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:22:52,766 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91192.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:23:15,499 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=91211.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:23:19,602 INFO [train.py:903] (3/4) Epoch 14, batch 2450, loss[loss=0.2339, simple_loss=0.3076, pruned_loss=0.08016, over 19776.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3014, pruned_loss=0.07539, over 3805300.08 frames. ], batch size: 63, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:23:23,094 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91217.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:23:37,773 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91229.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:24:08,242 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91254.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:24:19,340 INFO [train.py:903] (3/4) Epoch 14, batch 2500, loss[loss=0.1908, simple_loss=0.259, pruned_loss=0.06127, over 17411.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.301, pruned_loss=0.07504, over 3809837.63 frames. ], batch size: 38, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:24:32,860 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7312, 1.6372, 1.5017, 2.1248, 1.6467, 2.0387, 2.1287, 1.8963], + device='cuda:3'), covar=tensor([0.0791, 0.0874, 0.1061, 0.0775, 0.0835, 0.0692, 0.0823, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0223, 0.0224, 0.0241, 0.0228, 0.0209, 0.0191, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 02:24:35,621 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-04-02 02:24:39,623 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9712, 5.3563, 3.1068, 4.7602, 1.2590, 5.3051, 5.3367, 5.4926], + device='cuda:3'), covar=tensor([0.0463, 0.1005, 0.1882, 0.0756, 0.4201, 0.0597, 0.0671, 0.0990], + device='cuda:3'), in_proj_covar=tensor([0.0452, 0.0380, 0.0454, 0.0330, 0.0392, 0.0386, 0.0378, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:24:42,800 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.074e+02 5.867e+02 6.968e+02 8.618e+02 1.617e+03, threshold=1.394e+03, percent-clipped=2.0 +2023-04-02 02:24:46,256 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3493, 1.3326, 1.7406, 1.4569, 2.7778, 3.8363, 3.5787, 4.0626], + device='cuda:3'), covar=tensor([0.1486, 0.3419, 0.3000, 0.1967, 0.0515, 0.0146, 0.0170, 0.0175], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0301, 0.0330, 0.0253, 0.0223, 0.0166, 0.0206, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 02:24:58,182 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0761, 2.0424, 1.8462, 2.3396, 2.5530, 1.8311, 1.8637, 2.2611], + device='cuda:3'), covar=tensor([0.1124, 0.1746, 0.1686, 0.1167, 0.1286, 0.0882, 0.1554, 0.0891], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0352, 0.0294, 0.0240, 0.0294, 0.0240, 0.0288, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:25:06,908 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91303.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:25:19,880 INFO [train.py:903] (3/4) Epoch 14, batch 2550, loss[loss=0.2358, simple_loss=0.3141, pruned_loss=0.07877, over 19631.00 frames. ], tot_loss[loss=0.2243, simple_loss=0.2997, pruned_loss=0.07448, over 3819362.16 frames. ], batch size: 57, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:25:33,180 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91326.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:26:12,384 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 02:26:19,133 INFO [train.py:903] (3/4) Epoch 14, batch 2600, loss[loss=0.2234, simple_loss=0.3094, pruned_loss=0.06869, over 19764.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.3011, pruned_loss=0.07521, over 3817028.36 frames. ], batch size: 54, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:26:44,917 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.041e+02 5.557e+02 7.203e+02 9.059e+02 1.995e+03, threshold=1.441e+03, percent-clipped=7.0 +2023-04-02 02:27:21,520 INFO [train.py:903] (3/4) Epoch 14, batch 2650, loss[loss=0.2451, simple_loss=0.3182, pruned_loss=0.08599, over 19524.00 frames. ], tot_loss[loss=0.2254, simple_loss=0.3009, pruned_loss=0.07502, over 3818939.03 frames. ], batch size: 64, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:27:41,101 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 02:28:11,793 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7513, 1.3364, 1.5300, 1.5679, 3.2783, 1.0872, 2.2667, 3.7150], + device='cuda:3'), covar=tensor([0.0427, 0.2579, 0.2786, 0.1728, 0.0684, 0.2459, 0.1299, 0.0218], + device='cuda:3'), in_proj_covar=tensor([0.0371, 0.0346, 0.0363, 0.0327, 0.0355, 0.0338, 0.0347, 0.0369], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:28:18,985 INFO [train.py:903] (3/4) Epoch 14, batch 2700, loss[loss=0.2022, simple_loss=0.2732, pruned_loss=0.06558, over 19355.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3016, pruned_loss=0.07537, over 3815118.78 frames. ], batch size: 47, lr: 5.96e-03, grad_scale: 8.0 +2023-04-02 02:28:43,685 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.192e+02 5.145e+02 6.524e+02 8.606e+02 2.089e+03, threshold=1.305e+03, percent-clipped=4.0 +2023-04-02 02:29:13,058 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.82 vs. limit=5.0 +2023-04-02 02:29:20,024 INFO [train.py:903] (3/4) Epoch 14, batch 2750, loss[loss=0.221, simple_loss=0.3133, pruned_loss=0.06438, over 19129.00 frames. ], tot_loss[loss=0.2254, simple_loss=0.301, pruned_loss=0.07485, over 3815431.70 frames. ], batch size: 69, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:29:44,615 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-02 02:30:14,284 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91560.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:30:18,378 INFO [train.py:903] (3/4) Epoch 14, batch 2800, loss[loss=0.2305, simple_loss=0.3073, pruned_loss=0.07686, over 19850.00 frames. ], tot_loss[loss=0.2271, simple_loss=0.3027, pruned_loss=0.07578, over 3818177.17 frames. ], batch size: 52, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:30:41,199 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91582.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:30:44,141 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.583e+02 4.991e+02 6.176e+02 8.428e+02 2.269e+03, threshold=1.235e+03, percent-clipped=6.0 +2023-04-02 02:31:10,040 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91607.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:31:19,257 INFO [train.py:903] (3/4) Epoch 14, batch 2850, loss[loss=0.2235, simple_loss=0.3016, pruned_loss=0.07265, over 17397.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.3026, pruned_loss=0.07605, over 3820570.64 frames. ], batch size: 101, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:31:58,530 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=91647.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:32:20,013 INFO [train.py:903] (3/4) Epoch 14, batch 2900, loss[loss=0.2435, simple_loss=0.3182, pruned_loss=0.08444, over 17382.00 frames. ], tot_loss[loss=0.2265, simple_loss=0.3023, pruned_loss=0.07538, over 3829099.96 frames. ], batch size: 101, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:32:20,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 02:32:44,081 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.659e+02 5.126e+02 6.310e+02 7.919e+02 2.445e+03, threshold=1.262e+03, percent-clipped=4.0 +2023-04-02 02:33:15,679 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-02 02:33:19,569 INFO [train.py:903] (3/4) Epoch 14, batch 2950, loss[loss=0.2284, simple_loss=0.2999, pruned_loss=0.07846, over 19607.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.3018, pruned_loss=0.07526, over 3830385.86 frames. ], batch size: 50, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:34:17,811 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91762.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:34:19,630 INFO [train.py:903] (3/4) Epoch 14, batch 3000, loss[loss=0.2702, simple_loss=0.3371, pruned_loss=0.1016, over 19731.00 frames. ], tot_loss[loss=0.2264, simple_loss=0.3019, pruned_loss=0.07546, over 3823311.36 frames. ], batch size: 63, lr: 5.95e-03, grad_scale: 8.0 +2023-04-02 02:34:19,631 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 02:34:36,637 INFO [train.py:937] (3/4) Epoch 14, validation: loss=0.1742, simple_loss=0.2751, pruned_loss=0.03671, over 944034.00 frames. +2023-04-02 02:34:36,639 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 02:34:42,025 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 02:34:45,929 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5124, 2.3898, 1.6491, 1.6062, 2.1323, 1.3100, 1.2437, 1.9318], + device='cuda:3'), covar=tensor([0.1036, 0.0635, 0.1012, 0.0690, 0.0503, 0.1176, 0.0819, 0.0496], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0307, 0.0325, 0.0249, 0.0239, 0.0324, 0.0294, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:35:02,710 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.309e+02 5.261e+02 6.370e+02 8.625e+02 1.479e+03, threshold=1.274e+03, percent-clipped=4.0 +2023-04-02 02:35:25,863 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2945, 3.0188, 2.2180, 2.7592, 0.7941, 2.9308, 2.8499, 2.9109], + device='cuda:3'), covar=tensor([0.1171, 0.1403, 0.2018, 0.1032, 0.3899, 0.1002, 0.1115, 0.1344], + device='cuda:3'), in_proj_covar=tensor([0.0450, 0.0380, 0.0454, 0.0329, 0.0390, 0.0388, 0.0380, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:35:37,798 INFO [train.py:903] (3/4) Epoch 14, batch 3050, loss[loss=0.2318, simple_loss=0.3109, pruned_loss=0.07633, over 18718.00 frames. ], tot_loss[loss=0.226, simple_loss=0.3019, pruned_loss=0.07503, over 3830461.19 frames. ], batch size: 74, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:36:09,848 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3080, 2.2193, 1.8743, 1.7394, 1.7448, 1.7881, 0.6260, 1.1521], + device='cuda:3'), covar=tensor([0.0495, 0.0495, 0.0395, 0.0685, 0.0933, 0.0765, 0.0996, 0.0797], + device='cuda:3'), in_proj_covar=tensor([0.0342, 0.0339, 0.0334, 0.0364, 0.0432, 0.0361, 0.0315, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 02:36:12,272 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-02 02:36:37,020 INFO [train.py:903] (3/4) Epoch 14, batch 3100, loss[loss=0.2083, simple_loss=0.2762, pruned_loss=0.07025, over 19702.00 frames. ], tot_loss[loss=0.2266, simple_loss=0.3022, pruned_loss=0.07557, over 3816062.17 frames. ], batch size: 51, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:37:02,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.205e+02 5.361e+02 6.622e+02 8.860e+02 2.580e+03, threshold=1.324e+03, percent-clipped=11.0 +2023-04-02 02:37:09,339 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.54 vs. limit=2.0 +2023-04-02 02:37:24,417 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=91904.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 02:37:31,814 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91909.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:37:37,892 INFO [train.py:903] (3/4) Epoch 14, batch 3150, loss[loss=0.2229, simple_loss=0.2987, pruned_loss=0.07359, over 19515.00 frames. ], tot_loss[loss=0.2261, simple_loss=0.3016, pruned_loss=0.07528, over 3827333.56 frames. ], batch size: 54, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:37:44,909 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.32 vs. limit=5.0 +2023-04-02 02:38:04,260 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 02:38:37,357 INFO [train.py:903] (3/4) Epoch 14, batch 3200, loss[loss=0.2488, simple_loss=0.3288, pruned_loss=0.0844, over 19651.00 frames. ], tot_loss[loss=0.2263, simple_loss=0.3017, pruned_loss=0.07548, over 3830954.92 frames. ], batch size: 55, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:38:48,681 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91973.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:39:02,850 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.458e+02 5.150e+02 6.206e+02 7.874e+02 1.849e+03, threshold=1.241e+03, percent-clipped=5.0 +2023-04-02 02:39:04,390 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6124, 1.2779, 1.5346, 1.5768, 3.1670, 0.9227, 2.1825, 3.4641], + device='cuda:3'), covar=tensor([0.0445, 0.2705, 0.2781, 0.1705, 0.0724, 0.2576, 0.1326, 0.0282], + device='cuda:3'), in_proj_covar=tensor([0.0373, 0.0347, 0.0364, 0.0328, 0.0358, 0.0338, 0.0350, 0.0371], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:39:20,986 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91999.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 02:39:39,708 INFO [train.py:903] (3/4) Epoch 14, batch 3250, loss[loss=0.1873, simple_loss=0.2615, pruned_loss=0.05658, over 19736.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3002, pruned_loss=0.07461, over 3831191.14 frames. ], batch size: 46, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:39:44,736 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92018.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:39:45,902 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92019.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:40:15,594 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92043.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:40:39,291 INFO [train.py:903] (3/4) Epoch 14, batch 3300, loss[loss=0.2679, simple_loss=0.3309, pruned_loss=0.1024, over 19717.00 frames. ], tot_loss[loss=0.2251, simple_loss=0.3008, pruned_loss=0.07467, over 3819317.36 frames. ], batch size: 51, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:40:44,827 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 02:41:01,033 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0338, 1.7163, 1.8713, 1.8498, 4.4420, 0.9454, 2.6718, 4.8295], + device='cuda:3'), covar=tensor([0.0402, 0.2776, 0.2728, 0.1897, 0.0703, 0.2968, 0.1381, 0.0201], + device='cuda:3'), in_proj_covar=tensor([0.0373, 0.0347, 0.0364, 0.0328, 0.0357, 0.0338, 0.0349, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:41:04,997 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.031e+02 4.974e+02 6.176e+02 7.406e+02 2.018e+03, threshold=1.235e+03, percent-clipped=5.0 +2023-04-02 02:41:16,629 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9949, 1.5671, 1.5777, 1.9218, 1.5986, 1.7028, 1.5151, 1.7953], + device='cuda:3'), covar=tensor([0.0938, 0.1527, 0.1370, 0.0925, 0.1288, 0.0519, 0.1278, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0351, 0.0293, 0.0241, 0.0294, 0.0244, 0.0287, 0.0241], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:41:41,638 INFO [train.py:903] (3/4) Epoch 14, batch 3350, loss[loss=0.1954, simple_loss=0.2825, pruned_loss=0.05416, over 19647.00 frames. ], tot_loss[loss=0.2232, simple_loss=0.2996, pruned_loss=0.07345, over 3829584.07 frames. ], batch size: 55, lr: 5.94e-03, grad_scale: 8.0 +2023-04-02 02:42:40,620 INFO [train.py:903] (3/4) Epoch 14, batch 3400, loss[loss=0.2324, simple_loss=0.3099, pruned_loss=0.07746, over 19327.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2998, pruned_loss=0.07358, over 3833603.21 frames. ], batch size: 66, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:42:41,122 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5015, 1.4896, 1.7018, 1.7093, 2.4914, 2.3044, 2.5717, 1.0141], + device='cuda:3'), covar=tensor([0.2176, 0.3875, 0.2368, 0.1704, 0.1351, 0.1847, 0.1296, 0.3867], + device='cuda:3'), in_proj_covar=tensor([0.0502, 0.0589, 0.0634, 0.0448, 0.0598, 0.0500, 0.0645, 0.0506], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 02:43:05,849 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.972e+02 4.934e+02 6.017e+02 7.496e+02 1.650e+03, threshold=1.203e+03, percent-clipped=3.0 +2023-04-02 02:43:27,413 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-04-02 02:43:42,230 INFO [train.py:903] (3/4) Epoch 14, batch 3450, loss[loss=0.2109, simple_loss=0.2916, pruned_loss=0.06515, over 19703.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2987, pruned_loss=0.07267, over 3831406.20 frames. ], batch size: 59, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:43:44,673 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 02:44:28,720 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=92253.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:44:40,672 INFO [train.py:903] (3/4) Epoch 14, batch 3500, loss[loss=0.238, simple_loss=0.317, pruned_loss=0.07952, over 17351.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.3002, pruned_loss=0.07365, over 3825654.07 frames. ], batch size: 101, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:44:54,874 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92275.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:45:05,728 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.029e+02 5.239e+02 6.377e+02 7.871e+02 1.606e+03, threshold=1.275e+03, percent-clipped=5.0 +2023-04-02 02:45:24,194 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92300.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:45:37,117 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9402, 1.6526, 1.4625, 1.9147, 1.6506, 1.6826, 1.5395, 1.8433], + device='cuda:3'), covar=tensor([0.0973, 0.1412, 0.1523, 0.1030, 0.1228, 0.0517, 0.1243, 0.0732], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0353, 0.0297, 0.0244, 0.0298, 0.0246, 0.0290, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:45:41,111 INFO [train.py:903] (3/4) Epoch 14, batch 3550, loss[loss=0.2269, simple_loss=0.3066, pruned_loss=0.07358, over 19728.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3008, pruned_loss=0.07432, over 3812441.46 frames. ], batch size: 63, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:45:44,523 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=92317.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:46:15,212 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=92343.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 02:46:39,903 INFO [train.py:903] (3/4) Epoch 14, batch 3600, loss[loss=0.2341, simple_loss=0.316, pruned_loss=0.07611, over 19693.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.3018, pruned_loss=0.07497, over 3810930.92 frames. ], batch size: 59, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:46:44,980 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92368.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:47:04,639 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.193e+02 5.209e+02 6.321e+02 7.842e+02 1.520e+03, threshold=1.264e+03, percent-clipped=2.0 +2023-04-02 02:47:40,755 INFO [train.py:903] (3/4) Epoch 14, batch 3650, loss[loss=0.211, simple_loss=0.2996, pruned_loss=0.06122, over 19256.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.3012, pruned_loss=0.07497, over 3801781.33 frames. ], batch size: 66, lr: 5.93e-03, grad_scale: 8.0 +2023-04-02 02:48:03,008 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92432.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:48:34,087 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92458.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:48:40,398 INFO [train.py:903] (3/4) Epoch 14, batch 3700, loss[loss=0.2536, simple_loss=0.3249, pruned_loss=0.09119, over 19281.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.3015, pruned_loss=0.0751, over 3818444.67 frames. ], batch size: 66, lr: 5.92e-03, grad_scale: 8.0 +2023-04-02 02:49:05,816 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.433e+02 4.888e+02 6.023e+02 8.004e+02 1.682e+03, threshold=1.205e+03, percent-clipped=3.0 +2023-04-02 02:49:41,646 INFO [train.py:903] (3/4) Epoch 14, batch 3750, loss[loss=0.2322, simple_loss=0.3126, pruned_loss=0.07594, over 19548.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.3018, pruned_loss=0.07487, over 3816786.87 frames. ], batch size: 61, lr: 5.92e-03, grad_scale: 8.0 +2023-04-02 02:50:14,881 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9519, 4.4457, 2.7634, 3.9558, 1.0543, 4.3215, 4.3163, 4.4245], + device='cuda:3'), covar=tensor([0.0502, 0.0949, 0.1881, 0.0755, 0.3873, 0.0627, 0.0737, 0.0966], + device='cuda:3'), in_proj_covar=tensor([0.0447, 0.0375, 0.0448, 0.0327, 0.0388, 0.0386, 0.0375, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 02:50:42,053 INFO [train.py:903] (3/4) Epoch 14, batch 3800, loss[loss=0.1852, simple_loss=0.2531, pruned_loss=0.05861, over 19751.00 frames. ], tot_loss[loss=0.2245, simple_loss=0.3007, pruned_loss=0.07419, over 3828220.38 frames. ], batch size: 45, lr: 5.92e-03, grad_scale: 8.0 +2023-04-02 02:51:06,367 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.816e+02 4.992e+02 6.384e+02 8.353e+02 1.667e+03, threshold=1.277e+03, percent-clipped=5.0 +2023-04-02 02:51:11,003 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 02:51:15,768 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2888, 1.3387, 1.5818, 1.4707, 2.2850, 1.9350, 2.2876, 0.8637], + device='cuda:3'), covar=tensor([0.2440, 0.4001, 0.2442, 0.1881, 0.1399, 0.2199, 0.1333, 0.4069], + device='cuda:3'), in_proj_covar=tensor([0.0507, 0.0596, 0.0642, 0.0453, 0.0606, 0.0506, 0.0654, 0.0510], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 02:51:42,078 INFO [train.py:903] (3/4) Epoch 14, batch 3850, loss[loss=0.2389, simple_loss=0.3075, pruned_loss=0.08518, over 19383.00 frames. ], tot_loss[loss=0.224, simple_loss=0.3006, pruned_loss=0.07371, over 3835155.02 frames. ], batch size: 48, lr: 5.92e-03, grad_scale: 8.0 +2023-04-02 02:51:45,129 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-02 02:51:54,820 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92624.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:52:19,048 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92643.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:52:26,059 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92649.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:52:43,607 INFO [train.py:903] (3/4) Epoch 14, batch 3900, loss[loss=0.2254, simple_loss=0.3048, pruned_loss=0.07302, over 19672.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2991, pruned_loss=0.07338, over 3823810.17 frames. ], batch size: 58, lr: 5.92e-03, grad_scale: 4.0 +2023-04-02 02:53:10,396 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.910e+02 5.705e+02 7.277e+02 9.203e+02 2.913e+03, threshold=1.455e+03, percent-clipped=9.0 +2023-04-02 02:53:13,149 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92688.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:53:43,652 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92713.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:53:44,353 INFO [train.py:903] (3/4) Epoch 14, batch 3950, loss[loss=0.2523, simple_loss=0.3198, pruned_loss=0.09237, over 19774.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2987, pruned_loss=0.07294, over 3828485.80 frames. ], batch size: 56, lr: 5.92e-03, grad_scale: 4.0 +2023-04-02 02:53:44,806 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92714.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 02:53:48,544 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 02:54:12,952 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92738.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:54:14,290 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92739.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 02:54:45,454 INFO [train.py:903] (3/4) Epoch 14, batch 4000, loss[loss=0.2237, simple_loss=0.3046, pruned_loss=0.07141, over 19612.00 frames. ], tot_loss[loss=0.2225, simple_loss=0.2987, pruned_loss=0.07316, over 3811199.83 frames. ], batch size: 61, lr: 5.91e-03, grad_scale: 8.0 +2023-04-02 02:55:11,211 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.263e+02 4.983e+02 6.436e+02 8.255e+02 1.908e+03, threshold=1.287e+03, percent-clipped=2.0 +2023-04-02 02:55:33,048 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 02:55:45,574 INFO [train.py:903] (3/4) Epoch 14, batch 4050, loss[loss=0.2334, simple_loss=0.2909, pruned_loss=0.08797, over 19802.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2981, pruned_loss=0.07321, over 3820894.95 frames. ], batch size: 48, lr: 5.91e-03, grad_scale: 4.0 +2023-04-02 02:56:28,987 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92849.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:56:45,576 INFO [train.py:903] (3/4) Epoch 14, batch 4100, loss[loss=0.2259, simple_loss=0.2982, pruned_loss=0.07681, over 19535.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2978, pruned_loss=0.07322, over 3822568.02 frames. ], batch size: 54, lr: 5.91e-03, grad_scale: 4.0 +2023-04-02 02:57:13,888 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.303e+02 5.414e+02 6.604e+02 8.302e+02 1.654e+03, threshold=1.321e+03, percent-clipped=7.0 +2023-04-02 02:57:21,841 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 02:57:45,522 INFO [train.py:903] (3/4) Epoch 14, batch 4150, loss[loss=0.1877, simple_loss=0.2741, pruned_loss=0.05061, over 19661.00 frames. ], tot_loss[loss=0.2212, simple_loss=0.2971, pruned_loss=0.0727, over 3832557.69 frames. ], batch size: 53, lr: 5.91e-03, grad_scale: 4.0 +2023-04-02 02:57:51,061 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92917.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:58:28,818 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92949.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:58:48,463 INFO [train.py:903] (3/4) Epoch 14, batch 4200, loss[loss=0.2123, simple_loss=0.2939, pruned_loss=0.06532, over 19511.00 frames. ], tot_loss[loss=0.2213, simple_loss=0.2974, pruned_loss=0.07263, over 3822017.37 frames. ], batch size: 64, lr: 5.91e-03, grad_scale: 4.0 +2023-04-02 02:58:51,710 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 02:59:15,442 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.424e+02 5.036e+02 6.286e+02 7.807e+02 1.684e+03, threshold=1.257e+03, percent-clipped=4.0 +2023-04-02 02:59:15,622 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=92987.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:59:16,394 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.64 vs. limit=5.0 +2023-04-02 02:59:22,708 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92993.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 02:59:27,853 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6924, 4.1873, 4.4230, 4.3742, 1.6180, 4.1313, 3.6252, 4.1008], + device='cuda:3'), covar=tensor([0.1606, 0.0843, 0.0557, 0.0650, 0.5661, 0.0798, 0.0674, 0.1078], + device='cuda:3'), in_proj_covar=tensor([0.0711, 0.0643, 0.0847, 0.0732, 0.0762, 0.0597, 0.0514, 0.0780], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 02:59:48,126 INFO [train.py:903] (3/4) Epoch 14, batch 4250, loss[loss=0.2157, simple_loss=0.2843, pruned_loss=0.07359, over 19787.00 frames. ], tot_loss[loss=0.2218, simple_loss=0.2976, pruned_loss=0.07294, over 3818830.46 frames. ], batch size: 48, lr: 5.91e-03, grad_scale: 4.0 +2023-04-02 03:00:03,123 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 03:00:05,162 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93028.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:00:15,023 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 03:00:37,755 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93054.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:00:47,386 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.80 vs. limit=5.0 +2023-04-02 03:00:48,958 INFO [train.py:903] (3/4) Epoch 14, batch 4300, loss[loss=0.2384, simple_loss=0.3127, pruned_loss=0.0821, over 19398.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2979, pruned_loss=0.07317, over 3803689.90 frames. ], batch size: 70, lr: 5.90e-03, grad_scale: 4.0 +2023-04-02 03:01:12,696 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93082.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:01:18,348 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.279e+02 5.152e+02 6.308e+02 8.497e+02 2.668e+03, threshold=1.262e+03, percent-clipped=7.0 +2023-04-02 03:01:18,745 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9726, 1.9012, 1.6039, 2.1187, 2.0488, 1.6881, 1.6214, 1.9457], + device='cuda:3'), covar=tensor([0.0998, 0.1476, 0.1477, 0.0905, 0.1203, 0.0597, 0.1320, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0349, 0.0294, 0.0241, 0.0295, 0.0244, 0.0287, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:01:36,203 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93102.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:01:42,517 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 03:01:50,402 INFO [train.py:903] (3/4) Epoch 14, batch 4350, loss[loss=0.241, simple_loss=0.3188, pruned_loss=0.08157, over 19455.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2978, pruned_loss=0.07307, over 3812063.96 frames. ], batch size: 64, lr: 5.90e-03, grad_scale: 4.0 +2023-04-02 03:02:14,541 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93133.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:02:52,483 INFO [train.py:903] (3/4) Epoch 14, batch 4400, loss[loss=0.2491, simple_loss=0.3226, pruned_loss=0.08787, over 18235.00 frames. ], tot_loss[loss=0.2231, simple_loss=0.2989, pruned_loss=0.07365, over 3813038.23 frames. ], batch size: 83, lr: 5.90e-03, grad_scale: 8.0 +2023-04-02 03:03:15,289 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 03:03:18,721 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.086e+02 5.134e+02 6.154e+02 7.916e+02 2.681e+03, threshold=1.231e+03, percent-clipped=4.0 +2023-04-02 03:03:25,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 03:03:26,657 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93193.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:03:32,761 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93197.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:03:47,759 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3015, 3.0023, 2.2299, 2.2993, 2.0755, 2.5475, 0.8239, 2.1995], + device='cuda:3'), covar=tensor([0.0453, 0.0471, 0.0562, 0.0894, 0.0909, 0.0824, 0.1140, 0.0814], + device='cuda:3'), in_proj_covar=tensor([0.0341, 0.0337, 0.0333, 0.0362, 0.0437, 0.0362, 0.0315, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:03:52,928 INFO [train.py:903] (3/4) Epoch 14, batch 4450, loss[loss=0.2467, simple_loss=0.3244, pruned_loss=0.08454, over 19439.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2994, pruned_loss=0.07393, over 3827256.35 frames. ], batch size: 64, lr: 5.90e-03, grad_scale: 8.0 +2023-04-02 03:04:49,433 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93261.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:04:52,745 INFO [train.py:903] (3/4) Epoch 14, batch 4500, loss[loss=0.2335, simple_loss=0.3091, pruned_loss=0.07896, over 19683.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2998, pruned_loss=0.07375, over 3837337.51 frames. ], batch size: 60, lr: 5.90e-03, grad_scale: 8.0 +2023-04-02 03:05:21,652 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.340e+02 5.314e+02 6.344e+02 7.896e+02 1.749e+03, threshold=1.269e+03, percent-clipped=5.0 +2023-04-02 03:05:25,892 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-02 03:05:28,485 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93293.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:05:45,631 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93308.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:05:52,289 INFO [train.py:903] (3/4) Epoch 14, batch 4550, loss[loss=0.2197, simple_loss=0.2988, pruned_loss=0.07031, over 19522.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2997, pruned_loss=0.07379, over 3839638.84 frames. ], batch size: 54, lr: 5.90e-03, grad_scale: 8.0 +2023-04-02 03:06:06,164 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 03:06:21,908 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93337.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:06:27,656 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 03:06:39,765 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.37 vs. limit=5.0 +2023-04-02 03:06:47,781 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93358.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:06:55,967 INFO [train.py:903] (3/4) Epoch 14, batch 4600, loss[loss=0.1721, simple_loss=0.2499, pruned_loss=0.04718, over 16444.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2998, pruned_loss=0.07359, over 3822748.00 frames. ], batch size: 36, lr: 5.90e-03, grad_scale: 8.0 +2023-04-02 03:07:05,042 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93372.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:07:09,558 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93376.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:07:17,804 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93383.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:07:22,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.600e+02 5.339e+02 6.388e+02 8.227e+02 2.509e+03, threshold=1.278e+03, percent-clipped=2.0 +2023-04-02 03:07:35,575 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93398.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:07:49,478 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93408.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:07:56,022 INFO [train.py:903] (3/4) Epoch 14, batch 4650, loss[loss=0.2139, simple_loss=0.2751, pruned_loss=0.07635, over 19044.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.3, pruned_loss=0.07394, over 3822081.73 frames. ], batch size: 42, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:07:56,743 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-02 03:08:12,046 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 03:08:23,200 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 03:08:43,030 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93452.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:08:44,270 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93453.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:08:45,669 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.66 vs. limit=5.0 +2023-04-02 03:08:56,302 INFO [train.py:903] (3/4) Epoch 14, batch 4700, loss[loss=0.2103, simple_loss=0.2911, pruned_loss=0.06477, over 17181.00 frames. ], tot_loss[loss=0.2245, simple_loss=0.3003, pruned_loss=0.07435, over 3816342.19 frames. ], batch size: 101, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:09:11,839 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=93477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:09:14,188 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93478.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:09:20,324 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 03:09:25,719 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.143e+02 5.199e+02 6.337e+02 7.857e+02 1.524e+03, threshold=1.267e+03, percent-clipped=2.0 +2023-04-02 03:09:26,101 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93487.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:09:55,144 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93513.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:09:55,940 INFO [train.py:903] (3/4) Epoch 14, batch 4750, loss[loss=0.2248, simple_loss=0.2971, pruned_loss=0.07626, over 19361.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.3001, pruned_loss=0.07384, over 3824434.32 frames. ], batch size: 47, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:10:55,741 INFO [train.py:903] (3/4) Epoch 14, batch 4800, loss[loss=0.2452, simple_loss=0.3222, pruned_loss=0.0841, over 17315.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2994, pruned_loss=0.07332, over 3823495.91 frames. ], batch size: 101, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:10:56,177 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93564.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:11:18,655 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1632, 1.1870, 1.5639, 1.1806, 2.6901, 3.4754, 3.2266, 3.6883], + device='cuda:3'), covar=tensor([0.1666, 0.3672, 0.3240, 0.2279, 0.0537, 0.0162, 0.0211, 0.0220], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0301, 0.0330, 0.0254, 0.0222, 0.0164, 0.0207, 0.0216], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:11:22,951 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.531e+02 5.541e+02 6.642e+02 8.296e+02 2.320e+03, threshold=1.328e+03, percent-clipped=4.0 +2023-04-02 03:11:25,732 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93589.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:11:29,130 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93592.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:11:39,594 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 03:11:43,651 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7313, 3.2287, 3.2984, 3.2732, 1.3473, 3.1383, 2.7398, 3.0129], + device='cuda:3'), covar=tensor([0.1699, 0.1086, 0.0842, 0.0891, 0.5397, 0.1029, 0.0823, 0.1438], + device='cuda:3'), in_proj_covar=tensor([0.0710, 0.0644, 0.0851, 0.0729, 0.0757, 0.0594, 0.0510, 0.0776], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 03:11:57,048 INFO [train.py:903] (3/4) Epoch 14, batch 4850, loss[loss=0.2251, simple_loss=0.3024, pruned_loss=0.07387, over 19344.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2982, pruned_loss=0.07291, over 3832784.34 frames. ], batch size: 66, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:12:17,802 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93632.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:12:23,245 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 03:12:42,757 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 03:12:47,278 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 03:12:48,581 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 03:12:48,976 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93657.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:12:56,528 INFO [train.py:903] (3/4) Epoch 14, batch 4900, loss[loss=0.2504, simple_loss=0.3196, pruned_loss=0.09057, over 19651.00 frames. ], tot_loss[loss=0.222, simple_loss=0.298, pruned_loss=0.07297, over 3827367.53 frames. ], batch size: 60, lr: 5.89e-03, grad_scale: 8.0 +2023-04-02 03:12:56,907 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93664.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:12:57,692 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 03:13:18,102 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 03:13:25,557 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.007e+02 4.842e+02 5.818e+02 7.123e+02 1.786e+03, threshold=1.164e+03, percent-clipped=2.0 +2023-04-02 03:13:26,355 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 03:13:26,358 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.53 vs. limit=2.0 +2023-04-02 03:13:28,158 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93689.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:13:33,908 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7294, 1.7966, 2.0637, 1.9038, 3.1380, 2.6022, 3.3509, 1.7723], + device='cuda:3'), covar=tensor([0.2025, 0.3484, 0.2227, 0.1616, 0.1379, 0.1760, 0.1391, 0.3277], + device='cuda:3'), in_proj_covar=tensor([0.0501, 0.0592, 0.0638, 0.0452, 0.0602, 0.0504, 0.0649, 0.0508], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:13:49,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93708.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:13:55,992 INFO [train.py:903] (3/4) Epoch 14, batch 4950, loss[loss=0.2889, simple_loss=0.349, pruned_loss=0.1144, over 12852.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2985, pruned_loss=0.07296, over 3811564.82 frames. ], batch size: 136, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:14:16,295 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 03:14:21,123 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93733.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:14:32,402 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93743.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:14:36,690 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 03:14:58,213 INFO [train.py:903] (3/4) Epoch 14, batch 5000, loss[loss=0.2897, simple_loss=0.3546, pruned_loss=0.1124, over 17244.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2993, pruned_loss=0.07378, over 3795171.75 frames. ], batch size: 101, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:15:04,117 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93768.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:15:05,336 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93769.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:15:07,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 03:15:17,362 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 03:15:18,125 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.75 vs. limit=5.0 +2023-04-02 03:15:25,225 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 4.057e+02 5.347e+02 6.962e+02 9.103e+02 2.417e+03, threshold=1.392e+03, percent-clipped=9.0 +2023-04-02 03:15:26,635 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93788.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:15:33,740 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:15:56,434 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3357, 3.8056, 2.2522, 2.3376, 3.5768, 2.1291, 1.6462, 2.2523], + device='cuda:3'), covar=tensor([0.1097, 0.0448, 0.0832, 0.0695, 0.0374, 0.0927, 0.0799, 0.0633], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0303, 0.0324, 0.0246, 0.0236, 0.0323, 0.0288, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:15:59,313 INFO [train.py:903] (3/4) Epoch 14, batch 5050, loss[loss=0.2372, simple_loss=0.3148, pruned_loss=0.07974, over 19535.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.3002, pruned_loss=0.07403, over 3806776.32 frames. ], batch size: 54, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:16:35,024 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 03:16:40,897 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:16:59,033 INFO [train.py:903] (3/4) Epoch 14, batch 5100, loss[loss=0.1975, simple_loss=0.2705, pruned_loss=0.06222, over 15218.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2993, pruned_loss=0.0733, over 3813765.73 frames. ], batch size: 33, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:17:09,121 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93873.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:17:09,907 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 03:17:13,178 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 03:17:16,704 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 03:17:26,256 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.008e+02 4.904e+02 5.934e+02 7.645e+02 1.361e+03, threshold=1.187e+03, percent-clipped=0.0 +2023-04-02 03:17:56,877 INFO [train.py:903] (3/4) Epoch 14, batch 5150, loss[loss=0.1716, simple_loss=0.2478, pruned_loss=0.04773, over 19766.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2996, pruned_loss=0.07363, over 3816407.61 frames. ], batch size: 45, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:18:09,224 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 03:18:43,207 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 03:18:58,008 INFO [train.py:903] (3/4) Epoch 14, batch 5200, loss[loss=0.2958, simple_loss=0.356, pruned_loss=0.1178, over 19491.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.3, pruned_loss=0.07392, over 3811711.27 frames. ], batch size: 64, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:19:13,844 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 03:19:25,473 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.052e+02 5.268e+02 6.485e+02 8.631e+02 2.638e+03, threshold=1.297e+03, percent-clipped=7.0 +2023-04-02 03:19:39,265 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93999.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:19:57,305 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 03:19:59,389 INFO [train.py:903] (3/4) Epoch 14, batch 5250, loss[loss=0.2456, simple_loss=0.3233, pruned_loss=0.08396, over 19769.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2998, pruned_loss=0.07365, over 3820700.77 frames. ], batch size: 54, lr: 5.88e-03, grad_scale: 8.0 +2023-04-02 03:20:59,248 INFO [train.py:903] (3/4) Epoch 14, batch 5300, loss[loss=0.2454, simple_loss=0.3203, pruned_loss=0.08522, over 19741.00 frames. ], tot_loss[loss=0.2243, simple_loss=0.3007, pruned_loss=0.07395, over 3817124.81 frames. ], batch size: 63, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:21:16,449 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 03:21:27,964 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.389e+02 5.368e+02 7.020e+02 9.283e+02 2.840e+03, threshold=1.404e+03, percent-clipped=4.0 +2023-04-02 03:21:59,008 INFO [train.py:903] (3/4) Epoch 14, batch 5350, loss[loss=0.268, simple_loss=0.336, pruned_loss=0.1, over 19290.00 frames. ], tot_loss[loss=0.2246, simple_loss=0.3008, pruned_loss=0.07421, over 3824415.02 frames. ], batch size: 66, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:22:23,156 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=94132.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:22:34,758 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 03:23:01,401 INFO [train.py:903] (3/4) Epoch 14, batch 5400, loss[loss=0.2394, simple_loss=0.3049, pruned_loss=0.08692, over 19613.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.3015, pruned_loss=0.0747, over 3814486.59 frames. ], batch size: 50, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:23:29,187 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.364e+02 5.537e+02 7.248e+02 8.700e+02 2.021e+03, threshold=1.450e+03, percent-clipped=3.0 +2023-04-02 03:24:03,211 INFO [train.py:903] (3/4) Epoch 14, batch 5450, loss[loss=0.1963, simple_loss=0.2756, pruned_loss=0.05846, over 19416.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.3015, pruned_loss=0.07472, over 3804490.88 frames. ], batch size: 48, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:24:34,817 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94241.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:24:43,519 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94247.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:24:54,154 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4877, 1.6764, 2.0711, 1.7621, 3.1804, 2.7924, 3.4789, 1.4936], + device='cuda:3'), covar=tensor([0.2430, 0.4161, 0.2600, 0.1927, 0.1577, 0.1915, 0.1666, 0.4163], + device='cuda:3'), in_proj_covar=tensor([0.0502, 0.0591, 0.0639, 0.0450, 0.0604, 0.0505, 0.0649, 0.0510], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:25:02,869 INFO [train.py:903] (3/4) Epoch 14, batch 5500, loss[loss=0.2164, simple_loss=0.2983, pruned_loss=0.0672, over 19508.00 frames. ], tot_loss[loss=0.2244, simple_loss=0.3006, pruned_loss=0.07409, over 3818073.83 frames. ], batch size: 64, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:25:24,860 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 03:25:30,872 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.158e+02 4.805e+02 5.794e+02 7.462e+02 1.465e+03, threshold=1.159e+03, percent-clipped=1.0 +2023-04-02 03:25:46,411 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2153, 2.8902, 2.1954, 2.2332, 2.0394, 2.4404, 1.0204, 2.1310], + device='cuda:3'), covar=tensor([0.0525, 0.0530, 0.0580, 0.0915, 0.0929, 0.0935, 0.1027, 0.0820], + device='cuda:3'), in_proj_covar=tensor([0.0344, 0.0342, 0.0336, 0.0363, 0.0437, 0.0363, 0.0316, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:26:01,461 INFO [train.py:903] (3/4) Epoch 14, batch 5550, loss[loss=0.1854, simple_loss=0.2593, pruned_loss=0.05572, over 19764.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.2996, pruned_loss=0.07411, over 3813972.38 frames. ], batch size: 47, lr: 5.87e-03, grad_scale: 8.0 +2023-04-02 03:26:08,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 03:26:23,617 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.24 vs. limit=5.0 +2023-04-02 03:26:37,718 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=94343.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:26:53,090 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-02 03:26:57,901 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 03:27:01,450 INFO [train.py:903] (3/4) Epoch 14, batch 5600, loss[loss=0.2089, simple_loss=0.2825, pruned_loss=0.06763, over 19494.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2979, pruned_loss=0.07301, over 3827770.06 frames. ], batch size: 49, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:27:05,669 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94366.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:27:30,031 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.083e+02 5.188e+02 6.005e+02 7.911e+02 1.925e+03, threshold=1.201e+03, percent-clipped=8.0 +2023-04-02 03:28:03,376 INFO [train.py:903] (3/4) Epoch 14, batch 5650, loss[loss=0.2337, simple_loss=0.3041, pruned_loss=0.0816, over 19346.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2987, pruned_loss=0.07395, over 3818635.66 frames. ], batch size: 66, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:28:49,693 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 03:28:55,926 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94458.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:29:02,020 INFO [train.py:903] (3/4) Epoch 14, batch 5700, loss[loss=0.2826, simple_loss=0.3407, pruned_loss=0.1122, over 17581.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2991, pruned_loss=0.07397, over 3814513.53 frames. ], batch size: 101, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:29:29,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.310e+02 4.949e+02 6.008e+02 7.817e+02 2.884e+03, threshold=1.202e+03, percent-clipped=11.0 +2023-04-02 03:29:50,190 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94503.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:30:02,384 INFO [train.py:903] (3/4) Epoch 14, batch 5750, loss[loss=0.2671, simple_loss=0.3293, pruned_loss=0.1025, over 19363.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2987, pruned_loss=0.07358, over 3813519.12 frames. ], batch size: 70, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:30:04,715 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 03:30:09,051 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.58 vs. limit=5.0 +2023-04-02 03:30:11,543 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 03:30:17,756 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 03:30:21,236 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=94528.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:31:04,685 INFO [train.py:903] (3/4) Epoch 14, batch 5800, loss[loss=0.184, simple_loss=0.255, pruned_loss=0.05645, over 19313.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2974, pruned_loss=0.0728, over 3810387.63 frames. ], batch size: 44, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:31:30,529 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=94585.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:31:32,532 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.175e+02 5.430e+02 7.155e+02 9.192e+02 1.752e+03, threshold=1.431e+03, percent-clipped=10.0 +2023-04-02 03:32:06,963 INFO [train.py:903] (3/4) Epoch 14, batch 5850, loss[loss=0.2207, simple_loss=0.3013, pruned_loss=0.07009, over 19561.00 frames. ], tot_loss[loss=0.2213, simple_loss=0.2973, pruned_loss=0.07267, over 3810120.84 frames. ], batch size: 61, lr: 5.86e-03, grad_scale: 8.0 +2023-04-02 03:32:43,951 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.56 vs. limit=2.0 +2023-04-02 03:33:06,740 INFO [train.py:903] (3/4) Epoch 14, batch 5900, loss[loss=0.2049, simple_loss=0.2848, pruned_loss=0.06253, over 19379.00 frames. ], tot_loss[loss=0.2205, simple_loss=0.2965, pruned_loss=0.07227, over 3817322.59 frames. ], batch size: 48, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:33:07,936 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 03:33:27,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 03:33:32,178 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0713, 5.4725, 2.8336, 4.7123, 1.1637, 5.3938, 5.4053, 5.5371], + device='cuda:3'), covar=tensor([0.0450, 0.1023, 0.2232, 0.0742, 0.4159, 0.0575, 0.0694, 0.0997], + device='cuda:3'), in_proj_covar=tensor([0.0449, 0.0378, 0.0451, 0.0324, 0.0389, 0.0382, 0.0376, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:33:33,163 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.177e+02 5.122e+02 5.971e+02 8.409e+02 2.018e+03, threshold=1.194e+03, percent-clipped=4.0 +2023-04-02 03:33:50,039 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94700.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:34:01,325 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=94710.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:34:05,812 INFO [train.py:903] (3/4) Epoch 14, batch 5950, loss[loss=0.2289, simple_loss=0.3116, pruned_loss=0.0731, over 19668.00 frames. ], tot_loss[loss=0.2213, simple_loss=0.297, pruned_loss=0.07281, over 3830788.09 frames. ], batch size: 53, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:34:06,287 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94714.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:34:18,445 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3349, 3.0374, 2.2457, 2.7883, 0.8794, 2.9113, 2.8797, 2.9559], + device='cuda:3'), covar=tensor([0.1015, 0.1279, 0.1972, 0.0971, 0.3539, 0.0993, 0.1022, 0.1231], + device='cuda:3'), in_proj_covar=tensor([0.0447, 0.0374, 0.0449, 0.0321, 0.0385, 0.0379, 0.0373, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:34:37,293 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=94739.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:35:04,629 INFO [train.py:903] (3/4) Epoch 14, batch 6000, loss[loss=0.2792, simple_loss=0.3271, pruned_loss=0.1157, over 19389.00 frames. ], tot_loss[loss=0.2231, simple_loss=0.298, pruned_loss=0.07412, over 3826112.91 frames. ], batch size: 48, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:35:04,629 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 03:35:17,173 INFO [train.py:937] (3/4) Epoch 14, validation: loss=0.1744, simple_loss=0.2748, pruned_loss=0.03705, over 944034.00 frames. +2023-04-02 03:35:17,174 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 03:35:27,294 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-02 03:35:47,198 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.285e+02 5.018e+02 6.191e+02 7.483e+02 1.325e+03, threshold=1.238e+03, percent-clipped=2.0 +2023-04-02 03:36:17,849 INFO [train.py:903] (3/4) Epoch 14, batch 6050, loss[loss=0.2419, simple_loss=0.3148, pruned_loss=0.08455, over 19793.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2989, pruned_loss=0.07429, over 3820470.22 frames. ], batch size: 56, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:36:33,142 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94825.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:36:46,396 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0087, 5.0346, 5.7953, 5.7649, 1.8035, 5.4138, 4.6601, 5.4238], + device='cuda:3'), covar=tensor([0.1294, 0.0735, 0.0485, 0.0483, 0.5957, 0.0547, 0.0536, 0.1016], + device='cuda:3'), in_proj_covar=tensor([0.0712, 0.0641, 0.0850, 0.0726, 0.0756, 0.0595, 0.0510, 0.0783], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 03:37:02,993 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-04-02 03:37:20,908 INFO [train.py:903] (3/4) Epoch 14, batch 6100, loss[loss=0.1984, simple_loss=0.2764, pruned_loss=0.06023, over 19752.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2991, pruned_loss=0.07378, over 3829127.06 frames. ], batch size: 54, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:37:37,762 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1697, 1.3541, 1.7760, 1.4137, 2.7159, 3.5391, 3.2915, 3.7118], + device='cuda:3'), covar=tensor([0.1610, 0.3442, 0.3015, 0.2188, 0.0542, 0.0216, 0.0209, 0.0211], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0302, 0.0329, 0.0252, 0.0223, 0.0166, 0.0207, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:37:48,988 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.227e+02 5.260e+02 6.294e+02 8.137e+02 1.551e+03, threshold=1.259e+03, percent-clipped=3.0 +2023-04-02 03:38:21,670 INFO [train.py:903] (3/4) Epoch 14, batch 6150, loss[loss=0.2401, simple_loss=0.3207, pruned_loss=0.07973, over 19699.00 frames. ], tot_loss[loss=0.2225, simple_loss=0.2986, pruned_loss=0.07325, over 3840218.90 frames. ], batch size: 59, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:38:24,606 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7482, 1.4919, 1.4579, 1.8050, 1.4213, 1.5930, 1.4749, 1.6830], + device='cuda:3'), covar=tensor([0.0957, 0.1283, 0.1270, 0.0842, 0.1154, 0.0506, 0.1158, 0.0652], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0351, 0.0297, 0.0244, 0.0299, 0.0244, 0.0289, 0.0244], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:38:48,815 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 03:39:13,137 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94956.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:39:21,874 INFO [train.py:903] (3/4) Epoch 14, batch 6200, loss[loss=0.2086, simple_loss=0.2937, pruned_loss=0.06172, over 19647.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2989, pruned_loss=0.07346, over 3824402.98 frames. ], batch size: 55, lr: 5.85e-03, grad_scale: 8.0 +2023-04-02 03:39:44,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8519, 4.3537, 2.6638, 3.8387, 1.2253, 4.3105, 4.2236, 4.3097], + device='cuda:3'), covar=tensor([0.0574, 0.0931, 0.2027, 0.0881, 0.3787, 0.0624, 0.0788, 0.0995], + device='cuda:3'), in_proj_covar=tensor([0.0447, 0.0371, 0.0451, 0.0318, 0.0384, 0.0379, 0.0372, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:39:44,499 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=94981.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 03:39:51,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.111e+02 5.470e+02 6.385e+02 8.085e+02 2.296e+03, threshold=1.277e+03, percent-clipped=5.0 +2023-04-02 03:40:22,505 INFO [train.py:903] (3/4) Epoch 14, batch 6250, loss[loss=0.2069, simple_loss=0.2744, pruned_loss=0.06965, over 19318.00 frames. ], tot_loss[loss=0.2231, simple_loss=0.299, pruned_loss=0.07355, over 3821810.83 frames. ], batch size: 44, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:40:55,017 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 03:41:13,034 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0469, 1.1627, 1.6654, 1.1276, 2.6352, 3.6228, 3.3660, 3.8664], + device='cuda:3'), covar=tensor([0.1680, 0.3703, 0.3100, 0.2247, 0.0539, 0.0176, 0.0206, 0.0185], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0302, 0.0328, 0.0252, 0.0223, 0.0167, 0.0207, 0.0217], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:41:24,166 INFO [train.py:903] (3/4) Epoch 14, batch 6300, loss[loss=0.2157, simple_loss=0.2798, pruned_loss=0.07579, over 18677.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2998, pruned_loss=0.07385, over 3820486.80 frames. ], batch size: 41, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:41:44,526 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95081.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:41:51,892 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.537e+02 5.238e+02 6.215e+02 7.195e+02 1.642e+03, threshold=1.243e+03, percent-clipped=4.0 +2023-04-02 03:42:15,025 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95106.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:42:24,134 INFO [train.py:903] (3/4) Epoch 14, batch 6350, loss[loss=0.2834, simple_loss=0.3354, pruned_loss=0.1157, over 13144.00 frames. ], tot_loss[loss=0.2244, simple_loss=0.3001, pruned_loss=0.07442, over 3807364.65 frames. ], batch size: 136, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:42:34,691 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95123.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:43:23,508 INFO [train.py:903] (3/4) Epoch 14, batch 6400, loss[loss=0.2247, simple_loss=0.3044, pruned_loss=0.07253, over 19589.00 frames. ], tot_loss[loss=0.226, simple_loss=0.3014, pruned_loss=0.0753, over 3810850.80 frames. ], batch size: 61, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:43:31,461 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9371, 1.7986, 1.5704, 2.0723, 1.8534, 1.7135, 1.7056, 1.8879], + device='cuda:3'), covar=tensor([0.0937, 0.1477, 0.1435, 0.0947, 0.1253, 0.0515, 0.1135, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0352, 0.0295, 0.0244, 0.0298, 0.0243, 0.0287, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:43:52,823 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.332e+02 5.689e+02 7.116e+02 8.755e+02 2.889e+03, threshold=1.423e+03, percent-clipped=3.0 +2023-04-02 03:44:23,642 INFO [train.py:903] (3/4) Epoch 14, batch 6450, loss[loss=0.2408, simple_loss=0.3215, pruned_loss=0.08005, over 19673.00 frames. ], tot_loss[loss=0.2264, simple_loss=0.3019, pruned_loss=0.07547, over 3811067.18 frames. ], batch size: 58, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:45:09,474 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 03:45:25,843 INFO [train.py:903] (3/4) Epoch 14, batch 6500, loss[loss=0.2522, simple_loss=0.3323, pruned_loss=0.08607, over 19698.00 frames. ], tot_loss[loss=0.2248, simple_loss=0.3008, pruned_loss=0.07442, over 3830970.96 frames. ], batch size: 60, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:45:32,329 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 03:45:43,539 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95278.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:45:52,232 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-04-02 03:45:52,736 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95286.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:45:54,574 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.481e+02 5.245e+02 6.559e+02 8.783e+02 2.152e+03, threshold=1.312e+03, percent-clipped=6.0 +2023-04-02 03:46:27,877 INFO [train.py:903] (3/4) Epoch 14, batch 6550, loss[loss=0.187, simple_loss=0.2763, pruned_loss=0.04886, over 19613.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2992, pruned_loss=0.07326, over 3841310.44 frames. ], batch size: 57, lr: 5.84e-03, grad_scale: 8.0 +2023-04-02 03:47:20,335 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95357.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:47:28,138 INFO [train.py:903] (3/4) Epoch 14, batch 6600, loss[loss=0.1914, simple_loss=0.2637, pruned_loss=0.05958, over 19748.00 frames. ], tot_loss[loss=0.2218, simple_loss=0.2979, pruned_loss=0.07283, over 3814453.73 frames. ], batch size: 47, lr: 5.83e-03, grad_scale: 8.0 +2023-04-02 03:47:41,991 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9579, 4.3727, 4.6584, 4.6513, 1.8159, 4.3504, 3.8453, 4.3575], + device='cuda:3'), covar=tensor([0.1517, 0.0700, 0.0576, 0.0567, 0.5030, 0.0544, 0.0597, 0.1065], + device='cuda:3'), in_proj_covar=tensor([0.0726, 0.0648, 0.0863, 0.0740, 0.0765, 0.0603, 0.0517, 0.0793], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 03:47:53,875 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-02 03:47:57,395 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.401e+02 5.166e+02 6.061e+02 7.266e+02 1.890e+03, threshold=1.212e+03, percent-clipped=2.0 +2023-04-02 03:48:28,442 INFO [train.py:903] (3/4) Epoch 14, batch 6650, loss[loss=0.2162, simple_loss=0.2948, pruned_loss=0.06877, over 19358.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.298, pruned_loss=0.07293, over 3816563.05 frames. ], batch size: 47, lr: 5.83e-03, grad_scale: 8.0 +2023-04-02 03:48:52,536 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0085, 4.3604, 4.7098, 4.7003, 1.6233, 4.3719, 3.8587, 4.3338], + device='cuda:3'), covar=tensor([0.1351, 0.0867, 0.0532, 0.0526, 0.5701, 0.0670, 0.0599, 0.1141], + device='cuda:3'), in_proj_covar=tensor([0.0719, 0.0644, 0.0859, 0.0737, 0.0760, 0.0601, 0.0515, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 03:49:29,375 INFO [train.py:903] (3/4) Epoch 14, batch 6700, loss[loss=0.2674, simple_loss=0.3303, pruned_loss=0.1022, over 13001.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.2998, pruned_loss=0.07398, over 3807844.96 frames. ], batch size: 135, lr: 5.83e-03, grad_scale: 8.0 +2023-04-02 03:49:33,794 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=95467.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 03:49:38,883 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-02 03:49:50,129 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0977, 1.8557, 1.7211, 2.0460, 1.9552, 1.7183, 1.5813, 2.0285], + device='cuda:3'), covar=tensor([0.0904, 0.1309, 0.1307, 0.0898, 0.1160, 0.0516, 0.1271, 0.0617], + device='cuda:3'), in_proj_covar=tensor([0.0257, 0.0352, 0.0296, 0.0245, 0.0298, 0.0244, 0.0287, 0.0244], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:49:57,462 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.175e+02 5.267e+02 5.901e+02 8.158e+02 1.902e+03, threshold=1.180e+03, percent-clipped=6.0 +2023-04-02 03:50:01,513 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-02 03:50:12,944 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0527, 3.3701, 1.8714, 2.0272, 3.0336, 1.6198, 1.3426, 2.0769], + device='cuda:3'), covar=tensor([0.1270, 0.0595, 0.1001, 0.0762, 0.0512, 0.1153, 0.0947, 0.0704], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0304, 0.0325, 0.0246, 0.0238, 0.0326, 0.0291, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:50:25,737 INFO [train.py:903] (3/4) Epoch 14, batch 6750, loss[loss=0.262, simple_loss=0.3287, pruned_loss=0.09765, over 18779.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3005, pruned_loss=0.07445, over 3813932.63 frames. ], batch size: 74, lr: 5.83e-03, grad_scale: 8.0 +2023-04-02 03:51:01,613 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2587, 1.5459, 1.9679, 1.3924, 2.9534, 4.7580, 4.6790, 5.1672], + device='cuda:3'), covar=tensor([0.1547, 0.3281, 0.2870, 0.2018, 0.0499, 0.0150, 0.0150, 0.0115], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0303, 0.0329, 0.0253, 0.0223, 0.0167, 0.0207, 0.0219], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:51:21,229 INFO [train.py:903] (3/4) Epoch 14, batch 6800, loss[loss=0.2323, simple_loss=0.3071, pruned_loss=0.07877, over 17399.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.3011, pruned_loss=0.07528, over 3809632.69 frames. ], batch size: 101, lr: 5.83e-03, grad_scale: 8.0 +2023-04-02 03:51:41,543 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95582.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 03:51:46,738 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.931e+02 5.198e+02 6.166e+02 8.008e+02 1.508e+03, threshold=1.233e+03, percent-clipped=6.0 +2023-04-02 03:52:06,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 03:52:07,147 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 03:52:10,211 INFO [train.py:903] (3/4) Epoch 15, batch 0, loss[loss=0.2034, simple_loss=0.2754, pruned_loss=0.06575, over 19415.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2754, pruned_loss=0.06575, over 19415.00 frames. ], batch size: 48, lr: 5.63e-03, grad_scale: 8.0 +2023-04-02 03:52:10,211 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 03:52:21,739 INFO [train.py:937] (3/4) Epoch 15, validation: loss=0.1744, simple_loss=0.2751, pruned_loss=0.03681, over 944034.00 frames. +2023-04-02 03:52:21,741 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 03:52:28,903 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2508, 1.2563, 1.5633, 1.4793, 2.4988, 2.0973, 2.6740, 1.0716], + device='cuda:3'), covar=tensor([0.2447, 0.4132, 0.2458, 0.1892, 0.1490, 0.2026, 0.1424, 0.4025], + device='cuda:3'), in_proj_covar=tensor([0.0503, 0.0592, 0.0642, 0.0452, 0.0602, 0.0507, 0.0650, 0.0510], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 03:52:33,162 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 03:52:58,966 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=95622.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:53:08,239 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=95630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:53:14,653 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95635.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:53:22,140 INFO [train.py:903] (3/4) Epoch 15, batch 50, loss[loss=0.2512, simple_loss=0.3228, pruned_loss=0.08979, over 19335.00 frames. ], tot_loss[loss=0.2285, simple_loss=0.3041, pruned_loss=0.07643, over 861140.38 frames. ], batch size: 66, lr: 5.63e-03, grad_scale: 8.0 +2023-04-02 03:53:25,960 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8832, 1.5276, 1.4940, 1.4565, 3.4594, 0.9960, 2.3090, 3.8147], + device='cuda:3'), covar=tensor([0.0417, 0.2424, 0.2582, 0.1825, 0.0678, 0.2503, 0.1203, 0.0235], + device='cuda:3'), in_proj_covar=tensor([0.0376, 0.0349, 0.0369, 0.0328, 0.0358, 0.0337, 0.0346, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:53:58,798 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 03:54:20,263 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.439e+02 5.472e+02 6.461e+02 8.212e+02 1.912e+03, threshold=1.292e+03, percent-clipped=7.0 +2023-04-02 03:54:26,855 INFO [train.py:903] (3/4) Epoch 15, batch 100, loss[loss=0.2298, simple_loss=0.311, pruned_loss=0.07429, over 19755.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2992, pruned_loss=0.07237, over 1515497.49 frames. ], batch size: 63, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:54:29,739 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8797, 1.4724, 1.3987, 1.7872, 1.5678, 1.5504, 1.3722, 1.7257], + device='cuda:3'), covar=tensor([0.0939, 0.1386, 0.1546, 0.0911, 0.1196, 0.0552, 0.1399, 0.0731], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0355, 0.0299, 0.0246, 0.0299, 0.0247, 0.0289, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:54:37,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 03:54:37,597 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=95701.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:55:22,878 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95737.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:55:28,252 INFO [train.py:903] (3/4) Epoch 15, batch 150, loss[loss=0.2126, simple_loss=0.2972, pruned_loss=0.06399, over 19759.00 frames. ], tot_loss[loss=0.2207, simple_loss=0.2975, pruned_loss=0.07189, over 2037059.91 frames. ], batch size: 54, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:55:32,090 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95745.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:56:23,970 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.923e+02 5.351e+02 6.322e+02 7.452e+02 1.833e+03, threshold=1.264e+03, percent-clipped=1.0 +2023-04-02 03:56:27,313 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 03:56:28,499 INFO [train.py:903] (3/4) Epoch 15, batch 200, loss[loss=0.2031, simple_loss=0.2917, pruned_loss=0.05729, over 19617.00 frames. ], tot_loss[loss=0.2248, simple_loss=0.3011, pruned_loss=0.0742, over 2437789.87 frames. ], batch size: 57, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:56:51,115 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0520, 1.2963, 1.8302, 1.3837, 2.8101, 4.4270, 4.4103, 4.8945], + device='cuda:3'), covar=tensor([0.1639, 0.3575, 0.3076, 0.2128, 0.0604, 0.0188, 0.0151, 0.0129], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0302, 0.0328, 0.0252, 0.0223, 0.0166, 0.0205, 0.0218], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:56:59,737 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95816.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:57:15,609 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95830.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 03:57:24,900 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95838.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 03:57:29,641 INFO [train.py:903] (3/4) Epoch 15, batch 250, loss[loss=0.2449, simple_loss=0.3185, pruned_loss=0.08564, over 19659.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.3001, pruned_loss=0.07388, over 2750229.02 frames. ], batch size: 60, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:57:39,955 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5541, 4.1437, 2.4761, 3.6797, 1.1183, 3.9008, 3.8929, 3.9323], + device='cuda:3'), covar=tensor([0.0650, 0.0964, 0.2313, 0.0824, 0.4075, 0.0783, 0.0927, 0.1241], + device='cuda:3'), in_proj_covar=tensor([0.0454, 0.0377, 0.0457, 0.0322, 0.0391, 0.0387, 0.0381, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 03:57:56,159 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95863.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 03:57:58,221 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95865.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 03:58:24,514 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.450e+02 5.277e+02 6.948e+02 9.039e+02 3.101e+03, threshold=1.390e+03, percent-clipped=9.0 +2023-04-02 03:58:30,111 INFO [train.py:903] (3/4) Epoch 15, batch 300, loss[loss=0.2498, simple_loss=0.3226, pruned_loss=0.08846, over 19684.00 frames. ], tot_loss[loss=0.2259, simple_loss=0.3017, pruned_loss=0.07502, over 2995880.99 frames. ], batch size: 60, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:59:12,004 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3143, 1.4126, 1.7125, 1.5824, 2.6768, 2.2488, 2.6070, 1.0845], + device='cuda:3'), covar=tensor([0.2281, 0.4050, 0.2417, 0.1789, 0.1322, 0.1932, 0.1427, 0.3913], + device='cuda:3'), in_proj_covar=tensor([0.0498, 0.0587, 0.0636, 0.0448, 0.0598, 0.0503, 0.0643, 0.0506], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 03:59:32,836 INFO [train.py:903] (3/4) Epoch 15, batch 350, loss[loss=0.1821, simple_loss=0.2547, pruned_loss=0.05476, over 19724.00 frames. ], tot_loss[loss=0.2228, simple_loss=0.2991, pruned_loss=0.07323, over 3191067.85 frames. ], batch size: 46, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 03:59:33,870 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 04:00:17,460 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=95979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:00:24,009 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1560, 1.3062, 1.8154, 1.5598, 2.9078, 4.5108, 4.5861, 4.9930], + device='cuda:3'), covar=tensor([0.1679, 0.3541, 0.3118, 0.2037, 0.0536, 0.0182, 0.0145, 0.0153], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0304, 0.0332, 0.0254, 0.0224, 0.0168, 0.0207, 0.0220], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:00:28,194 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.384e+02 5.018e+02 5.906e+02 6.897e+02 1.495e+03, threshold=1.181e+03, percent-clipped=1.0 +2023-04-02 04:00:32,778 INFO [train.py:903] (3/4) Epoch 15, batch 400, loss[loss=0.2149, simple_loss=0.2876, pruned_loss=0.0711, over 19853.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.298, pruned_loss=0.07286, over 3335973.42 frames. ], batch size: 52, lr: 5.62e-03, grad_scale: 8.0 +2023-04-02 04:00:34,351 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95993.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:00:39,927 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95998.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:00:44,425 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96001.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:00:53,872 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9794, 1.0688, 1.5410, 0.7597, 2.2492, 2.9995, 2.7221, 3.2084], + device='cuda:3'), covar=tensor([0.1727, 0.4007, 0.3487, 0.2653, 0.0591, 0.0200, 0.0242, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0305, 0.0332, 0.0255, 0.0225, 0.0168, 0.0207, 0.0221], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:01:04,807 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96018.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:01:16,068 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96026.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:01:33,843 INFO [train.py:903] (3/4) Epoch 15, batch 450, loss[loss=0.2083, simple_loss=0.2907, pruned_loss=0.0629, over 19746.00 frames. ], tot_loss[loss=0.2216, simple_loss=0.2976, pruned_loss=0.07278, over 3449776.93 frames. ], batch size: 54, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:02:07,788 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 04:02:07,823 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 04:02:12,916 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96072.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:02:19,227 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 04:02:31,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.196e+02 4.837e+02 5.995e+02 7.453e+02 1.580e+03, threshold=1.199e+03, percent-clipped=6.0 +2023-04-02 04:02:36,670 INFO [train.py:903] (3/4) Epoch 15, batch 500, loss[loss=0.2177, simple_loss=0.295, pruned_loss=0.07022, over 19853.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2979, pruned_loss=0.07247, over 3530121.68 frames. ], batch size: 52, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:02:39,295 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:02:43,773 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96097.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:02:44,803 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96098.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:03:38,897 INFO [train.py:903] (3/4) Epoch 15, batch 550, loss[loss=0.2426, simple_loss=0.3189, pruned_loss=0.08319, over 18827.00 frames. ], tot_loss[loss=0.2204, simple_loss=0.2972, pruned_loss=0.07177, over 3613871.85 frames. ], batch size: 74, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:03:45,757 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7058, 1.4200, 1.3874, 1.7609, 1.4492, 1.4904, 1.4494, 1.6433], + device='cuda:3'), covar=tensor([0.1008, 0.1387, 0.1434, 0.0983, 0.1252, 0.0558, 0.1268, 0.0738], + device='cuda:3'), in_proj_covar=tensor([0.0255, 0.0349, 0.0295, 0.0243, 0.0295, 0.0244, 0.0285, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:04:18,076 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96174.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 04:04:29,243 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96183.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:04:35,630 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.160e+02 5.342e+02 6.491e+02 8.104e+02 1.503e+03, threshold=1.298e+03, percent-clipped=3.0 +2023-04-02 04:04:40,050 INFO [train.py:903] (3/4) Epoch 15, batch 600, loss[loss=0.1954, simple_loss=0.2776, pruned_loss=0.05655, over 19788.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2988, pruned_loss=0.0726, over 3667968.60 frames. ], batch size: 47, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:05:00,836 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96209.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:05:20,396 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 04:05:33,430 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 04:05:43,306 INFO [train.py:903] (3/4) Epoch 15, batch 650, loss[loss=0.1983, simple_loss=0.2744, pruned_loss=0.06107, over 19475.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2989, pruned_loss=0.07265, over 3707568.48 frames. ], batch size: 49, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:06:10,064 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1025, 3.5505, 2.0777, 2.1105, 3.2511, 1.8798, 1.2881, 2.2783], + device='cuda:3'), covar=tensor([0.1215, 0.0515, 0.0978, 0.0799, 0.0457, 0.1109, 0.1025, 0.0660], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0303, 0.0322, 0.0245, 0.0236, 0.0324, 0.0287, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:06:10,082 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2578, 3.0675, 2.2870, 2.3599, 2.1547, 2.5934, 0.7746, 2.1394], + device='cuda:3'), covar=tensor([0.0555, 0.0478, 0.0603, 0.0932, 0.0984, 0.0940, 0.1343, 0.0913], + device='cuda:3'), in_proj_covar=tensor([0.0344, 0.0340, 0.0335, 0.0364, 0.0438, 0.0365, 0.0319, 0.0324], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:06:20,840 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3218, 3.7784, 3.9019, 3.9133, 1.5360, 3.6671, 3.2691, 3.5967], + device='cuda:3'), covar=tensor([0.1505, 0.0828, 0.0674, 0.0702, 0.5366, 0.0781, 0.0647, 0.1233], + device='cuda:3'), in_proj_covar=tensor([0.0720, 0.0648, 0.0857, 0.0736, 0.0760, 0.0595, 0.0516, 0.0786], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:06:41,549 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.029e+02 5.105e+02 6.385e+02 8.770e+02 1.706e+03, threshold=1.277e+03, percent-clipped=3.0 +2023-04-02 04:06:42,946 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96289.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:06:46,046 INFO [train.py:903] (3/4) Epoch 15, batch 700, loss[loss=0.226, simple_loss=0.3022, pruned_loss=0.07494, over 18291.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2988, pruned_loss=0.07287, over 3731879.92 frames. ], batch size: 84, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:07:24,629 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.53 vs. limit=2.0 +2023-04-02 04:07:26,395 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96324.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:07:47,397 INFO [train.py:903] (3/4) Epoch 15, batch 750, loss[loss=0.2245, simple_loss=0.3052, pruned_loss=0.07189, over 19760.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.3001, pruned_loss=0.07379, over 3754437.47 frames. ], batch size: 63, lr: 5.61e-03, grad_scale: 8.0 +2023-04-02 04:07:47,565 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96342.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:07:47,728 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96342.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:07:58,659 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96350.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:08:29,483 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96375.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:08:44,438 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.337e+02 5.276e+02 6.207e+02 7.536e+02 1.572e+03, threshold=1.241e+03, percent-clipped=2.0 +2023-04-02 04:08:49,811 INFO [train.py:903] (3/4) Epoch 15, batch 800, loss[loss=0.1567, simple_loss=0.2325, pruned_loss=0.0404, over 19762.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2996, pruned_loss=0.07361, over 3780018.41 frames. ], batch size: 47, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:09:04,707 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 04:09:09,611 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96408.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:09:12,036 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9915, 2.7764, 1.9687, 2.0710, 1.8533, 2.2556, 0.9695, 1.8980], + device='cuda:3'), covar=tensor([0.0617, 0.0525, 0.0609, 0.0940, 0.0986, 0.0952, 0.1127, 0.0905], + device='cuda:3'), in_proj_covar=tensor([0.0345, 0.0342, 0.0336, 0.0366, 0.0440, 0.0367, 0.0321, 0.0326], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:09:50,619 INFO [train.py:903] (3/4) Epoch 15, batch 850, loss[loss=0.2914, simple_loss=0.3498, pruned_loss=0.1165, over 13173.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2998, pruned_loss=0.07382, over 3790883.73 frames. ], batch size: 136, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:09:51,925 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96442.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:09:53,261 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3236, 3.7848, 3.8945, 3.8913, 1.5254, 3.6685, 3.2505, 3.6130], + device='cuda:3'), covar=tensor([0.1445, 0.0894, 0.0635, 0.0682, 0.5251, 0.0811, 0.0637, 0.1074], + device='cuda:3'), in_proj_covar=tensor([0.0725, 0.0649, 0.0863, 0.0739, 0.0767, 0.0602, 0.0519, 0.0792], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:09:57,238 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.76 vs. limit=5.0 +2023-04-02 04:10:10,299 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96457.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:10:41,147 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 04:10:47,710 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.001e+02 5.238e+02 6.465e+02 7.879e+02 1.664e+03, threshold=1.293e+03, percent-clipped=4.0 +2023-04-02 04:10:52,477 INFO [train.py:903] (3/4) Epoch 15, batch 900, loss[loss=0.1746, simple_loss=0.2535, pruned_loss=0.04787, over 19795.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.3, pruned_loss=0.07409, over 3792951.71 frames. ], batch size: 47, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:11:15,482 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9701, 3.6197, 2.4594, 3.2175, 0.9139, 3.5361, 3.4368, 3.5315], + device='cuda:3'), covar=tensor([0.0844, 0.1263, 0.2080, 0.0935, 0.3924, 0.0768, 0.0914, 0.1088], + device='cuda:3'), in_proj_covar=tensor([0.0453, 0.0381, 0.0455, 0.0323, 0.0390, 0.0388, 0.0382, 0.0412], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:11:36,537 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96527.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:11:55,051 INFO [train.py:903] (3/4) Epoch 15, batch 950, loss[loss=0.1951, simple_loss=0.2816, pruned_loss=0.05428, over 19845.00 frames. ], tot_loss[loss=0.2236, simple_loss=0.2998, pruned_loss=0.0737, over 3807852.65 frames. ], batch size: 52, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:11:56,235 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 04:11:59,952 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96545.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:12:14,123 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96557.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:12:30,391 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96570.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:12:41,732 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96580.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:12:52,449 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.674e+02 5.225e+02 6.143e+02 7.839e+02 1.754e+03, threshold=1.229e+03, percent-clipped=1.0 +2023-04-02 04:12:57,220 INFO [train.py:903] (3/4) Epoch 15, batch 1000, loss[loss=0.2076, simple_loss=0.2883, pruned_loss=0.06341, over 19742.00 frames. ], tot_loss[loss=0.2236, simple_loss=0.2996, pruned_loss=0.07379, over 3820802.32 frames. ], batch size: 51, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:12:58,910 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7521, 1.8873, 2.1043, 2.5993, 1.7893, 2.4624, 2.2202, 1.9382], + device='cuda:3'), covar=tensor([0.4003, 0.3469, 0.1594, 0.1847, 0.3717, 0.1695, 0.4123, 0.3014], + device='cuda:3'), in_proj_covar=tensor([0.0826, 0.0865, 0.0666, 0.0896, 0.0812, 0.0747, 0.0802, 0.0731], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 04:13:13,527 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96605.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:13:51,310 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 04:13:59,278 INFO [train.py:903] (3/4) Epoch 15, batch 1050, loss[loss=0.2077, simple_loss=0.2771, pruned_loss=0.0692, over 19393.00 frames. ], tot_loss[loss=0.2248, simple_loss=0.3006, pruned_loss=0.07448, over 3817930.05 frames. ], batch size: 48, lr: 5.60e-03, grad_scale: 8.0 +2023-04-02 04:13:59,604 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96642.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:14:28,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9869, 3.6356, 2.4023, 3.2485, 0.8974, 3.5064, 3.4573, 3.4863], + device='cuda:3'), covar=tensor([0.0791, 0.1126, 0.2121, 0.0864, 0.3781, 0.0766, 0.0854, 0.1183], + device='cuda:3'), in_proj_covar=tensor([0.0453, 0.0380, 0.0456, 0.0323, 0.0389, 0.0388, 0.0381, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:14:31,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 04:14:53,799 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96686.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:14:57,022 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.831e+02 5.170e+02 6.505e+02 8.376e+02 1.590e+03, threshold=1.301e+03, percent-clipped=4.0 +2023-04-02 04:15:01,344 INFO [train.py:903] (3/4) Epoch 15, batch 1100, loss[loss=0.2017, simple_loss=0.268, pruned_loss=0.06768, over 19403.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2997, pruned_loss=0.07382, over 3817250.28 frames. ], batch size: 48, lr: 5.60e-03, grad_scale: 4.0 +2023-04-02 04:15:28,122 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96713.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:15:58,790 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96738.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:16:02,840 INFO [train.py:903] (3/4) Epoch 15, batch 1150, loss[loss=0.2509, simple_loss=0.3173, pruned_loss=0.09223, over 19638.00 frames. ], tot_loss[loss=0.2225, simple_loss=0.2988, pruned_loss=0.0731, over 3816600.80 frames. ], batch size: 55, lr: 5.59e-03, grad_scale: 4.0 +2023-04-02 04:16:08,669 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1428, 5.5528, 3.0391, 4.8207, 1.3545, 5.5976, 5.4748, 5.6558], + device='cuda:3'), covar=tensor([0.0401, 0.0764, 0.1865, 0.0677, 0.3448, 0.0561, 0.0659, 0.0947], + device='cuda:3'), in_proj_covar=tensor([0.0452, 0.0380, 0.0456, 0.0323, 0.0388, 0.0388, 0.0381, 0.0412], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:16:16,218 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=96752.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:16:55,637 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7971, 1.8555, 2.1493, 2.3592, 1.6534, 2.3142, 2.3107, 1.9784], + device='cuda:3'), covar=tensor([0.3738, 0.3378, 0.1651, 0.1888, 0.3423, 0.1685, 0.3993, 0.2973], + device='cuda:3'), in_proj_covar=tensor([0.0828, 0.0865, 0.0668, 0.0902, 0.0813, 0.0749, 0.0805, 0.0735], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 04:17:01,609 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.369e+02 5.122e+02 6.460e+02 8.216e+02 1.619e+03, threshold=1.292e+03, percent-clipped=5.0 +2023-04-02 04:17:06,203 INFO [train.py:903] (3/4) Epoch 15, batch 1200, loss[loss=0.2214, simple_loss=0.3053, pruned_loss=0.06881, over 19555.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2996, pruned_loss=0.07348, over 3819557.95 frames. ], batch size: 56, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:17:08,711 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:17:18,031 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96801.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:17:32,634 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96813.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:17:39,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 04:18:03,086 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96838.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:18:08,000 INFO [train.py:903] (3/4) Epoch 15, batch 1250, loss[loss=0.2155, simple_loss=0.2953, pruned_loss=0.0679, over 19333.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2998, pruned_loss=0.07379, over 3814521.16 frames. ], batch size: 70, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:18:23,404 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7245, 3.9327, 4.2730, 4.2540, 2.4390, 4.0204, 3.6591, 4.0266], + device='cuda:3'), covar=tensor([0.1187, 0.2102, 0.0526, 0.0586, 0.4018, 0.0912, 0.0522, 0.0830], + device='cuda:3'), in_proj_covar=tensor([0.0724, 0.0653, 0.0863, 0.0739, 0.0770, 0.0604, 0.0519, 0.0793], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:18:38,305 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96867.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:19:05,690 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.585e+02 5.442e+02 6.897e+02 8.528e+02 1.967e+03, threshold=1.379e+03, percent-clipped=7.0 +2023-04-02 04:19:09,075 INFO [train.py:903] (3/4) Epoch 15, batch 1300, loss[loss=0.2269, simple_loss=0.3048, pruned_loss=0.07454, over 19616.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2993, pruned_loss=0.07368, over 3818921.36 frames. ], batch size: 57, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:19:17,641 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96898.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:19:48,652 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96923.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:19:57,838 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9354, 2.0082, 2.2679, 2.6617, 1.9247, 2.6628, 2.3112, 2.0629], + device='cuda:3'), covar=tensor([0.3930, 0.3479, 0.1695, 0.2198, 0.3762, 0.1783, 0.4150, 0.2992], + device='cuda:3'), in_proj_covar=tensor([0.0831, 0.0865, 0.0669, 0.0902, 0.0814, 0.0750, 0.0807, 0.0735], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 04:20:12,477 INFO [train.py:903] (3/4) Epoch 15, batch 1350, loss[loss=0.216, simple_loss=0.297, pruned_loss=0.06749, over 19655.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2994, pruned_loss=0.07375, over 3829359.33 frames. ], batch size: 60, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:21:11,447 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.829e+02 4.520e+02 5.515e+02 7.272e+02 1.782e+03, threshold=1.103e+03, percent-clipped=1.0 +2023-04-02 04:21:15,867 INFO [train.py:903] (3/4) Epoch 15, batch 1400, loss[loss=0.2233, simple_loss=0.3036, pruned_loss=0.07144, over 19675.00 frames. ], tot_loss[loss=0.221, simple_loss=0.2974, pruned_loss=0.07231, over 3833401.41 frames. ], batch size: 60, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:21:49,713 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97018.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:22:19,081 INFO [train.py:903] (3/4) Epoch 15, batch 1450, loss[loss=0.2217, simple_loss=0.3079, pruned_loss=0.06777, over 19535.00 frames. ], tot_loss[loss=0.2211, simple_loss=0.2973, pruned_loss=0.07246, over 3824349.65 frames. ], batch size: 54, lr: 5.59e-03, grad_scale: 8.0 +2023-04-02 04:22:20,293 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 04:22:38,951 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97057.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:23:09,851 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97082.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:23:18,577 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.200e+02 5.198e+02 6.417e+02 9.294e+02 1.968e+03, threshold=1.283e+03, percent-clipped=11.0 +2023-04-02 04:23:21,917 INFO [train.py:903] (3/4) Epoch 15, batch 1500, loss[loss=0.1886, simple_loss=0.2633, pruned_loss=0.05701, over 19777.00 frames. ], tot_loss[loss=0.221, simple_loss=0.2974, pruned_loss=0.07235, over 3827275.90 frames. ], batch size: 48, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:23:22,383 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1156, 1.7856, 1.6100, 2.0198, 1.7369, 1.8093, 1.5880, 2.0517], + device='cuda:3'), covar=tensor([0.0872, 0.1403, 0.1421, 0.0969, 0.1346, 0.0493, 0.1307, 0.0663], + device='cuda:3'), in_proj_covar=tensor([0.0256, 0.0349, 0.0297, 0.0242, 0.0293, 0.0245, 0.0287, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:23:46,474 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.40 vs. limit=5.0 +2023-04-02 04:24:01,311 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97123.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:24:20,475 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97138.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:24:24,799 INFO [train.py:903] (3/4) Epoch 15, batch 1550, loss[loss=0.186, simple_loss=0.2593, pruned_loss=0.05634, over 19385.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2971, pruned_loss=0.07179, over 3834089.66 frames. ], batch size: 48, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:24:31,868 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97148.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:25:22,551 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.982e+02 5.223e+02 6.490e+02 8.468e+02 1.572e+03, threshold=1.298e+03, percent-clipped=7.0 +2023-04-02 04:25:25,973 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1128, 1.2217, 1.4594, 1.3615, 2.6805, 1.0883, 2.1112, 2.9693], + device='cuda:3'), covar=tensor([0.0525, 0.2777, 0.2625, 0.1722, 0.0764, 0.2248, 0.1106, 0.0358], + device='cuda:3'), in_proj_covar=tensor([0.0373, 0.0346, 0.0363, 0.0326, 0.0351, 0.0335, 0.0346, 0.0371], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:25:26,785 INFO [train.py:903] (3/4) Epoch 15, batch 1600, loss[loss=0.214, simple_loss=0.2882, pruned_loss=0.06995, over 19733.00 frames. ], tot_loss[loss=0.2211, simple_loss=0.2975, pruned_loss=0.07233, over 3826015.16 frames. ], batch size: 51, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:25:40,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3931, 1.4257, 1.7242, 1.6368, 2.6653, 2.2746, 2.8105, 1.0449], + device='cuda:3'), covar=tensor([0.2268, 0.3984, 0.2580, 0.1784, 0.1492, 0.1963, 0.1422, 0.4109], + device='cuda:3'), in_proj_covar=tensor([0.0507, 0.0592, 0.0644, 0.0451, 0.0605, 0.0506, 0.0649, 0.0509], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:25:51,567 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 04:26:28,830 INFO [train.py:903] (3/4) Epoch 15, batch 1650, loss[loss=0.2608, simple_loss=0.3415, pruned_loss=0.09005, over 19327.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2984, pruned_loss=0.07302, over 3817792.29 frames. ], batch size: 70, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:26:40,205 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0463, 1.5822, 1.8770, 1.6360, 4.5602, 0.8191, 2.6138, 4.8874], + device='cuda:3'), covar=tensor([0.0345, 0.2760, 0.2587, 0.1932, 0.0715, 0.2886, 0.1297, 0.0179], + device='cuda:3'), in_proj_covar=tensor([0.0374, 0.0347, 0.0365, 0.0327, 0.0353, 0.0336, 0.0348, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:26:42,671 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97253.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:27:14,827 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97278.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:27:27,487 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.574e+02 5.168e+02 6.563e+02 8.938e+02 1.305e+03, threshold=1.313e+03, percent-clipped=1.0 +2023-04-02 04:27:30,919 INFO [train.py:903] (3/4) Epoch 15, batch 1700, loss[loss=0.2435, simple_loss=0.317, pruned_loss=0.08502, over 19404.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2991, pruned_loss=0.07341, over 3816439.56 frames. ], batch size: 70, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:27:36,039 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4991, 4.0628, 4.2149, 4.2029, 1.6544, 3.9605, 3.4911, 3.9403], + device='cuda:3'), covar=tensor([0.1400, 0.0770, 0.0535, 0.0590, 0.5148, 0.0811, 0.0622, 0.0933], + device='cuda:3'), in_proj_covar=tensor([0.0717, 0.0647, 0.0855, 0.0735, 0.0763, 0.0599, 0.0516, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:28:11,291 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 04:28:32,694 INFO [train.py:903] (3/4) Epoch 15, batch 1750, loss[loss=0.217, simple_loss=0.2947, pruned_loss=0.0697, over 19585.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.2994, pruned_loss=0.07386, over 3813691.59 frames. ], batch size: 61, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:28:55,664 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97360.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:28:57,858 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97362.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:29:01,665 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97365.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:29:30,674 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.259e+02 5.096e+02 6.393e+02 8.037e+02 1.428e+03, threshold=1.279e+03, percent-clipped=5.0 +2023-04-02 04:29:33,928 INFO [train.py:903] (3/4) Epoch 15, batch 1800, loss[loss=0.2384, simple_loss=0.3203, pruned_loss=0.07823, over 19531.00 frames. ], tot_loss[loss=0.2232, simple_loss=0.2991, pruned_loss=0.07368, over 3821244.25 frames. ], batch size: 54, lr: 5.58e-03, grad_scale: 8.0 +2023-04-02 04:29:34,202 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97392.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:30:32,231 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 04:30:36,840 INFO [train.py:903] (3/4) Epoch 15, batch 1850, loss[loss=0.2529, simple_loss=0.3211, pruned_loss=0.09238, over 19312.00 frames. ], tot_loss[loss=0.2227, simple_loss=0.2988, pruned_loss=0.07325, over 3811261.28 frames. ], batch size: 66, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:31:10,809 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 04:31:21,359 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:31:35,888 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.266e+02 4.535e+02 5.674e+02 7.772e+02 1.771e+03, threshold=1.135e+03, percent-clipped=3.0 +2023-04-02 04:31:39,238 INFO [train.py:903] (3/4) Epoch 15, batch 1900, loss[loss=0.2315, simple_loss=0.3117, pruned_loss=0.07569, over 19651.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2982, pruned_loss=0.07295, over 3806543.71 frames. ], batch size: 60, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:31:58,009 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 04:32:00,970 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97509.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:32:02,865 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 04:32:28,308 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 04:32:32,288 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97534.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:32:40,694 INFO [train.py:903] (3/4) Epoch 15, batch 1950, loss[loss=0.2401, simple_loss=0.3179, pruned_loss=0.08114, over 18320.00 frames. ], tot_loss[loss=0.2224, simple_loss=0.2988, pruned_loss=0.07298, over 3822382.64 frames. ], batch size: 84, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:33:19,864 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97573.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:33:28,685 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-02 04:33:39,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.621e+02 4.980e+02 6.456e+02 8.636e+02 2.349e+03, threshold=1.291e+03, percent-clipped=8.0 +2023-04-02 04:33:43,291 INFO [train.py:903] (3/4) Epoch 15, batch 2000, loss[loss=0.2529, simple_loss=0.3138, pruned_loss=0.09598, over 14019.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2999, pruned_loss=0.07381, over 3803767.34 frames. ], batch size: 136, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:34:03,603 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.65 vs. limit=5.0 +2023-04-02 04:34:20,860 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97622.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:34:42,256 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 04:34:47,042 INFO [train.py:903] (3/4) Epoch 15, batch 2050, loss[loss=0.2222, simple_loss=0.3073, pruned_loss=0.06851, over 19291.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.3, pruned_loss=0.07379, over 3803738.57 frames. ], batch size: 66, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:35:01,890 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 04:35:03,049 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 04:35:23,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 04:35:47,011 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.214e+02 4.998e+02 6.604e+02 7.928e+02 1.987e+03, threshold=1.321e+03, percent-clipped=5.0 +2023-04-02 04:35:50,611 INFO [train.py:903] (3/4) Epoch 15, batch 2100, loss[loss=0.2065, simple_loss=0.2856, pruned_loss=0.06366, over 19591.00 frames. ], tot_loss[loss=0.2247, simple_loss=0.3008, pruned_loss=0.07427, over 3786007.23 frames. ], batch size: 52, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:35:50,932 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97692.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:36:04,943 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97704.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:36:10,965 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-04-02 04:36:11,598 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97709.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:36:20,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 04:36:41,918 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1020, 1.7877, 1.7024, 2.1388, 1.9086, 1.8059, 1.5598, 2.0327], + device='cuda:3'), covar=tensor([0.0875, 0.1560, 0.1371, 0.0970, 0.1209, 0.0503, 0.1331, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0353, 0.0298, 0.0245, 0.0296, 0.0247, 0.0290, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:36:41,957 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97733.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:36:43,967 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 04:36:45,262 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97736.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:36:46,629 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97737.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:36:51,867 INFO [train.py:903] (3/4) Epoch 15, batch 2150, loss[loss=0.1996, simple_loss=0.2799, pruned_loss=0.05958, over 19577.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.3003, pruned_loss=0.07393, over 3784033.27 frames. ], batch size: 52, lr: 5.57e-03, grad_scale: 8.0 +2023-04-02 04:36:53,795 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-04-02 04:37:12,302 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97758.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:37:49,744 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.357e+02 4.862e+02 6.186e+02 7.185e+02 1.323e+03, threshold=1.237e+03, percent-clipped=1.0 +2023-04-02 04:37:53,998 INFO [train.py:903] (3/4) Epoch 15, batch 2200, loss[loss=0.2107, simple_loss=0.2945, pruned_loss=0.06347, over 19604.00 frames. ], tot_loss[loss=0.2231, simple_loss=0.2995, pruned_loss=0.07338, over 3784915.14 frames. ], batch size: 57, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:38:28,072 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97819.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:38:33,718 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97824.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:38:36,554 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 04:38:57,676 INFO [train.py:903] (3/4) Epoch 15, batch 2250, loss[loss=0.2944, simple_loss=0.3491, pruned_loss=0.1198, over 19351.00 frames. ], tot_loss[loss=0.2235, simple_loss=0.3, pruned_loss=0.07354, over 3791539.85 frames. ], batch size: 70, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:39:09,183 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97851.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:39:09,869 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.19 vs. limit=5.0 +2023-04-02 04:39:56,865 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.382e+02 4.977e+02 6.292e+02 8.077e+02 1.831e+03, threshold=1.258e+03, percent-clipped=5.0 +2023-04-02 04:40:00,327 INFO [train.py:903] (3/4) Epoch 15, batch 2300, loss[loss=0.2286, simple_loss=0.2846, pruned_loss=0.0863, over 16121.00 frames. ], tot_loss[loss=0.2227, simple_loss=0.2992, pruned_loss=0.07309, over 3790333.43 frames. ], batch size: 35, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:40:12,679 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 04:40:16,282 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97906.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:40:21,313 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.41 vs. limit=5.0 +2023-04-02 04:40:29,883 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=97917.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:41:01,537 INFO [train.py:903] (3/4) Epoch 15, batch 2350, loss[loss=0.2431, simple_loss=0.3137, pruned_loss=0.08624, over 19297.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2999, pruned_loss=0.07339, over 3800921.00 frames. ], batch size: 66, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:41:11,490 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-04-02 04:41:44,921 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 04:41:58,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.032e+02 5.416e+02 6.244e+02 7.823e+02 1.587e+03, threshold=1.249e+03, percent-clipped=4.0 +2023-04-02 04:41:58,341 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 04:42:02,751 INFO [train.py:903] (3/4) Epoch 15, batch 2400, loss[loss=0.2004, simple_loss=0.2739, pruned_loss=0.06346, over 19851.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2985, pruned_loss=0.07286, over 3799819.59 frames. ], batch size: 52, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:42:04,441 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97993.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:42:36,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98018.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:42:53,733 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98032.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:42:58,098 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=98036.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:43:05,611 INFO [train.py:903] (3/4) Epoch 15, batch 2450, loss[loss=0.2086, simple_loss=0.286, pruned_loss=0.06559, over 19771.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2987, pruned_loss=0.07299, over 3798889.85 frames. ], batch size: 54, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:43:47,589 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7509, 1.6125, 1.5883, 2.1867, 1.7419, 2.1746, 2.1262, 1.9910], + device='cuda:3'), covar=tensor([0.0819, 0.0919, 0.0993, 0.0839, 0.0824, 0.0622, 0.0815, 0.0602], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0220, 0.0222, 0.0241, 0.0226, 0.0206, 0.0190, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 04:43:47,647 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98075.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:43:55,097 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98080.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:44:05,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.133e+02 5.305e+02 6.204e+02 8.091e+02 1.870e+03, threshold=1.241e+03, percent-clipped=5.0 +2023-04-02 04:44:09,724 INFO [train.py:903] (3/4) Epoch 15, batch 2500, loss[loss=0.2267, simple_loss=0.3046, pruned_loss=0.07435, over 19562.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2968, pruned_loss=0.07187, over 3813044.60 frames. ], batch size: 61, lr: 5.56e-03, grad_scale: 8.0 +2023-04-02 04:44:19,650 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98100.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:44:25,576 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98105.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:44:27,987 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98107.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:44:51,252 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98124.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:45:01,484 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98132.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:45:01,673 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 04:45:02,547 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1991, 3.7198, 3.8497, 3.8427, 1.4744, 3.6762, 3.1642, 3.5613], + device='cuda:3'), covar=tensor([0.1720, 0.0935, 0.0715, 0.0774, 0.5558, 0.0890, 0.0769, 0.1204], + device='cuda:3'), in_proj_covar=tensor([0.0724, 0.0654, 0.0862, 0.0742, 0.0770, 0.0608, 0.0520, 0.0787], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:45:12,195 INFO [train.py:903] (3/4) Epoch 15, batch 2550, loss[loss=0.2943, simple_loss=0.3605, pruned_loss=0.114, over 13683.00 frames. ], tot_loss[loss=0.2218, simple_loss=0.2982, pruned_loss=0.07266, over 3809547.45 frames. ], batch size: 136, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:45:23,247 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98151.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 04:46:07,096 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 04:46:10,535 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.981e+02 5.247e+02 6.339e+02 8.638e+02 2.352e+03, threshold=1.268e+03, percent-clipped=5.0 +2023-04-02 04:46:14,040 INFO [train.py:903] (3/4) Epoch 15, batch 2600, loss[loss=0.2256, simple_loss=0.3012, pruned_loss=0.07503, over 19589.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2977, pruned_loss=0.0726, over 3796551.98 frames. ], batch size: 52, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:47:18,023 INFO [train.py:903] (3/4) Epoch 15, batch 2650, loss[loss=0.1967, simple_loss=0.2715, pruned_loss=0.06091, over 19744.00 frames. ], tot_loss[loss=0.22, simple_loss=0.2965, pruned_loss=0.0718, over 3810787.26 frames. ], batch size: 45, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:47:28,823 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=98250.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:47:39,872 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 04:48:17,333 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98288.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:48:18,024 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.302e+02 4.990e+02 6.118e+02 7.575e+02 1.335e+03, threshold=1.224e+03, percent-clipped=1.0 +2023-04-02 04:48:21,600 INFO [train.py:903] (3/4) Epoch 15, batch 2700, loss[loss=0.2249, simple_loss=0.2996, pruned_loss=0.07513, over 19607.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2978, pruned_loss=0.07255, over 3799809.12 frames. ], batch size: 61, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:48:40,275 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6943, 1.6075, 1.5526, 2.2326, 1.8505, 1.9902, 2.0474, 1.8177], + device='cuda:3'), covar=tensor([0.0759, 0.0846, 0.0954, 0.0658, 0.0777, 0.0695, 0.0792, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0221, 0.0224, 0.0242, 0.0229, 0.0208, 0.0190, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 04:48:47,480 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98313.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:49:24,161 INFO [train.py:903] (3/4) Epoch 15, batch 2750, loss[loss=0.1937, simple_loss=0.2649, pruned_loss=0.06129, over 19755.00 frames. ], tot_loss[loss=0.2211, simple_loss=0.2974, pruned_loss=0.07245, over 3805771.12 frames. ], batch size: 46, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:49:39,888 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 04:49:54,726 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98365.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:50:23,801 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.213e+02 5.173e+02 6.107e+02 7.991e+02 1.431e+03, threshold=1.221e+03, percent-clipped=3.0 +2023-04-02 04:50:27,281 INFO [train.py:903] (3/4) Epoch 15, batch 2800, loss[loss=0.2289, simple_loss=0.3092, pruned_loss=0.0743, over 19097.00 frames. ], tot_loss[loss=0.2204, simple_loss=0.2969, pruned_loss=0.07192, over 3816115.41 frames. ], batch size: 69, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:50:34,375 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-04-02 04:50:48,806 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98407.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:51:13,908 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98428.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:51:18,547 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98432.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 04:51:30,819 INFO [train.py:903] (3/4) Epoch 15, batch 2850, loss[loss=0.1926, simple_loss=0.2659, pruned_loss=0.0597, over 19741.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2982, pruned_loss=0.07242, over 3804245.13 frames. ], batch size: 46, lr: 5.55e-03, grad_scale: 8.0 +2023-04-02 04:51:53,692 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7433, 3.2133, 3.2550, 3.3034, 1.3642, 3.1490, 2.7198, 2.9989], + device='cuda:3'), covar=tensor([0.1713, 0.1115, 0.0866, 0.0916, 0.5196, 0.0974, 0.0874, 0.1449], + device='cuda:3'), in_proj_covar=tensor([0.0726, 0.0658, 0.0862, 0.0741, 0.0773, 0.0611, 0.0519, 0.0799], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 04:52:03,936 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=98468.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:52:30,664 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.993e+02 5.164e+02 6.295e+02 8.927e+02 2.262e+03, threshold=1.259e+03, percent-clipped=4.0 +2023-04-02 04:52:34,166 INFO [train.py:903] (3/4) Epoch 15, batch 2900, loss[loss=0.1928, simple_loss=0.2644, pruned_loss=0.06063, over 19740.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2982, pruned_loss=0.0724, over 3821138.13 frames. ], batch size: 46, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:52:35,432 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 04:52:49,640 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4776, 1.3430, 1.3973, 1.8200, 1.4904, 1.6775, 1.7722, 1.5665], + device='cuda:3'), covar=tensor([0.0884, 0.1022, 0.1086, 0.0720, 0.0842, 0.0801, 0.0811, 0.0734], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0224, 0.0243, 0.0230, 0.0209, 0.0191, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 04:53:14,485 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8024, 4.4435, 3.2013, 3.7470, 1.9900, 4.2236, 4.2009, 4.3049], + device='cuda:3'), covar=tensor([0.0524, 0.1037, 0.1788, 0.0988, 0.2844, 0.0750, 0.0828, 0.1150], + device='cuda:3'), in_proj_covar=tensor([0.0463, 0.0384, 0.0460, 0.0326, 0.0389, 0.0394, 0.0385, 0.0418], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:53:36,705 INFO [train.py:903] (3/4) Epoch 15, batch 2950, loss[loss=0.2214, simple_loss=0.3049, pruned_loss=0.06891, over 19537.00 frames. ], tot_loss[loss=0.2228, simple_loss=0.2994, pruned_loss=0.07313, over 3817777.78 frames. ], batch size: 56, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:54:00,570 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98561.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:54:29,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:54:35,377 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.243e+02 5.043e+02 6.244e+02 8.249e+02 2.456e+03, threshold=1.249e+03, percent-clipped=2.0 +2023-04-02 04:54:38,823 INFO [train.py:903] (3/4) Epoch 15, batch 3000, loss[loss=0.2161, simple_loss=0.2916, pruned_loss=0.0703, over 19575.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2994, pruned_loss=0.07331, over 3821187.50 frames. ], batch size: 61, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:54:38,824 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 04:54:51,336 INFO [train.py:937] (3/4) Epoch 15, validation: loss=0.1735, simple_loss=0.2738, pruned_loss=0.0366, over 944034.00 frames. +2023-04-02 04:54:51,337 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 04:54:53,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 04:55:08,335 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4590, 2.2367, 1.6981, 1.3932, 2.0237, 1.3264, 1.3907, 1.9138], + device='cuda:3'), covar=tensor([0.0851, 0.0736, 0.0957, 0.0797, 0.0500, 0.1203, 0.0616, 0.0418], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0305, 0.0325, 0.0249, 0.0238, 0.0329, 0.0292, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 04:55:28,665 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98621.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:55:52,974 INFO [train.py:903] (3/4) Epoch 15, batch 3050, loss[loss=0.1932, simple_loss=0.2803, pruned_loss=0.05305, over 19655.00 frames. ], tot_loss[loss=0.223, simple_loss=0.2993, pruned_loss=0.07335, over 3826134.81 frames. ], batch size: 55, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:55:57,857 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98646.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:56:51,964 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.865e+02 5.607e+02 7.054e+02 9.171e+02 2.046e+03, threshold=1.411e+03, percent-clipped=6.0 +2023-04-02 04:56:54,317 INFO [train.py:903] (3/4) Epoch 15, batch 3100, loss[loss=0.2436, simple_loss=0.3211, pruned_loss=0.08299, over 19605.00 frames. ], tot_loss[loss=0.224, simple_loss=0.3001, pruned_loss=0.07396, over 3820430.46 frames. ], batch size: 61, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:57:58,294 INFO [train.py:903] (3/4) Epoch 15, batch 3150, loss[loss=0.2122, simple_loss=0.2908, pruned_loss=0.06675, over 19664.00 frames. ], tot_loss[loss=0.2248, simple_loss=0.3007, pruned_loss=0.07444, over 3809218.45 frames. ], batch size: 53, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:58:26,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 04:58:31,706 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-04-02 04:58:34,647 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=98772.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 04:58:39,659 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1935, 2.1109, 1.9056, 1.7216, 1.5480, 1.7697, 0.7117, 1.2150], + device='cuda:3'), covar=tensor([0.0523, 0.0471, 0.0359, 0.0658, 0.1022, 0.0737, 0.0988, 0.0788], + device='cuda:3'), in_proj_covar=tensor([0.0342, 0.0340, 0.0337, 0.0369, 0.0441, 0.0365, 0.0319, 0.0327], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 04:58:58,671 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.173e+02 5.220e+02 6.366e+02 8.908e+02 1.802e+03, threshold=1.273e+03, percent-clipped=4.0 +2023-04-02 04:59:01,103 INFO [train.py:903] (3/4) Epoch 15, batch 3200, loss[loss=0.2174, simple_loss=0.2988, pruned_loss=0.06798, over 19665.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2997, pruned_loss=0.07356, over 3804886.84 frames. ], batch size: 53, lr: 5.54e-03, grad_scale: 8.0 +2023-04-02 04:59:59,440 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98839.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:00:02,405 INFO [train.py:903] (3/4) Epoch 15, batch 3250, loss[loss=0.2394, simple_loss=0.3162, pruned_loss=0.08133, over 19669.00 frames. ], tot_loss[loss=0.2228, simple_loss=0.2993, pruned_loss=0.07318, over 3808414.71 frames. ], batch size: 55, lr: 5.53e-03, grad_scale: 8.0 +2023-04-02 05:00:30,847 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98864.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:00:56,827 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98887.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:00:59,975 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.355e+02 4.885e+02 6.090e+02 7.322e+02 1.846e+03, threshold=1.218e+03, percent-clipped=3.0 +2023-04-02 05:01:02,403 INFO [train.py:903] (3/4) Epoch 15, batch 3300, loss[loss=0.2256, simple_loss=0.3003, pruned_loss=0.07552, over 19789.00 frames. ], tot_loss[loss=0.2232, simple_loss=0.2992, pruned_loss=0.07355, over 3813386.57 frames. ], batch size: 56, lr: 5.53e-03, grad_scale: 8.0 +2023-04-02 05:01:08,225 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 05:01:20,964 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=98905.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:01:25,264 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98908.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:01:30,573 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98913.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:01:41,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9153, 2.0176, 2.2081, 2.6677, 1.8660, 2.5172, 2.4010, 2.0395], + device='cuda:3'), covar=tensor([0.3971, 0.3529, 0.1702, 0.2114, 0.3949, 0.1839, 0.3995, 0.3058], + device='cuda:3'), in_proj_covar=tensor([0.0831, 0.0873, 0.0669, 0.0903, 0.0815, 0.0748, 0.0807, 0.0735], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 05:02:07,326 INFO [train.py:903] (3/4) Epoch 15, batch 3350, loss[loss=0.2151, simple_loss=0.2841, pruned_loss=0.07301, over 19802.00 frames. ], tot_loss[loss=0.2232, simple_loss=0.2994, pruned_loss=0.07346, over 3822176.05 frames. ], batch size: 49, lr: 5.53e-03, grad_scale: 8.0 +2023-04-02 05:02:09,397 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-02 05:03:06,865 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98989.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:03:07,570 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.074e+02 5.628e+02 7.317e+02 9.748e+02 2.071e+03, threshold=1.463e+03, percent-clipped=8.0 +2023-04-02 05:03:09,826 INFO [train.py:903] (3/4) Epoch 15, batch 3400, loss[loss=0.2385, simple_loss=0.3143, pruned_loss=0.08141, over 19642.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2986, pruned_loss=0.07299, over 3822986.62 frames. ], batch size: 60, lr: 5.53e-03, grad_scale: 8.0 +2023-04-02 05:03:44,093 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99020.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:04:10,174 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5255, 2.2864, 1.6623, 1.5770, 2.1021, 1.3388, 1.4499, 1.8747], + device='cuda:3'), covar=tensor([0.0953, 0.0714, 0.0948, 0.0691, 0.0478, 0.1089, 0.0654, 0.0494], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0307, 0.0324, 0.0251, 0.0239, 0.0327, 0.0293, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:04:10,837 INFO [train.py:903] (3/4) Epoch 15, batch 3450, loss[loss=0.2572, simple_loss=0.3289, pruned_loss=0.09278, over 19621.00 frames. ], tot_loss[loss=0.2225, simple_loss=0.2989, pruned_loss=0.07306, over 3822644.82 frames. ], batch size: 61, lr: 5.53e-03, grad_scale: 4.0 +2023-04-02 05:04:14,066 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 05:04:34,710 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1752, 1.1033, 1.1488, 1.3091, 1.0661, 1.3228, 1.2555, 1.2397], + device='cuda:3'), covar=tensor([0.0863, 0.0982, 0.1011, 0.0669, 0.0881, 0.0793, 0.0843, 0.0732], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0221, 0.0221, 0.0240, 0.0225, 0.0207, 0.0189, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 05:04:45,800 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99069.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:05:11,152 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.718e+02 4.946e+02 5.886e+02 7.201e+02 1.354e+03, threshold=1.177e+03, percent-clipped=0.0 +2023-04-02 05:05:12,315 INFO [train.py:903] (3/4) Epoch 15, batch 3500, loss[loss=0.2203, simple_loss=0.3031, pruned_loss=0.06875, over 19464.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2992, pruned_loss=0.07336, over 3818658.47 frames. ], batch size: 64, lr: 5.53e-03, grad_scale: 4.0 +2023-04-02 05:06:15,627 INFO [train.py:903] (3/4) Epoch 15, batch 3550, loss[loss=0.2566, simple_loss=0.3317, pruned_loss=0.0907, over 19308.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2991, pruned_loss=0.07333, over 3809125.47 frames. ], batch size: 66, lr: 5.53e-03, grad_scale: 4.0 +2023-04-02 05:06:18,404 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99143.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:06:48,316 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99168.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:06:53,056 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2867, 1.2089, 1.2589, 1.3662, 1.0122, 1.3894, 1.3334, 1.3210], + device='cuda:3'), covar=tensor([0.0876, 0.0981, 0.1044, 0.0681, 0.0885, 0.0806, 0.0847, 0.0749], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0221, 0.0221, 0.0240, 0.0226, 0.0208, 0.0189, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 05:07:18,042 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.289e+02 4.734e+02 6.110e+02 7.809e+02 1.736e+03, threshold=1.222e+03, percent-clipped=8.0 +2023-04-02 05:07:19,077 INFO [train.py:903] (3/4) Epoch 15, batch 3600, loss[loss=0.2332, simple_loss=0.3048, pruned_loss=0.08078, over 19762.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2988, pruned_loss=0.07294, over 3811795.93 frames. ], batch size: 56, lr: 5.52e-03, grad_scale: 8.0 +2023-04-02 05:08:20,329 INFO [train.py:903] (3/4) Epoch 15, batch 3650, loss[loss=0.2081, simple_loss=0.2929, pruned_loss=0.06159, over 19726.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2996, pruned_loss=0.07349, over 3808176.97 frames. ], batch size: 63, lr: 5.52e-03, grad_scale: 8.0 +2023-04-02 05:08:26,434 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99247.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:08:32,076 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99252.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:08:37,709 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99257.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:09:03,832 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99276.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:09:15,265 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99286.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:09:20,757 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.641e+02 5.562e+02 6.527e+02 8.007e+02 1.277e+03, threshold=1.305e+03, percent-clipped=3.0 +2023-04-02 05:09:21,909 INFO [train.py:903] (3/4) Epoch 15, batch 3700, loss[loss=0.1842, simple_loss=0.2657, pruned_loss=0.0513, over 19742.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2985, pruned_loss=0.07273, over 3826277.00 frames. ], batch size: 51, lr: 5.52e-03, grad_scale: 8.0 +2023-04-02 05:09:34,139 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99301.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:10:13,725 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99333.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:10:24,820 INFO [train.py:903] (3/4) Epoch 15, batch 3750, loss[loss=0.2532, simple_loss=0.321, pruned_loss=0.09272, over 19531.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2985, pruned_loss=0.07305, over 3823987.33 frames. ], batch size: 56, lr: 5.52e-03, grad_scale: 8.0 +2023-04-02 05:10:25,126 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99342.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:10:40,368 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4512, 2.1528, 2.1851, 2.5738, 2.4243, 2.1466, 2.0139, 2.5326], + device='cuda:3'), covar=tensor([0.0873, 0.1595, 0.1192, 0.0947, 0.1241, 0.0483, 0.1161, 0.0600], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0354, 0.0297, 0.0244, 0.0298, 0.0249, 0.0292, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:10:56,148 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99367.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:11:01,981 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99372.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:11:26,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.684e+02 5.640e+02 6.861e+02 8.808e+02 1.909e+03, threshold=1.372e+03, percent-clipped=3.0 +2023-04-02 05:11:28,567 INFO [train.py:903] (3/4) Epoch 15, batch 3800, loss[loss=0.2156, simple_loss=0.2825, pruned_loss=0.07433, over 19805.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2987, pruned_loss=0.07281, over 3829769.82 frames. ], batch size: 47, lr: 5.52e-03, grad_scale: 8.0 +2023-04-02 05:11:41,830 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4061, 2.1575, 1.9018, 1.8878, 1.6400, 1.8435, 0.4939, 1.2017], + device='cuda:3'), covar=tensor([0.0484, 0.0482, 0.0411, 0.0618, 0.0891, 0.0613, 0.1030, 0.0798], + device='cuda:3'), in_proj_covar=tensor([0.0346, 0.0346, 0.0341, 0.0373, 0.0444, 0.0369, 0.0322, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:11:53,197 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99413.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:11:58,859 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 05:12:26,300 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99438.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:12:30,508 INFO [train.py:903] (3/4) Epoch 15, batch 3850, loss[loss=0.2328, simple_loss=0.3094, pruned_loss=0.0781, over 18140.00 frames. ], tot_loss[loss=0.2236, simple_loss=0.2997, pruned_loss=0.07371, over 3817459.42 frames. ], batch size: 84, lr: 5.52e-03, grad_scale: 4.0 +2023-04-02 05:12:37,705 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99448.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:12:44,226 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-04-02 05:13:25,909 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99486.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:13:32,485 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.342e+02 4.923e+02 6.054e+02 7.410e+02 1.518e+03, threshold=1.211e+03, percent-clipped=1.0 +2023-04-02 05:13:32,508 INFO [train.py:903] (3/4) Epoch 15, batch 3900, loss[loss=0.2139, simple_loss=0.2946, pruned_loss=0.06661, over 19543.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2987, pruned_loss=0.07277, over 3818933.47 frames. ], batch size: 54, lr: 5.52e-03, grad_scale: 4.0 +2023-04-02 05:14:18,191 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99528.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:14:33,477 INFO [train.py:903] (3/4) Epoch 15, batch 3950, loss[loss=0.2058, simple_loss=0.2945, pruned_loss=0.05851, over 19679.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.298, pruned_loss=0.07289, over 3816594.77 frames. ], batch size: 58, lr: 5.52e-03, grad_scale: 4.0 +2023-04-02 05:14:41,169 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 05:15:28,365 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99586.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:15:35,965 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99591.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:15:36,879 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.680e+02 5.329e+02 6.331e+02 8.679e+02 1.427e+03, threshold=1.266e+03, percent-clipped=6.0 +2023-04-02 05:15:36,896 INFO [train.py:903] (3/4) Epoch 15, batch 4000, loss[loss=0.2375, simple_loss=0.2958, pruned_loss=0.0896, over 19393.00 frames. ], tot_loss[loss=0.223, simple_loss=0.299, pruned_loss=0.07349, over 3808219.42 frames. ], batch size: 47, lr: 5.51e-03, grad_scale: 8.0 +2023-04-02 05:15:54,666 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99607.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:16:14,214 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99623.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:16:21,043 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99628.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:16:23,022 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 05:16:23,148 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:16:30,441 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9570, 2.5805, 1.7610, 1.8322, 2.3654, 1.6017, 1.5764, 1.9182], + device='cuda:3'), covar=tensor([0.0934, 0.0668, 0.0718, 0.0711, 0.0491, 0.0965, 0.0674, 0.0569], + device='cuda:3'), in_proj_covar=tensor([0.0292, 0.0305, 0.0321, 0.0250, 0.0239, 0.0325, 0.0290, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:16:38,165 INFO [train.py:903] (3/4) Epoch 15, batch 4050, loss[loss=0.1971, simple_loss=0.2826, pruned_loss=0.05583, over 19678.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.2982, pruned_loss=0.07286, over 3803291.05 frames. ], batch size: 58, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:16:45,330 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99648.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:16:50,848 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99653.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:17:32,698 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99686.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:17:39,406 INFO [train.py:903] (3/4) Epoch 15, batch 4100, loss[loss=0.2569, simple_loss=0.319, pruned_loss=0.0974, over 17506.00 frames. ], tot_loss[loss=0.2226, simple_loss=0.2986, pruned_loss=0.07324, over 3784498.49 frames. ], batch size: 101, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:17:40,549 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.802e+02 5.566e+02 7.429e+02 9.161e+02 2.166e+03, threshold=1.486e+03, percent-clipped=8.0 +2023-04-02 05:17:47,000 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99698.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:17:55,016 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99704.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:17:58,101 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99706.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:18:14,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 05:18:25,797 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0732, 4.4027, 4.7758, 4.7752, 1.6761, 4.4560, 3.8777, 4.4195], + device='cuda:3'), covar=tensor([0.1402, 0.0843, 0.0539, 0.0542, 0.5576, 0.0694, 0.0620, 0.1083], + device='cuda:3'), in_proj_covar=tensor([0.0725, 0.0657, 0.0860, 0.0736, 0.0771, 0.0609, 0.0519, 0.0797], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 05:18:27,041 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99729.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:18:43,280 INFO [train.py:903] (3/4) Epoch 15, batch 4150, loss[loss=0.1824, simple_loss=0.2667, pruned_loss=0.04901, over 19735.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2984, pruned_loss=0.07297, over 3789131.92 frames. ], batch size: 51, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:18:47,084 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99745.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:19:11,133 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8746, 1.1075, 1.5075, 0.6208, 2.0277, 2.4217, 2.0942, 2.5651], + device='cuda:3'), covar=tensor([0.1608, 0.3809, 0.3198, 0.2603, 0.0587, 0.0250, 0.0355, 0.0313], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0304, 0.0332, 0.0252, 0.0226, 0.0169, 0.0207, 0.0224], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:19:13,809 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 05:19:31,732 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99782.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:19:35,421 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99784.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:19:45,183 INFO [train.py:903] (3/4) Epoch 15, batch 4200, loss[loss=0.2085, simple_loss=0.2961, pruned_loss=0.06048, over 19691.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2984, pruned_loss=0.07297, over 3799866.22 frames. ], batch size: 60, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:19:47,439 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.476e+02 5.549e+02 6.772e+02 8.720e+02 1.402e+03, threshold=1.354e+03, percent-clipped=0.0 +2023-04-02 05:19:50,942 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 05:19:57,173 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99801.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:20:06,509 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99809.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:20:32,495 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99830.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:20:47,388 INFO [train.py:903] (3/4) Epoch 15, batch 4250, loss[loss=0.2843, simple_loss=0.3429, pruned_loss=0.1129, over 19724.00 frames. ], tot_loss[loss=0.222, simple_loss=0.2983, pruned_loss=0.07289, over 3806621.83 frames. ], batch size: 63, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:21:03,966 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 05:21:15,081 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 05:21:47,833 INFO [train.py:903] (3/4) Epoch 15, batch 4300, loss[loss=0.2167, simple_loss=0.2939, pruned_loss=0.06981, over 19688.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.3002, pruned_loss=0.07398, over 3808362.74 frames. ], batch size: 59, lr: 5.51e-03, grad_scale: 4.0 +2023-04-02 05:21:48,973 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.369e+02 5.517e+02 6.494e+02 8.260e+02 1.741e+03, threshold=1.299e+03, percent-clipped=4.0 +2023-04-02 05:21:53,808 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99897.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:22:35,375 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99930.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:22:41,688 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 05:22:48,549 INFO [train.py:903] (3/4) Epoch 15, batch 4350, loss[loss=0.2319, simple_loss=0.311, pruned_loss=0.07646, over 19676.00 frames. ], tot_loss[loss=0.2253, simple_loss=0.3017, pruned_loss=0.07451, over 3800563.46 frames. ], batch size: 59, lr: 5.50e-03, grad_scale: 4.0 +2023-04-02 05:22:53,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99945.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 05:23:01,035 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=99951.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:23:15,966 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99962.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:23:46,367 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99987.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:23:51,929 INFO [train.py:903] (3/4) Epoch 15, batch 4400, loss[loss=0.1917, simple_loss=0.2753, pruned_loss=0.054, over 19862.00 frames. ], tot_loss[loss=0.2254, simple_loss=0.302, pruned_loss=0.0744, over 3810843.38 frames. ], batch size: 52, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:23:53,158 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.223e+02 4.844e+02 5.787e+02 7.470e+02 1.170e+03, threshold=1.157e+03, percent-clipped=0.0 +2023-04-02 05:23:57,370 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6081, 0.9598, 1.2921, 1.5439, 2.9476, 1.2493, 2.5014, 3.5523], + device='cuda:3'), covar=tensor([0.0610, 0.3691, 0.3292, 0.2069, 0.1123, 0.2783, 0.1215, 0.0356], + device='cuda:3'), in_proj_covar=tensor([0.0379, 0.0349, 0.0370, 0.0329, 0.0356, 0.0339, 0.0350, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:24:06,744 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100001.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:24:23,726 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 05:24:31,844 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 05:24:34,570 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100025.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 05:24:35,754 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100026.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:24:47,758 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-02 05:24:57,052 INFO [train.py:903] (3/4) Epoch 15, batch 4450, loss[loss=0.1974, simple_loss=0.2773, pruned_loss=0.05875, over 19607.00 frames. ], tot_loss[loss=0.2246, simple_loss=0.3014, pruned_loss=0.07391, over 3809521.53 frames. ], batch size: 50, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:24:57,242 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=100042.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:25:00,932 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100045.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:25:07,812 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8578, 1.3566, 1.5838, 1.4229, 3.3619, 0.9358, 2.2364, 3.7727], + device='cuda:3'), covar=tensor([0.0434, 0.2741, 0.2636, 0.1895, 0.0729, 0.2812, 0.1386, 0.0244], + device='cuda:3'), in_proj_covar=tensor([0.0379, 0.0350, 0.0369, 0.0330, 0.0357, 0.0339, 0.0350, 0.0372], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:25:14,940 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100057.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:25:25,381 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100066.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:25:47,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100082.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:25:59,178 INFO [train.py:903] (3/4) Epoch 15, batch 4500, loss[loss=0.2092, simple_loss=0.2791, pruned_loss=0.06967, over 15128.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.3003, pruned_loss=0.07326, over 3807781.20 frames. ], batch size: 33, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:25:59,590 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3040, 1.4121, 1.8150, 1.5660, 2.6650, 2.1026, 2.7651, 1.2130], + device='cuda:3'), covar=tensor([0.2383, 0.4045, 0.2373, 0.1881, 0.1522, 0.2160, 0.1553, 0.4020], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0600, 0.0651, 0.0456, 0.0606, 0.0508, 0.0646, 0.0515], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:26:00,181 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.254e+02 5.166e+02 6.659e+02 8.670e+02 1.796e+03, threshold=1.332e+03, percent-clipped=6.0 +2023-04-02 05:27:01,358 INFO [train.py:903] (3/4) Epoch 15, batch 4550, loss[loss=0.2057, simple_loss=0.2784, pruned_loss=0.06652, over 19713.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.3003, pruned_loss=0.07324, over 3811567.15 frames. ], batch size: 46, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:27:10,376 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 05:27:16,343 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100153.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:27:21,784 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100157.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:27:38,127 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 05:27:47,854 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100178.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:28:04,925 INFO [train.py:903] (3/4) Epoch 15, batch 4600, loss[loss=0.2051, simple_loss=0.2922, pruned_loss=0.05897, over 19520.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2994, pruned_loss=0.07252, over 3817096.83 frames. ], batch size: 54, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:28:06,059 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.047e+02 4.707e+02 5.654e+02 7.541e+02 1.184e+03, threshold=1.131e+03, percent-clipped=0.0 +2023-04-02 05:28:19,066 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100201.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 05:28:30,960 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7204, 4.8684, 5.4924, 5.4967, 2.0411, 5.1811, 4.4440, 5.1751], + device='cuda:3'), covar=tensor([0.1415, 0.0928, 0.0445, 0.0494, 0.5288, 0.0673, 0.0563, 0.0883], + device='cuda:3'), in_proj_covar=tensor([0.0728, 0.0664, 0.0865, 0.0744, 0.0771, 0.0612, 0.0522, 0.0799], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 05:28:47,079 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100226.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:29:08,990 INFO [train.py:903] (3/4) Epoch 15, batch 4650, loss[loss=0.2468, simple_loss=0.3303, pruned_loss=0.08162, over 19772.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.3005, pruned_loss=0.07348, over 3801894.74 frames. ], batch size: 56, lr: 5.50e-03, grad_scale: 8.0 +2023-04-02 05:29:25,489 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 05:29:25,762 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100256.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:29:35,836 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 05:29:44,956 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1840, 1.2642, 1.6679, 1.2070, 2.5112, 3.4001, 3.1382, 3.6096], + device='cuda:3'), covar=tensor([0.1536, 0.3573, 0.3119, 0.2255, 0.0524, 0.0166, 0.0220, 0.0244], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0305, 0.0334, 0.0252, 0.0225, 0.0170, 0.0208, 0.0225], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:30:11,077 INFO [train.py:903] (3/4) Epoch 15, batch 4700, loss[loss=0.2594, simple_loss=0.331, pruned_loss=0.09392, over 13615.00 frames. ], tot_loss[loss=0.2229, simple_loss=0.2997, pruned_loss=0.07307, over 3791460.89 frames. ], batch size: 136, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:30:12,228 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.373e+02 5.601e+02 6.328e+02 8.331e+02 3.311e+03, threshold=1.266e+03, percent-clipped=13.0 +2023-04-02 05:30:22,240 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100301.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:30:33,229 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 05:30:50,899 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100322.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:30:55,476 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100326.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:31:13,906 INFO [train.py:903] (3/4) Epoch 15, batch 4750, loss[loss=0.2459, simple_loss=0.3208, pruned_loss=0.0855, over 19544.00 frames. ], tot_loss[loss=0.221, simple_loss=0.2981, pruned_loss=0.0719, over 3806217.14 frames. ], batch size: 56, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:31:21,212 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:31:49,667 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=100369.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:32:16,847 INFO [train.py:903] (3/4) Epoch 15, batch 4800, loss[loss=0.2247, simple_loss=0.3054, pruned_loss=0.07201, over 19745.00 frames. ], tot_loss[loss=0.2216, simple_loss=0.2983, pruned_loss=0.07241, over 3803744.52 frames. ], batch size: 63, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:32:18,028 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.926e+02 4.713e+02 6.387e+02 8.455e+02 2.006e+03, threshold=1.277e+03, percent-clipped=3.0 +2023-04-02 05:32:44,444 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100413.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:32:52,650 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100420.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:33:17,263 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100438.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:33:21,583 INFO [train.py:903] (3/4) Epoch 15, batch 4850, loss[loss=0.2026, simple_loss=0.2906, pruned_loss=0.05734, over 19175.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2967, pruned_loss=0.07154, over 3811805.45 frames. ], batch size: 69, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:33:48,715 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 05:34:10,979 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 05:34:14,955 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100484.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 05:34:15,723 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 05:34:17,764 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 05:34:24,801 INFO [train.py:903] (3/4) Epoch 15, batch 4900, loss[loss=0.2398, simple_loss=0.3219, pruned_loss=0.07887, over 18864.00 frames. ], tot_loss[loss=0.22, simple_loss=0.2969, pruned_loss=0.0715, over 3822244.05 frames. ], batch size: 74, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:34:25,926 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.390e+02 5.042e+02 5.846e+02 7.925e+02 1.600e+03, threshold=1.169e+03, percent-clipped=3.0 +2023-04-02 05:34:25,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 05:34:46,437 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 05:35:19,255 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7930, 2.5170, 2.4786, 2.8532, 2.6272, 2.5560, 2.2025, 2.9686], + device='cuda:3'), covar=tensor([0.0783, 0.1466, 0.1251, 0.0981, 0.1310, 0.0431, 0.1193, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0355, 0.0297, 0.0242, 0.0298, 0.0245, 0.0292, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:35:27,072 INFO [train.py:903] (3/4) Epoch 15, batch 4950, loss[loss=0.2634, simple_loss=0.3364, pruned_loss=0.09514, over 19327.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2959, pruned_loss=0.07105, over 3821338.12 frames. ], batch size: 66, lr: 5.49e-03, grad_scale: 8.0 +2023-04-02 05:35:45,798 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 05:36:05,342 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7707, 4.2660, 4.5281, 4.4879, 1.6441, 4.2238, 3.7219, 4.2535], + device='cuda:3'), covar=tensor([0.1540, 0.0721, 0.0557, 0.0631, 0.5593, 0.0675, 0.0610, 0.1042], + device='cuda:3'), in_proj_covar=tensor([0.0735, 0.0669, 0.0873, 0.0750, 0.0778, 0.0619, 0.0528, 0.0808], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 05:36:09,661 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 05:36:29,684 INFO [train.py:903] (3/4) Epoch 15, batch 5000, loss[loss=0.2001, simple_loss=0.2858, pruned_loss=0.05724, over 19708.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2965, pruned_loss=0.0717, over 3816453.84 frames. ], batch size: 59, lr: 5.49e-03, grad_scale: 4.0 +2023-04-02 05:36:31,854 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.571e+02 5.199e+02 6.623e+02 8.418e+02 1.165e+03, threshold=1.325e+03, percent-clipped=0.0 +2023-04-02 05:36:40,514 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 05:36:40,832 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=100600.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:36:52,299 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 05:37:33,618 INFO [train.py:903] (3/4) Epoch 15, batch 5050, loss[loss=0.2797, simple_loss=0.3516, pruned_loss=0.1039, over 18177.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2976, pruned_loss=0.07255, over 3813284.13 frames. ], batch size: 83, lr: 5.49e-03, grad_scale: 4.0 +2023-04-02 05:38:09,699 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 05:38:37,502 INFO [train.py:903] (3/4) Epoch 15, batch 5100, loss[loss=0.2217, simple_loss=0.3049, pruned_loss=0.06926, over 19661.00 frames. ], tot_loss[loss=0.2209, simple_loss=0.2975, pruned_loss=0.07216, over 3819410.00 frames. ], batch size: 55, lr: 5.48e-03, grad_scale: 4.0 +2023-04-02 05:38:39,901 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.188e+02 5.134e+02 6.124e+02 7.830e+02 1.941e+03, threshold=1.225e+03, percent-clipped=4.0 +2023-04-02 05:38:49,391 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 05:38:51,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 05:38:57,375 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 05:38:57,742 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0296, 1.1962, 1.7346, 1.2357, 2.7274, 3.7681, 3.4939, 3.9971], + device='cuda:3'), covar=tensor([0.1617, 0.3713, 0.3144, 0.2279, 0.0526, 0.0176, 0.0206, 0.0191], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0302, 0.0332, 0.0251, 0.0225, 0.0169, 0.0206, 0.0223], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:39:05,825 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100715.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:39:37,098 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100740.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:39:38,886 INFO [train.py:903] (3/4) Epoch 15, batch 5150, loss[loss=0.2128, simple_loss=0.2958, pruned_loss=0.06493, over 19679.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2986, pruned_loss=0.07278, over 3797376.67 frames. ], batch size: 53, lr: 5.48e-03, grad_scale: 4.0 +2023-04-02 05:39:39,312 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9580, 1.8577, 1.7085, 1.4947, 1.3256, 1.4366, 0.3501, 0.8148], + device='cuda:3'), covar=tensor([0.0496, 0.0548, 0.0372, 0.0609, 0.1144, 0.0705, 0.1088, 0.0937], + device='cuda:3'), in_proj_covar=tensor([0.0347, 0.0342, 0.0340, 0.0369, 0.0442, 0.0370, 0.0322, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:39:47,029 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100748.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:39:51,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 05:40:07,667 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=100764.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:40:09,141 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100765.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:40:28,241 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 05:40:41,945 INFO [train.py:903] (3/4) Epoch 15, batch 5200, loss[loss=0.2076, simple_loss=0.2821, pruned_loss=0.0665, over 19846.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.2981, pruned_loss=0.07291, over 3807214.22 frames. ], batch size: 52, lr: 5.48e-03, grad_scale: 8.0 +2023-04-02 05:40:44,524 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.350e+02 5.325e+02 6.515e+02 8.501e+02 1.618e+03, threshold=1.303e+03, percent-clipped=5.0 +2023-04-02 05:40:58,646 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 05:41:43,070 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 05:41:47,614 INFO [train.py:903] (3/4) Epoch 15, batch 5250, loss[loss=0.2311, simple_loss=0.3068, pruned_loss=0.07773, over 19658.00 frames. ], tot_loss[loss=0.221, simple_loss=0.2974, pruned_loss=0.07228, over 3811057.61 frames. ], batch size: 55, lr: 5.48e-03, grad_scale: 8.0 +2023-04-02 05:42:21,034 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1064, 2.1751, 2.4168, 2.8814, 2.1464, 2.7447, 2.5610, 2.2208], + device='cuda:3'), covar=tensor([0.3803, 0.3729, 0.1663, 0.2142, 0.3682, 0.1779, 0.3925, 0.3011], + device='cuda:3'), in_proj_covar=tensor([0.0835, 0.0877, 0.0675, 0.0906, 0.0817, 0.0752, 0.0812, 0.0739], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 05:42:25,537 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1524, 1.2767, 1.7349, 1.1720, 2.4894, 3.2268, 3.0120, 3.5045], + device='cuda:3'), covar=tensor([0.1680, 0.3743, 0.3144, 0.2469, 0.0624, 0.0278, 0.0255, 0.0250], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0304, 0.0333, 0.0252, 0.0225, 0.0170, 0.0208, 0.0224], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:42:33,480 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100879.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:42:50,397 INFO [train.py:903] (3/4) Epoch 15, batch 5300, loss[loss=0.2136, simple_loss=0.2957, pruned_loss=0.06571, over 19584.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2979, pruned_loss=0.07244, over 3822918.56 frames. ], batch size: 61, lr: 5.48e-03, grad_scale: 8.0 +2023-04-02 05:42:52,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.288e+02 5.131e+02 6.120e+02 8.353e+02 1.768e+03, threshold=1.224e+03, percent-clipped=4.0 +2023-04-02 05:43:08,004 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 05:43:52,923 INFO [train.py:903] (3/4) Epoch 15, batch 5350, loss[loss=0.1784, simple_loss=0.264, pruned_loss=0.04644, over 19483.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2981, pruned_loss=0.0725, over 3813533.14 frames. ], batch size: 49, lr: 5.48e-03, grad_scale: 8.0 +2023-04-02 05:44:08,513 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2179, 2.3257, 2.5544, 3.3279, 2.3108, 3.0565, 2.7252, 2.3744], + device='cuda:3'), covar=tensor([0.4149, 0.3841, 0.1612, 0.2187, 0.4279, 0.1825, 0.4077, 0.2962], + device='cuda:3'), in_proj_covar=tensor([0.0834, 0.0877, 0.0675, 0.0906, 0.0819, 0.0752, 0.0813, 0.0739], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 05:44:29,366 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 05:44:30,896 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100971.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:44:56,023 INFO [train.py:903] (3/4) Epoch 15, batch 5400, loss[loss=0.2038, simple_loss=0.2844, pruned_loss=0.06162, over 19471.00 frames. ], tot_loss[loss=0.2206, simple_loss=0.2973, pruned_loss=0.07198, over 3817598.07 frames. ], batch size: 49, lr: 5.48e-03, grad_scale: 8.0 +2023-04-02 05:44:58,255 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.408e+02 5.355e+02 6.832e+02 8.412e+02 2.240e+03, threshold=1.366e+03, percent-clipped=7.0 +2023-04-02 05:45:01,686 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100996.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:46:00,083 INFO [train.py:903] (3/4) Epoch 15, batch 5450, loss[loss=0.2102, simple_loss=0.2919, pruned_loss=0.06428, over 19765.00 frames. ], tot_loss[loss=0.2211, simple_loss=0.2978, pruned_loss=0.07217, over 3826209.28 frames. ], batch size: 54, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:46:30,076 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6111, 2.3610, 1.7773, 1.6432, 2.2290, 1.3462, 1.3908, 1.9440], + device='cuda:3'), covar=tensor([0.1023, 0.0729, 0.0938, 0.0793, 0.0466, 0.1187, 0.0730, 0.0455], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0307, 0.0326, 0.0252, 0.0240, 0.0330, 0.0294, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:47:04,596 INFO [train.py:903] (3/4) Epoch 15, batch 5500, loss[loss=0.2414, simple_loss=0.3131, pruned_loss=0.08488, over 19602.00 frames. ], tot_loss[loss=0.2208, simple_loss=0.2978, pruned_loss=0.07184, over 3837041.35 frames. ], batch size: 61, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:47:04,781 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=101092.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:47:06,826 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.049e+02 5.715e+02 6.860e+02 8.623e+02 1.531e+03, threshold=1.372e+03, percent-clipped=1.0 +2023-04-02 05:47:27,320 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 05:47:45,288 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101125.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:47:45,819 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-02 05:47:58,047 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101135.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:48:05,666 INFO [train.py:903] (3/4) Epoch 15, batch 5550, loss[loss=0.268, simple_loss=0.3381, pruned_loss=0.09898, over 19695.00 frames. ], tot_loss[loss=0.2219, simple_loss=0.2988, pruned_loss=0.07249, over 3830644.76 frames. ], batch size: 59, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:48:12,793 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 05:48:28,880 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101160.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:49:04,239 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 05:49:08,930 INFO [train.py:903] (3/4) Epoch 15, batch 5600, loss[loss=0.2206, simple_loss=0.2947, pruned_loss=0.07326, over 19756.00 frames. ], tot_loss[loss=0.2213, simple_loss=0.2984, pruned_loss=0.07214, over 3824367.41 frames. ], batch size: 54, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:49:11,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.323e+02 5.272e+02 6.555e+02 8.017e+02 1.573e+03, threshold=1.311e+03, percent-clipped=2.0 +2023-04-02 05:49:28,272 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=101207.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 05:50:11,396 INFO [train.py:903] (3/4) Epoch 15, batch 5650, loss[loss=0.2125, simple_loss=0.2877, pruned_loss=0.06865, over 19600.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2985, pruned_loss=0.07219, over 3831377.74 frames. ], batch size: 50, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:50:59,678 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 05:51:14,087 INFO [train.py:903] (3/4) Epoch 15, batch 5700, loss[loss=0.2085, simple_loss=0.2925, pruned_loss=0.06221, over 19659.00 frames. ], tot_loss[loss=0.2218, simple_loss=0.299, pruned_loss=0.07236, over 3828122.21 frames. ], batch size: 55, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:51:17,702 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.786e+02 5.651e+02 6.610e+02 8.419e+02 1.957e+03, threshold=1.322e+03, percent-clipped=7.0 +2023-04-02 05:51:48,363 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0395, 1.9126, 1.7328, 2.0745, 1.8078, 1.8285, 1.6856, 2.0675], + device='cuda:3'), covar=tensor([0.0980, 0.1416, 0.1469, 0.1025, 0.1372, 0.0516, 0.1340, 0.0677], + device='cuda:3'), in_proj_covar=tensor([0.0258, 0.0351, 0.0296, 0.0244, 0.0297, 0.0244, 0.0291, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:52:17,928 INFO [train.py:903] (3/4) Epoch 15, batch 5750, loss[loss=0.2463, simple_loss=0.324, pruned_loss=0.0843, over 19594.00 frames. ], tot_loss[loss=0.2208, simple_loss=0.2981, pruned_loss=0.07175, over 3831824.54 frames. ], batch size: 57, lr: 5.47e-03, grad_scale: 8.0 +2023-04-02 05:52:20,191 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 05:52:28,390 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 05:52:33,007 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 05:52:51,725 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2121, 1.3396, 1.7120, 1.2501, 2.5747, 3.4023, 3.1294, 3.5956], + device='cuda:3'), covar=tensor([0.1514, 0.3486, 0.3050, 0.2148, 0.0496, 0.0161, 0.0210, 0.0241], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0304, 0.0334, 0.0253, 0.0225, 0.0170, 0.0208, 0.0224], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:53:10,196 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5645, 1.4510, 1.5205, 1.5420, 3.0744, 1.0420, 2.2038, 3.5511], + device='cuda:3'), covar=tensor([0.0446, 0.2605, 0.2688, 0.1852, 0.0712, 0.2603, 0.1312, 0.0233], + device='cuda:3'), in_proj_covar=tensor([0.0382, 0.0353, 0.0370, 0.0336, 0.0362, 0.0341, 0.0354, 0.0374], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:53:21,257 INFO [train.py:903] (3/4) Epoch 15, batch 5800, loss[loss=0.2149, simple_loss=0.3024, pruned_loss=0.0637, over 19617.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2989, pruned_loss=0.0719, over 3825079.24 frames. ], batch size: 57, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:53:23,494 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.554e+02 5.132e+02 6.075e+02 7.663e+02 2.298e+03, threshold=1.215e+03, percent-clipped=3.0 +2023-04-02 05:54:24,304 INFO [train.py:903] (3/4) Epoch 15, batch 5850, loss[loss=0.2252, simple_loss=0.2988, pruned_loss=0.07581, over 19777.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2972, pruned_loss=0.07125, over 3827791.84 frames. ], batch size: 56, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:54:37,790 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4241, 2.4347, 2.6121, 3.1012, 2.5073, 3.0512, 2.7011, 2.3838], + device='cuda:3'), covar=tensor([0.3338, 0.3050, 0.1457, 0.1940, 0.3182, 0.1492, 0.3211, 0.2398], + device='cuda:3'), in_proj_covar=tensor([0.0836, 0.0882, 0.0679, 0.0910, 0.0823, 0.0756, 0.0815, 0.0744], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 05:54:52,524 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101463.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:54:58,893 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=101469.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:55:24,168 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101488.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 05:55:28,445 INFO [train.py:903] (3/4) Epoch 15, batch 5900, loss[loss=0.2511, simple_loss=0.3292, pruned_loss=0.08647, over 19684.00 frames. ], tot_loss[loss=0.2216, simple_loss=0.2989, pruned_loss=0.07214, over 3820483.28 frames. ], batch size: 59, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:55:30,755 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.093e+02 5.163e+02 6.557e+02 7.983e+02 1.612e+03, threshold=1.311e+03, percent-clipped=3.0 +2023-04-02 05:55:30,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 05:55:51,839 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 05:56:03,796 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9747, 1.8403, 1.7841, 2.1331, 1.8047, 1.8156, 1.9139, 2.0818], + device='cuda:3'), covar=tensor([0.0945, 0.1476, 0.1336, 0.0912, 0.1313, 0.0510, 0.1114, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0259, 0.0352, 0.0296, 0.0245, 0.0297, 0.0244, 0.0291, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 05:56:20,834 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4557, 1.5484, 1.8360, 1.6960, 2.7156, 2.3927, 2.8017, 1.0778], + device='cuda:3'), covar=tensor([0.2317, 0.4201, 0.2452, 0.1835, 0.1394, 0.1923, 0.1400, 0.4065], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0608, 0.0654, 0.0457, 0.0609, 0.0513, 0.0650, 0.0516], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 05:56:30,791 INFO [train.py:903] (3/4) Epoch 15, batch 5950, loss[loss=0.2091, simple_loss=0.2968, pruned_loss=0.06069, over 19686.00 frames. ], tot_loss[loss=0.2212, simple_loss=0.2987, pruned_loss=0.07185, over 3826782.32 frames. ], batch size: 58, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:57:24,063 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=101584.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 05:57:34,460 INFO [train.py:903] (3/4) Epoch 15, batch 6000, loss[loss=0.2276, simple_loss=0.3, pruned_loss=0.07759, over 19398.00 frames. ], tot_loss[loss=0.2209, simple_loss=0.2983, pruned_loss=0.07173, over 3819726.14 frames. ], batch size: 48, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:57:34,460 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 05:57:47,184 INFO [train.py:937] (3/4) Epoch 15, validation: loss=0.1729, simple_loss=0.2735, pruned_loss=0.0362, over 944034.00 frames. +2023-04-02 05:57:47,186 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 05:57:49,630 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.269e+02 5.208e+02 6.128e+02 8.316e+02 1.573e+03, threshold=1.226e+03, percent-clipped=3.0 +2023-04-02 05:58:03,650 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-02 05:58:49,960 INFO [train.py:903] (3/4) Epoch 15, batch 6050, loss[loss=0.1982, simple_loss=0.2718, pruned_loss=0.06235, over 19737.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2978, pruned_loss=0.07145, over 3807417.08 frames. ], batch size: 51, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:59:52,019 INFO [train.py:903] (3/4) Epoch 15, batch 6100, loss[loss=0.188, simple_loss=0.2596, pruned_loss=0.05818, over 19764.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2963, pruned_loss=0.07068, over 3827875.41 frames. ], batch size: 47, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 05:59:55,083 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.053e+02 5.047e+02 6.326e+02 7.867e+02 1.574e+03, threshold=1.265e+03, percent-clipped=9.0 +2023-04-02 06:00:05,513 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101702.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:00:56,528 INFO [train.py:903] (3/4) Epoch 15, batch 6150, loss[loss=0.1949, simple_loss=0.2847, pruned_loss=0.05248, over 19542.00 frames. ], tot_loss[loss=0.2189, simple_loss=0.296, pruned_loss=0.07084, over 3823399.73 frames. ], batch size: 56, lr: 5.46e-03, grad_scale: 8.0 +2023-04-02 06:01:23,846 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 06:01:59,349 INFO [train.py:903] (3/4) Epoch 15, batch 6200, loss[loss=0.2434, simple_loss=0.3276, pruned_loss=0.0796, over 19668.00 frames. ], tot_loss[loss=0.2181, simple_loss=0.2953, pruned_loss=0.07045, over 3826458.52 frames. ], batch size: 58, lr: 5.45e-03, grad_scale: 8.0 +2023-04-02 06:02:01,556 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.367e+02 4.959e+02 6.303e+02 7.991e+02 1.526e+03, threshold=1.261e+03, percent-clipped=1.0 +2023-04-02 06:02:57,096 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5536, 1.4655, 1.4167, 1.9369, 1.5538, 1.8821, 1.8863, 1.6953], + device='cuda:3'), covar=tensor([0.0833, 0.0905, 0.1017, 0.0722, 0.0797, 0.0715, 0.0803, 0.0633], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0222, 0.0223, 0.0243, 0.0228, 0.0209, 0.0189, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 06:03:00,196 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101840.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:03:02,073 INFO [train.py:903] (3/4) Epoch 15, batch 6250, loss[loss=0.1862, simple_loss=0.2525, pruned_loss=0.05999, over 19744.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2959, pruned_loss=0.07128, over 3824366.70 frames. ], batch size: 45, lr: 5.45e-03, grad_scale: 8.0 +2023-04-02 06:03:16,054 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6934, 1.7276, 1.4890, 1.2577, 1.2465, 1.2634, 0.2946, 0.5976], + device='cuda:3'), covar=tensor([0.0749, 0.0651, 0.0455, 0.0718, 0.1380, 0.0929, 0.1177, 0.1195], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0340, 0.0339, 0.0369, 0.0443, 0.0368, 0.0322, 0.0331], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:03:30,465 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 06:03:30,808 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101865.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:03:34,313 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3689, 1.4133, 1.7515, 1.5883, 2.4791, 2.1885, 2.6321, 0.9843], + device='cuda:3'), covar=tensor([0.2414, 0.4166, 0.2585, 0.1917, 0.1548, 0.2088, 0.1434, 0.4303], + device='cuda:3'), in_proj_covar=tensor([0.0505, 0.0601, 0.0651, 0.0455, 0.0605, 0.0511, 0.0647, 0.0512], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:04:04,536 INFO [train.py:903] (3/4) Epoch 15, batch 6300, loss[loss=0.2306, simple_loss=0.3068, pruned_loss=0.07719, over 19751.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.296, pruned_loss=0.07138, over 3817135.39 frames. ], batch size: 63, lr: 5.45e-03, grad_scale: 4.0 +2023-04-02 06:04:07,999 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.483e+02 5.666e+02 6.616e+02 7.934e+02 1.912e+03, threshold=1.323e+03, percent-clipped=2.0 +2023-04-02 06:05:03,701 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101938.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:05:08,252 INFO [train.py:903] (3/4) Epoch 15, batch 6350, loss[loss=0.1887, simple_loss=0.2646, pruned_loss=0.05643, over 14695.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2948, pruned_loss=0.07059, over 3832071.21 frames. ], batch size: 32, lr: 5.45e-03, grad_scale: 2.0 +2023-04-02 06:05:50,374 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.56 vs. limit=2.0 +2023-04-02 06:06:11,828 INFO [train.py:903] (3/4) Epoch 15, batch 6400, loss[loss=0.186, simple_loss=0.2631, pruned_loss=0.05441, over 19738.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2953, pruned_loss=0.07057, over 3839083.84 frames. ], batch size: 51, lr: 5.45e-03, grad_scale: 4.0 +2023-04-02 06:06:13,934 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 06:06:16,594 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.303e+02 4.874e+02 5.936e+02 7.490e+02 2.019e+03, threshold=1.187e+03, percent-clipped=4.0 +2023-04-02 06:06:30,758 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102005.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:06:52,969 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-02 06:07:17,087 INFO [train.py:903] (3/4) Epoch 15, batch 6450, loss[loss=0.2508, simple_loss=0.3152, pruned_loss=0.09318, over 19280.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2959, pruned_loss=0.071, over 3846656.79 frames. ], batch size: 44, lr: 5.45e-03, grad_scale: 4.0 +2023-04-02 06:07:23,181 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=102046.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:08:03,623 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 06:08:20,717 INFO [train.py:903] (3/4) Epoch 15, batch 6500, loss[loss=0.2511, simple_loss=0.322, pruned_loss=0.09007, over 18193.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2955, pruned_loss=0.07095, over 3852790.18 frames. ], batch size: 83, lr: 5.45e-03, grad_scale: 4.0 +2023-04-02 06:08:25,528 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.034e+02 4.860e+02 6.100e+02 7.866e+02 2.286e+03, threshold=1.220e+03, percent-clipped=9.0 +2023-04-02 06:08:26,632 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 06:09:23,524 INFO [train.py:903] (3/4) Epoch 15, batch 6550, loss[loss=0.203, simple_loss=0.2839, pruned_loss=0.06104, over 19474.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.296, pruned_loss=0.0714, over 3837532.76 frames. ], batch size: 49, lr: 5.44e-03, grad_scale: 4.0 +2023-04-02 06:09:47,144 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102161.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:10:26,479 INFO [train.py:903] (3/4) Epoch 15, batch 6600, loss[loss=0.2244, simple_loss=0.3103, pruned_loss=0.06928, over 19663.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2968, pruned_loss=0.07192, over 3829693.74 frames. ], batch size: 60, lr: 5.44e-03, grad_scale: 4.0 +2023-04-02 06:10:31,178 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.095e+02 4.937e+02 6.611e+02 8.175e+02 1.787e+03, threshold=1.322e+03, percent-clipped=6.0 +2023-04-02 06:10:43,389 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7418, 4.1717, 4.4111, 4.4002, 1.6962, 4.1132, 3.6493, 4.1267], + device='cuda:3'), covar=tensor([0.1427, 0.1012, 0.0591, 0.0606, 0.5547, 0.0872, 0.0624, 0.1043], + device='cuda:3'), in_proj_covar=tensor([0.0737, 0.0671, 0.0880, 0.0756, 0.0785, 0.0628, 0.0528, 0.0816], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 06:11:29,847 INFO [train.py:903] (3/4) Epoch 15, batch 6650, loss[loss=0.1938, simple_loss=0.2631, pruned_loss=0.06221, over 19782.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.2963, pruned_loss=0.07141, over 3827035.96 frames. ], batch size: 47, lr: 5.44e-03, grad_scale: 4.0 +2023-04-02 06:11:44,398 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-02 06:12:20,501 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=102282.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:12:33,724 INFO [train.py:903] (3/4) Epoch 15, batch 6700, loss[loss=0.2282, simple_loss=0.3024, pruned_loss=0.077, over 19535.00 frames. ], tot_loss[loss=0.2195, simple_loss=0.2964, pruned_loss=0.07132, over 3826977.66 frames. ], batch size: 54, lr: 5.44e-03, grad_scale: 4.0 +2023-04-02 06:12:38,441 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.247e+02 5.015e+02 6.423e+02 7.841e+02 1.581e+03, threshold=1.285e+03, percent-clipped=1.0 +2023-04-02 06:13:32,252 INFO [train.py:903] (3/4) Epoch 15, batch 6750, loss[loss=0.2438, simple_loss=0.3193, pruned_loss=0.08411, over 19338.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2971, pruned_loss=0.07139, over 3826914.78 frames. ], batch size: 66, lr: 5.44e-03, grad_scale: 4.0 +2023-04-02 06:13:40,336 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=102349.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:14:30,398 INFO [train.py:903] (3/4) Epoch 15, batch 6800, loss[loss=0.226, simple_loss=0.3064, pruned_loss=0.07284, over 19369.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.2967, pruned_loss=0.07127, over 3834641.59 frames. ], batch size: 70, lr: 5.44e-03, grad_scale: 8.0 +2023-04-02 06:14:35,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.506e+02 4.944e+02 6.022e+02 7.665e+02 3.022e+03, threshold=1.204e+03, percent-clipped=5.0 +2023-04-02 06:14:37,028 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102397.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 06:14:57,426 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102417.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:15:15,153 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 06:15:16,183 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 06:15:19,105 INFO [train.py:903] (3/4) Epoch 16, batch 0, loss[loss=0.2505, simple_loss=0.3085, pruned_loss=0.09629, over 19459.00 frames. ], tot_loss[loss=0.2505, simple_loss=0.3085, pruned_loss=0.09629, over 19459.00 frames. ], batch size: 49, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:15:19,105 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 06:15:29,713 INFO [train.py:937] (3/4) Epoch 16, validation: loss=0.1737, simple_loss=0.2745, pruned_loss=0.03646, over 944034.00 frames. +2023-04-02 06:15:29,714 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 06:15:45,587 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 06:15:58,370 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=102442.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:16:06,636 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4111, 1.3356, 1.3321, 1.7019, 1.4118, 1.7005, 1.7951, 1.5690], + device='cuda:3'), covar=tensor([0.0821, 0.0918, 0.0995, 0.0710, 0.0775, 0.0729, 0.0762, 0.0651], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0222, 0.0222, 0.0243, 0.0226, 0.0209, 0.0189, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 06:16:24,719 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102464.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:16:33,054 INFO [train.py:903] (3/4) Epoch 16, batch 50, loss[loss=0.2251, simple_loss=0.3023, pruned_loss=0.07393, over 19358.00 frames. ], tot_loss[loss=0.2201, simple_loss=0.297, pruned_loss=0.07161, over 856792.28 frames. ], batch size: 70, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:17:04,308 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.709e+02 4.865e+02 6.426e+02 8.395e+02 1.744e+03, threshold=1.285e+03, percent-clipped=5.0 +2023-04-02 06:17:08,781 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 06:17:29,601 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8074, 1.4581, 1.5711, 1.4906, 3.3379, 0.9950, 2.4129, 3.7888], + device='cuda:3'), covar=tensor([0.0497, 0.2912, 0.2907, 0.1967, 0.0756, 0.2712, 0.1295, 0.0246], + device='cuda:3'), in_proj_covar=tensor([0.0382, 0.0350, 0.0369, 0.0334, 0.0359, 0.0338, 0.0350, 0.0374], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:17:33,675 INFO [train.py:903] (3/4) Epoch 16, batch 100, loss[loss=0.1938, simple_loss=0.2813, pruned_loss=0.05317, over 19521.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2964, pruned_loss=0.07109, over 1528270.28 frames. ], batch size: 56, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:17:47,702 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 06:18:13,827 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8631, 1.9639, 2.2137, 2.5023, 1.8210, 2.3626, 2.3044, 1.9596], + device='cuda:3'), covar=tensor([0.3926, 0.3470, 0.1636, 0.2044, 0.3693, 0.1853, 0.4167, 0.3079], + device='cuda:3'), in_proj_covar=tensor([0.0833, 0.0881, 0.0678, 0.0903, 0.0821, 0.0756, 0.0811, 0.0743], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 06:18:34,759 INFO [train.py:903] (3/4) Epoch 16, batch 150, loss[loss=0.1959, simple_loss=0.2793, pruned_loss=0.05623, over 19533.00 frames. ], tot_loss[loss=0.2176, simple_loss=0.2945, pruned_loss=0.07032, over 2045462.87 frames. ], batch size: 56, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:19:06,737 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8292, 1.6333, 1.4522, 1.9167, 1.8134, 1.5942, 1.4495, 1.7974], + device='cuda:3'), covar=tensor([0.1057, 0.1430, 0.1529, 0.0923, 0.1147, 0.0572, 0.1376, 0.0753], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0352, 0.0298, 0.0246, 0.0299, 0.0246, 0.0292, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:19:07,435 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.312e+02 5.578e+02 6.638e+02 8.298e+02 1.665e+03, threshold=1.328e+03, percent-clipped=4.0 +2023-04-02 06:19:36,780 INFO [train.py:903] (3/4) Epoch 16, batch 200, loss[loss=0.1977, simple_loss=0.2851, pruned_loss=0.05517, over 19750.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2953, pruned_loss=0.07067, over 2447079.30 frames. ], batch size: 63, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:19:38,849 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 06:20:18,522 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102653.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 06:20:29,352 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2716, 2.1495, 1.8892, 1.6930, 1.5887, 1.7459, 0.4870, 1.1770], + device='cuda:3'), covar=tensor([0.0500, 0.0488, 0.0384, 0.0698, 0.1066, 0.0735, 0.1129, 0.0828], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0342, 0.0339, 0.0371, 0.0444, 0.0371, 0.0324, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:20:38,278 INFO [train.py:903] (3/4) Epoch 16, batch 250, loss[loss=0.2234, simple_loss=0.307, pruned_loss=0.06991, over 17997.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2958, pruned_loss=0.07078, over 2754158.03 frames. ], batch size: 83, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:20:43,970 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-02 06:20:51,617 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=102678.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:21:12,162 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.486e+02 5.293e+02 5.976e+02 7.347e+02 1.638e+03, threshold=1.195e+03, percent-clipped=3.0 +2023-04-02 06:21:43,459 INFO [train.py:903] (3/4) Epoch 16, batch 300, loss[loss=0.2146, simple_loss=0.2889, pruned_loss=0.07009, over 19712.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.2961, pruned_loss=0.07157, over 2968314.05 frames. ], batch size: 51, lr: 5.26e-03, grad_scale: 8.0 +2023-04-02 06:21:43,893 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102720.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:22:13,600 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=102745.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:22:21,954 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0443, 2.1004, 2.3670, 2.8249, 2.0294, 2.6563, 2.5556, 2.1885], + device='cuda:3'), covar=tensor([0.4082, 0.3734, 0.1632, 0.2044, 0.3918, 0.1860, 0.4011, 0.2968], + device='cuda:3'), in_proj_covar=tensor([0.0840, 0.0887, 0.0680, 0.0909, 0.0826, 0.0759, 0.0816, 0.0745], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 06:22:44,953 INFO [train.py:903] (3/4) Epoch 16, batch 350, loss[loss=0.1754, simple_loss=0.2522, pruned_loss=0.04927, over 19393.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2952, pruned_loss=0.07092, over 3161019.31 frames. ], batch size: 48, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:22:47,853 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4624, 1.4909, 1.7034, 1.6095, 2.5500, 2.1576, 2.5224, 1.3238], + device='cuda:3'), covar=tensor([0.2130, 0.3855, 0.2383, 0.1726, 0.1321, 0.1972, 0.1366, 0.3791], + device='cuda:3'), in_proj_covar=tensor([0.0500, 0.0597, 0.0650, 0.0453, 0.0602, 0.0509, 0.0648, 0.0511], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:22:50,857 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 06:23:16,113 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.122e+02 5.254e+02 6.186e+02 7.503e+02 2.205e+03, threshold=1.237e+03, percent-clipped=4.0 +2023-04-02 06:23:39,980 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-02 06:23:47,562 INFO [train.py:903] (3/4) Epoch 16, batch 400, loss[loss=0.1724, simple_loss=0.252, pruned_loss=0.04644, over 19762.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2955, pruned_loss=0.0712, over 3312416.05 frames. ], batch size: 47, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:24:49,384 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102869.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:24:50,278 INFO [train.py:903] (3/4) Epoch 16, batch 450, loss[loss=0.2789, simple_loss=0.3347, pruned_loss=0.1115, over 13514.00 frames. ], tot_loss[loss=0.2202, simple_loss=0.2963, pruned_loss=0.07198, over 3423187.97 frames. ], batch size: 136, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:25:11,603 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 06:25:22,281 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.385e+02 4.837e+02 5.955e+02 8.003e+02 1.401e+03, threshold=1.191e+03, percent-clipped=4.0 +2023-04-02 06:25:24,676 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 06:25:25,876 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 06:25:30,873 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6353, 1.6392, 1.5606, 1.4262, 1.2865, 1.4071, 0.6748, 0.9743], + device='cuda:3'), covar=tensor([0.0427, 0.0449, 0.0278, 0.0390, 0.0739, 0.0501, 0.0827, 0.0664], + device='cuda:3'), in_proj_covar=tensor([0.0347, 0.0338, 0.0335, 0.0367, 0.0438, 0.0368, 0.0321, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:25:32,994 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102905.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:25:52,023 INFO [train.py:903] (3/4) Epoch 16, batch 500, loss[loss=0.2122, simple_loss=0.2943, pruned_loss=0.06508, over 19734.00 frames. ], tot_loss[loss=0.2206, simple_loss=0.2973, pruned_loss=0.07196, over 3508798.97 frames. ], batch size: 51, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:26:54,172 INFO [train.py:903] (3/4) Epoch 16, batch 550, loss[loss=0.1967, simple_loss=0.2714, pruned_loss=0.06096, over 19018.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.2961, pruned_loss=0.07133, over 3588971.88 frames. ], batch size: 42, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:27:24,958 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.789e+02 5.227e+02 6.443e+02 7.702e+02 1.436e+03, threshold=1.289e+03, percent-clipped=3.0 +2023-04-02 06:27:54,419 INFO [train.py:903] (3/4) Epoch 16, batch 600, loss[loss=0.2258, simple_loss=0.3026, pruned_loss=0.07452, over 17495.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2953, pruned_loss=0.07065, over 3646863.21 frames. ], batch size: 101, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:28:37,068 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 06:28:55,606 INFO [train.py:903] (3/4) Epoch 16, batch 650, loss[loss=0.2061, simple_loss=0.2724, pruned_loss=0.06993, over 18189.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.2937, pruned_loss=0.06962, over 3699700.72 frames. ], batch size: 40, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:29:28,782 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.210e+02 4.715e+02 6.008e+02 7.942e+02 1.451e+03, threshold=1.202e+03, percent-clipped=1.0 +2023-04-02 06:29:53,904 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103116.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:29:57,413 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.3021, 5.2642, 6.1143, 6.0914, 2.1794, 5.7264, 4.8731, 5.7173], + device='cuda:3'), covar=tensor([0.1324, 0.0622, 0.0487, 0.0452, 0.5190, 0.0503, 0.0557, 0.1068], + device='cuda:3'), in_proj_covar=tensor([0.0733, 0.0666, 0.0876, 0.0749, 0.0775, 0.0619, 0.0523, 0.0804], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 06:29:58,304 INFO [train.py:903] (3/4) Epoch 16, batch 700, loss[loss=0.1724, simple_loss=0.2427, pruned_loss=0.05101, over 19736.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2948, pruned_loss=0.06988, over 3726889.30 frames. ], batch size: 46, lr: 5.25e-03, grad_scale: 8.0 +2023-04-02 06:30:35,173 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103150.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:31:00,295 INFO [train.py:903] (3/4) Epoch 16, batch 750, loss[loss=0.2506, simple_loss=0.3183, pruned_loss=0.09145, over 19538.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2962, pruned_loss=0.07086, over 3733356.29 frames. ], batch size: 54, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:31:33,687 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.142e+02 5.236e+02 6.262e+02 8.063e+02 1.865e+03, threshold=1.252e+03, percent-clipped=5.0 +2023-04-02 06:31:55,101 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=103213.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:32:03,180 INFO [train.py:903] (3/4) Epoch 16, batch 800, loss[loss=0.1884, simple_loss=0.2677, pruned_loss=0.05453, over 19737.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2962, pruned_loss=0.07104, over 3761940.43 frames. ], batch size: 51, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:32:18,129 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 06:32:36,484 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9140, 1.8579, 1.5554, 2.1045, 1.7511, 1.6889, 1.7452, 1.9076], + device='cuda:3'), covar=tensor([0.0999, 0.1354, 0.1464, 0.0877, 0.1273, 0.0548, 0.1217, 0.0712], + device='cuda:3'), in_proj_covar=tensor([0.0260, 0.0352, 0.0296, 0.0244, 0.0298, 0.0245, 0.0292, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:32:38,496 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=103249.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:33:04,743 INFO [train.py:903] (3/4) Epoch 16, batch 850, loss[loss=0.1693, simple_loss=0.2478, pruned_loss=0.04543, over 19741.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2959, pruned_loss=0.07079, over 3785277.65 frames. ], batch size: 46, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:33:38,416 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.826e+02 4.865e+02 6.263e+02 7.829e+02 1.710e+03, threshold=1.253e+03, percent-clipped=2.0 +2023-04-02 06:33:57,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 06:34:06,720 INFO [train.py:903] (3/4) Epoch 16, batch 900, loss[loss=0.2214, simple_loss=0.2954, pruned_loss=0.07369, over 19682.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2955, pruned_loss=0.07048, over 3791173.58 frames. ], batch size: 53, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:34:17,413 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103328.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:34:52,836 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5802, 1.2479, 1.4553, 1.2292, 2.1819, 0.9459, 2.0629, 2.4372], + device='cuda:3'), covar=tensor([0.0659, 0.2679, 0.2618, 0.1635, 0.0893, 0.2052, 0.0953, 0.0441], + device='cuda:3'), in_proj_covar=tensor([0.0385, 0.0351, 0.0370, 0.0334, 0.0361, 0.0340, 0.0349, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:35:01,527 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103364.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:35:08,188 INFO [train.py:903] (3/4) Epoch 16, batch 950, loss[loss=0.2003, simple_loss=0.2823, pruned_loss=0.05915, over 19652.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2952, pruned_loss=0.07041, over 3793344.03 frames. ], batch size: 55, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:35:13,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 06:35:27,470 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103384.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 06:35:34,305 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5500, 1.1514, 1.3932, 1.2669, 2.1426, 1.0018, 1.9192, 2.4569], + device='cuda:3'), covar=tensor([0.0677, 0.2743, 0.2747, 0.1600, 0.0945, 0.1987, 0.1044, 0.0455], + device='cuda:3'), in_proj_covar=tensor([0.0385, 0.0352, 0.0370, 0.0334, 0.0361, 0.0339, 0.0350, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:35:40,823 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.146e+02 5.367e+02 6.234e+02 7.823e+02 2.113e+03, threshold=1.247e+03, percent-clipped=3.0 +2023-04-02 06:36:12,153 INFO [train.py:903] (3/4) Epoch 16, batch 1000, loss[loss=0.2258, simple_loss=0.3048, pruned_loss=0.07345, over 19685.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2959, pruned_loss=0.07107, over 3803134.28 frames. ], batch size: 53, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:37:03,122 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=103460.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:37:07,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 06:37:14,586 INFO [train.py:903] (3/4) Epoch 16, batch 1050, loss[loss=0.2383, simple_loss=0.3186, pruned_loss=0.07899, over 19518.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2957, pruned_loss=0.0709, over 3801110.04 frames. ], batch size: 56, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:37:43,711 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=103494.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:37:45,897 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.718e+02 5.497e+02 6.450e+02 8.583e+02 2.663e+03, threshold=1.290e+03, percent-clipped=6.0 +2023-04-02 06:37:49,195 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 06:38:15,967 INFO [train.py:903] (3/4) Epoch 16, batch 1100, loss[loss=0.2125, simple_loss=0.3018, pruned_loss=0.06156, over 19759.00 frames. ], tot_loss[loss=0.2198, simple_loss=0.2967, pruned_loss=0.07139, over 3803929.76 frames. ], batch size: 54, lr: 5.24e-03, grad_scale: 8.0 +2023-04-02 06:38:54,859 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0362, 1.7735, 1.6820, 2.0807, 1.8229, 1.8829, 1.7941, 1.9987], + device='cuda:3'), covar=tensor([0.0979, 0.1451, 0.1398, 0.0935, 0.1297, 0.0493, 0.1202, 0.0680], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0355, 0.0298, 0.0247, 0.0301, 0.0247, 0.0294, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:39:18,066 INFO [train.py:903] (3/4) Epoch 16, batch 1150, loss[loss=0.1861, simple_loss=0.2672, pruned_loss=0.05248, over 19423.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2944, pruned_loss=0.07001, over 3809862.08 frames. ], batch size: 48, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:39:25,933 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103575.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:39:37,385 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103584.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:39:50,623 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.809e+02 5.106e+02 6.190e+02 8.719e+02 1.567e+03, threshold=1.238e+03, percent-clipped=4.0 +2023-04-02 06:40:06,762 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103609.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:40:06,804 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103609.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:40:21,155 INFO [train.py:903] (3/4) Epoch 16, batch 1200, loss[loss=0.2145, simple_loss=0.289, pruned_loss=0.06998, over 19625.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2953, pruned_loss=0.07064, over 3818357.60 frames. ], batch size: 50, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:40:21,635 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103620.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:40:52,966 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103645.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:40:54,977 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 06:41:24,588 INFO [train.py:903] (3/4) Epoch 16, batch 1250, loss[loss=0.2093, simple_loss=0.2831, pruned_loss=0.06772, over 19759.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2944, pruned_loss=0.07021, over 3822935.92 frames. ], batch size: 51, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:41:56,589 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.406e+02 4.955e+02 6.144e+02 7.729e+02 1.641e+03, threshold=1.229e+03, percent-clipped=4.0 +2023-04-02 06:42:00,283 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5340, 1.3351, 1.4335, 2.0101, 3.0903, 1.5256, 2.3260, 3.5214], + device='cuda:3'), covar=tensor([0.0514, 0.2780, 0.2885, 0.1437, 0.0740, 0.1932, 0.1207, 0.0257], + device='cuda:3'), in_proj_covar=tensor([0.0383, 0.0351, 0.0368, 0.0333, 0.0359, 0.0339, 0.0348, 0.0375], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:42:24,682 INFO [train.py:903] (3/4) Epoch 16, batch 1300, loss[loss=0.2484, simple_loss=0.3224, pruned_loss=0.08724, over 19318.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2945, pruned_loss=0.07017, over 3826583.16 frames. ], batch size: 66, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:42:34,227 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5041, 2.2898, 1.6233, 1.5153, 2.0721, 1.2650, 1.4949, 1.8960], + device='cuda:3'), covar=tensor([0.1016, 0.0671, 0.1076, 0.0763, 0.0528, 0.1212, 0.0691, 0.0433], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0309, 0.0329, 0.0255, 0.0243, 0.0329, 0.0296, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:42:35,157 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=103728.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:43:12,241 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103758.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:43:26,072 INFO [train.py:903] (3/4) Epoch 16, batch 1350, loss[loss=0.2173, simple_loss=0.3003, pruned_loss=0.06711, over 19674.00 frames. ], tot_loss[loss=0.2177, simple_loss=0.2949, pruned_loss=0.07031, over 3833526.73 frames. ], batch size: 60, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:43:43,882 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103783.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:43:59,437 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.194e+02 5.214e+02 6.633e+02 9.392e+02 2.118e+03, threshold=1.327e+03, percent-clipped=8.0 +2023-04-02 06:44:30,116 INFO [train.py:903] (3/4) Epoch 16, batch 1400, loss[loss=0.2086, simple_loss=0.2968, pruned_loss=0.06025, over 19660.00 frames. ], tot_loss[loss=0.2178, simple_loss=0.2951, pruned_loss=0.07025, over 3827867.84 frames. ], batch size: 58, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:44:42,081 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.13 vs. limit=5.0 +2023-04-02 06:44:43,973 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103831.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:44:57,959 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103843.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 06:45:14,487 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103856.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:45:26,693 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103865.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:45:32,163 INFO [train.py:903] (3/4) Epoch 16, batch 1450, loss[loss=0.2163, simple_loss=0.2966, pruned_loss=0.06797, over 19741.00 frames. ], tot_loss[loss=0.2189, simple_loss=0.2959, pruned_loss=0.07092, over 3827849.31 frames. ], batch size: 51, lr: 5.23e-03, grad_scale: 8.0 +2023-04-02 06:45:32,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 06:45:56,458 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103890.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:46:01,056 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103894.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:46:03,878 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.072e+02 4.634e+02 5.962e+02 7.073e+02 1.523e+03, threshold=1.192e+03, percent-clipped=2.0 +2023-04-02 06:46:22,702 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5318, 2.3196, 1.6125, 1.4367, 2.1485, 1.2384, 1.2796, 2.0011], + device='cuda:3'), covar=tensor([0.1135, 0.0767, 0.1141, 0.0899, 0.0560, 0.1347, 0.0871, 0.0468], + device='cuda:3'), in_proj_covar=tensor([0.0299, 0.0311, 0.0331, 0.0257, 0.0244, 0.0333, 0.0297, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:46:33,191 INFO [train.py:903] (3/4) Epoch 16, batch 1500, loss[loss=0.268, simple_loss=0.3394, pruned_loss=0.09831, over 19603.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.2966, pruned_loss=0.07112, over 3831813.69 frames. ], batch size: 57, lr: 5.23e-03, grad_scale: 16.0 +2023-04-02 06:47:35,230 INFO [train.py:903] (3/4) Epoch 16, batch 1550, loss[loss=0.2449, simple_loss=0.3202, pruned_loss=0.08483, over 18771.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2972, pruned_loss=0.07132, over 3835650.45 frames. ], batch size: 74, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:48:02,907 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5137, 2.4326, 2.0877, 2.9138, 2.5029, 2.1776, 2.2427, 2.7578], + device='cuda:3'), covar=tensor([0.0925, 0.1668, 0.1437, 0.0889, 0.1316, 0.0525, 0.1239, 0.0590], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0355, 0.0299, 0.0247, 0.0302, 0.0247, 0.0295, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:48:09,263 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.399e+02 5.500e+02 6.727e+02 8.636e+02 1.580e+03, threshold=1.345e+03, percent-clipped=8.0 +2023-04-02 06:48:14,274 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4496, 1.3834, 1.3610, 1.8153, 1.4179, 1.7177, 1.7601, 1.5749], + device='cuda:3'), covar=tensor([0.0777, 0.0895, 0.0977, 0.0648, 0.0790, 0.0702, 0.0780, 0.0652], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0222, 0.0224, 0.0243, 0.0225, 0.0210, 0.0191, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 06:48:24,334 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104008.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:48:32,661 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104015.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:48:39,348 INFO [train.py:903] (3/4) Epoch 16, batch 1600, loss[loss=0.2009, simple_loss=0.2758, pruned_loss=0.06304, over 19401.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2964, pruned_loss=0.07105, over 3837873.28 frames. ], batch size: 48, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:49:05,527 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 06:49:30,573 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6517, 1.4049, 1.4173, 2.3094, 1.6514, 1.9831, 2.0687, 1.6451], + device='cuda:3'), covar=tensor([0.0843, 0.1039, 0.1129, 0.0802, 0.0887, 0.0757, 0.0896, 0.0754], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0226, 0.0245, 0.0227, 0.0211, 0.0192, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 06:49:38,525 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104067.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:49:41,609 INFO [train.py:903] (3/4) Epoch 16, batch 1650, loss[loss=0.2045, simple_loss=0.2902, pruned_loss=0.05942, over 19668.00 frames. ], tot_loss[loss=0.22, simple_loss=0.2969, pruned_loss=0.07158, over 3823494.29 frames. ], batch size: 55, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:50:14,819 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.025e+02 5.150e+02 6.179e+02 7.587e+02 1.568e+03, threshold=1.236e+03, percent-clipped=4.0 +2023-04-02 06:50:18,602 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104099.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:50:19,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9137, 2.0615, 2.1570, 2.7858, 2.0559, 2.5905, 2.3358, 1.9906], + device='cuda:3'), covar=tensor([0.3909, 0.3321, 0.1693, 0.1898, 0.3516, 0.1716, 0.4194, 0.3050], + device='cuda:3'), in_proj_covar=tensor([0.0842, 0.0889, 0.0683, 0.0914, 0.0827, 0.0764, 0.0816, 0.0748], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 06:50:22,743 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104102.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:50:43,581 INFO [train.py:903] (3/4) Epoch 16, batch 1700, loss[loss=0.2404, simple_loss=0.32, pruned_loss=0.08041, over 19579.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2991, pruned_loss=0.07262, over 3805782.62 frames. ], batch size: 52, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:50:43,968 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9160, 1.1700, 1.5434, 0.5712, 1.9939, 2.4246, 2.1039, 2.6008], + device='cuda:3'), covar=tensor([0.1555, 0.3631, 0.3076, 0.2503, 0.0564, 0.0291, 0.0329, 0.0296], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0308, 0.0336, 0.0257, 0.0229, 0.0173, 0.0209, 0.0229], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 06:50:48,647 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104124.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 06:50:52,856 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104127.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:51:25,773 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 06:51:44,968 INFO [train.py:903] (3/4) Epoch 16, batch 1750, loss[loss=0.2563, simple_loss=0.3231, pruned_loss=0.09475, over 19663.00 frames. ], tot_loss[loss=0.2227, simple_loss=0.2991, pruned_loss=0.07321, over 3818745.17 frames. ], batch size: 60, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:52:19,019 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.135e+02 4.884e+02 5.867e+02 6.930e+02 2.034e+03, threshold=1.173e+03, percent-clipped=2.0 +2023-04-02 06:52:45,665 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104217.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:52:47,903 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4345, 1.3312, 1.5323, 1.4893, 2.9797, 1.0939, 2.3523, 3.3017], + device='cuda:3'), covar=tensor([0.0478, 0.2727, 0.2706, 0.1767, 0.0704, 0.2471, 0.1129, 0.0286], + device='cuda:3'), in_proj_covar=tensor([0.0381, 0.0349, 0.0367, 0.0333, 0.0356, 0.0340, 0.0352, 0.0375], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:52:48,724 INFO [train.py:903] (3/4) Epoch 16, batch 1800, loss[loss=0.2142, simple_loss=0.2921, pruned_loss=0.06813, over 19793.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2981, pruned_loss=0.07239, over 3806789.94 frames. ], batch size: 56, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:53:09,937 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104238.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:53:14,912 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104242.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:53:24,021 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0905, 1.7567, 1.9035, 2.7732, 2.1307, 2.4095, 2.5965, 2.3862], + device='cuda:3'), covar=tensor([0.0827, 0.1003, 0.1012, 0.0884, 0.0865, 0.0770, 0.0811, 0.0666], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0224, 0.0226, 0.0245, 0.0228, 0.0212, 0.0192, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 06:53:46,177 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 06:53:49,825 INFO [train.py:903] (3/4) Epoch 16, batch 1850, loss[loss=0.2157, simple_loss=0.2824, pruned_loss=0.07446, over 19479.00 frames. ], tot_loss[loss=0.2201, simple_loss=0.297, pruned_loss=0.07163, over 3814946.67 frames. ], batch size: 49, lr: 5.22e-03, grad_scale: 8.0 +2023-04-02 06:54:21,794 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 06:54:22,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.793e+02 5.248e+02 6.749e+02 7.716e+02 1.558e+03, threshold=1.350e+03, percent-clipped=5.0 +2023-04-02 06:54:51,925 INFO [train.py:903] (3/4) Epoch 16, batch 1900, loss[loss=0.2145, simple_loss=0.2883, pruned_loss=0.07037, over 19733.00 frames. ], tot_loss[loss=0.2205, simple_loss=0.2973, pruned_loss=0.07182, over 3820664.73 frames. ], batch size: 45, lr: 5.22e-03, grad_scale: 4.0 +2023-04-02 06:55:09,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 06:55:15,665 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 06:55:18,494 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-02 06:55:31,877 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104352.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:55:34,086 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104353.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:55:39,887 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 06:55:41,223 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104359.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:55:53,978 INFO [train.py:903] (3/4) Epoch 16, batch 1950, loss[loss=0.2301, simple_loss=0.3123, pruned_loss=0.07393, over 19746.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2956, pruned_loss=0.07041, over 3833280.67 frames. ], batch size: 63, lr: 5.21e-03, grad_scale: 4.0 +2023-04-02 06:56:00,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4607, 2.4077, 2.2396, 2.8689, 2.5174, 2.2998, 2.1398, 2.5775], + device='cuda:3'), covar=tensor([0.0900, 0.1391, 0.1282, 0.0882, 0.1156, 0.0479, 0.1203, 0.0643], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0354, 0.0300, 0.0248, 0.0301, 0.0249, 0.0294, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 06:56:30,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.975e+02 4.688e+02 6.377e+02 8.120e+02 1.703e+03, threshold=1.275e+03, percent-clipped=4.0 +2023-04-02 06:56:46,834 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=104411.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:56:58,264 INFO [train.py:903] (3/4) Epoch 16, batch 2000, loss[loss=0.2376, simple_loss=0.3139, pruned_loss=0.08069, over 19359.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2965, pruned_loss=0.0709, over 3825080.21 frames. ], batch size: 70, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 06:57:57,345 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 06:57:57,646 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104467.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:58:00,776 INFO [train.py:903] (3/4) Epoch 16, batch 2050, loss[loss=0.2055, simple_loss=0.2871, pruned_loss=0.06192, over 19659.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2958, pruned_loss=0.07033, over 3828101.83 frames. ], batch size: 53, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 06:58:04,673 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104473.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:58:05,676 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104474.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:58:14,089 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 06:58:15,347 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 06:58:35,728 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.098e+02 4.885e+02 6.080e+02 8.555e+02 1.693e+03, threshold=1.216e+03, percent-clipped=6.0 +2023-04-02 06:58:36,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104498.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:58:36,188 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104498.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:58:40,139 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 06:59:02,741 INFO [train.py:903] (3/4) Epoch 16, batch 2100, loss[loss=0.2076, simple_loss=0.2748, pruned_loss=0.07015, over 19791.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2967, pruned_loss=0.07075, over 3823393.83 frames. ], batch size: 47, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 06:59:06,446 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104523.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:59:09,999 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104526.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 06:59:33,815 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 06:59:55,527 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 07:00:04,689 INFO [train.py:903] (3/4) Epoch 16, batch 2150, loss[loss=0.2349, simple_loss=0.2971, pruned_loss=0.08635, over 19816.00 frames. ], tot_loss[loss=0.2195, simple_loss=0.2967, pruned_loss=0.07114, over 3816522.95 frames. ], batch size: 49, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 07:00:39,765 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 5.547e+02 6.907e+02 8.298e+02 2.194e+03, threshold=1.381e+03, percent-clipped=3.0 +2023-04-02 07:00:53,810 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104609.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:01:08,291 INFO [train.py:903] (3/4) Epoch 16, batch 2200, loss[loss=0.2143, simple_loss=0.2872, pruned_loss=0.07069, over 19610.00 frames. ], tot_loss[loss=0.2201, simple_loss=0.2969, pruned_loss=0.07166, over 3819443.19 frames. ], batch size: 50, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 07:01:26,427 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104634.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:02:09,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3835, 2.1630, 1.5832, 1.4339, 1.9378, 1.1855, 1.3375, 1.8383], + device='cuda:3'), covar=tensor([0.0996, 0.0737, 0.1058, 0.0720, 0.0552, 0.1261, 0.0682, 0.0460], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0309, 0.0328, 0.0255, 0.0244, 0.0332, 0.0292, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:02:12,256 INFO [train.py:903] (3/4) Epoch 16, batch 2250, loss[loss=0.2465, simple_loss=0.318, pruned_loss=0.08746, over 19379.00 frames. ], tot_loss[loss=0.22, simple_loss=0.2962, pruned_loss=0.07184, over 3822518.67 frames. ], batch size: 66, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 07:02:46,758 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.071e+02 4.921e+02 5.898e+02 6.952e+02 1.646e+03, threshold=1.180e+03, percent-clipped=1.0 +2023-04-02 07:03:15,189 INFO [train.py:903] (3/4) Epoch 16, batch 2300, loss[loss=0.206, simple_loss=0.2701, pruned_loss=0.07098, over 19409.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2953, pruned_loss=0.07132, over 3811482.68 frames. ], batch size: 48, lr: 5.21e-03, grad_scale: 8.0 +2023-04-02 07:03:19,197 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104723.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:03:27,425 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104730.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:03:30,537 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 07:03:51,576 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104748.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:03:59,127 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.53 vs. limit=2.0 +2023-04-02 07:03:59,896 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104755.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:04:11,302 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6062, 2.3509, 2.1390, 2.7061, 2.3027, 2.4085, 2.0749, 2.5634], + device='cuda:3'), covar=tensor([0.0868, 0.1593, 0.1341, 0.1020, 0.1395, 0.0450, 0.1258, 0.0603], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0350, 0.0297, 0.0246, 0.0298, 0.0247, 0.0292, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:04:17,962 INFO [train.py:903] (3/4) Epoch 16, batch 2350, loss[loss=0.2047, simple_loss=0.2781, pruned_loss=0.06562, over 19336.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2961, pruned_loss=0.07185, over 3792610.15 frames. ], batch size: 48, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:04:34,937 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104782.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:04:53,849 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.557e+02 5.048e+02 6.738e+02 8.844e+02 1.580e+03, threshold=1.348e+03, percent-clipped=5.0 +2023-04-02 07:05:00,896 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 07:05:05,755 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104807.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:05:18,361 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 07:05:22,504 INFO [train.py:903] (3/4) Epoch 16, batch 2400, loss[loss=0.21, simple_loss=0.2966, pruned_loss=0.06172, over 19529.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2955, pruned_loss=0.0711, over 3792310.73 frames. ], batch size: 54, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:06:24,571 INFO [train.py:903] (3/4) Epoch 16, batch 2450, loss[loss=0.1896, simple_loss=0.2639, pruned_loss=0.05764, over 19751.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2949, pruned_loss=0.07085, over 3801873.42 frames. ], batch size: 46, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:06:54,712 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.05 vs. limit=5.0 +2023-04-02 07:07:00,039 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.524e+02 5.877e+02 7.599e+02 9.302e+02 2.010e+03, threshold=1.520e+03, percent-clipped=8.0 +2023-04-02 07:07:27,204 INFO [train.py:903] (3/4) Epoch 16, batch 2500, loss[loss=0.1895, simple_loss=0.2814, pruned_loss=0.04877, over 19321.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.295, pruned_loss=0.07096, over 3814294.51 frames. ], batch size: 70, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:08:29,672 INFO [train.py:903] (3/4) Epoch 16, batch 2550, loss[loss=0.1779, simple_loss=0.2696, pruned_loss=0.04313, over 19688.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2954, pruned_loss=0.07081, over 3818537.07 frames. ], batch size: 53, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:09:05,175 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.330e+02 5.239e+02 6.384e+02 8.143e+02 1.987e+03, threshold=1.277e+03, percent-clipped=1.0 +2023-04-02 07:09:24,638 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 07:09:32,170 INFO [train.py:903] (3/4) Epoch 16, batch 2600, loss[loss=0.2306, simple_loss=0.3108, pruned_loss=0.07519, over 19773.00 frames. ], tot_loss[loss=0.2189, simple_loss=0.2961, pruned_loss=0.07084, over 3815334.82 frames. ], batch size: 54, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:10:35,305 INFO [train.py:903] (3/4) Epoch 16, batch 2650, loss[loss=0.2165, simple_loss=0.2964, pruned_loss=0.06832, over 19373.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2962, pruned_loss=0.07042, over 3820590.27 frames. ], batch size: 66, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:10:54,820 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 07:11:09,899 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.780e+02 5.317e+02 6.316e+02 8.001e+02 1.365e+03, threshold=1.263e+03, percent-clipped=2.0 +2023-04-02 07:11:36,917 INFO [train.py:903] (3/4) Epoch 16, batch 2700, loss[loss=0.2571, simple_loss=0.3265, pruned_loss=0.09389, over 17387.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2966, pruned_loss=0.07093, over 3811396.34 frames. ], batch size: 101, lr: 5.20e-03, grad_scale: 8.0 +2023-04-02 07:11:55,883 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.72 vs. limit=5.0 +2023-04-02 07:12:39,772 INFO [train.py:903] (3/4) Epoch 16, batch 2750, loss[loss=0.1972, simple_loss=0.2835, pruned_loss=0.0555, over 19656.00 frames. ], tot_loss[loss=0.2202, simple_loss=0.2974, pruned_loss=0.07152, over 3815365.19 frames. ], batch size: 58, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:12:52,416 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 07:13:15,058 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.123e+02 5.170e+02 6.445e+02 8.543e+02 2.677e+03, threshold=1.289e+03, percent-clipped=7.0 +2023-04-02 07:13:41,785 INFO [train.py:903] (3/4) Epoch 16, batch 2800, loss[loss=0.2251, simple_loss=0.3097, pruned_loss=0.07026, over 19518.00 frames. ], tot_loss[loss=0.2207, simple_loss=0.2979, pruned_loss=0.07179, over 3819037.63 frames. ], batch size: 64, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:14:09,832 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-04-02 07:14:11,888 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3447, 1.4668, 1.8494, 1.6177, 3.0550, 2.5911, 3.3573, 1.4553], + device='cuda:3'), covar=tensor([0.2590, 0.4453, 0.2824, 0.1979, 0.1676, 0.2076, 0.1694, 0.4299], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0603, 0.0658, 0.0458, 0.0604, 0.0511, 0.0650, 0.0518], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 07:14:44,333 INFO [train.py:903] (3/4) Epoch 16, batch 2850, loss[loss=0.2079, simple_loss=0.2952, pruned_loss=0.06029, over 19662.00 frames. ], tot_loss[loss=0.2195, simple_loss=0.2967, pruned_loss=0.07113, over 3815202.78 frames. ], batch size: 55, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:15:18,938 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.908e+02 5.066e+02 6.361e+02 8.222e+02 2.548e+03, threshold=1.272e+03, percent-clipped=4.0 +2023-04-02 07:15:46,248 INFO [train.py:903] (3/4) Epoch 16, batch 2900, loss[loss=0.2072, simple_loss=0.2955, pruned_loss=0.05941, over 19665.00 frames. ], tot_loss[loss=0.2202, simple_loss=0.2971, pruned_loss=0.07169, over 3810422.58 frames. ], batch size: 60, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:15:46,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 07:16:35,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7291, 1.3246, 1.4739, 1.4422, 3.2564, 0.9412, 2.2980, 3.6758], + device='cuda:3'), covar=tensor([0.0446, 0.2851, 0.2822, 0.1861, 0.0692, 0.2611, 0.1269, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0383, 0.0350, 0.0367, 0.0333, 0.0357, 0.0339, 0.0353, 0.0374], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:16:48,802 INFO [train.py:903] (3/4) Epoch 16, batch 2950, loss[loss=0.236, simple_loss=0.3089, pruned_loss=0.08156, over 16460.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2959, pruned_loss=0.07065, over 3824968.77 frames. ], batch size: 36, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:17:23,862 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.147e+02 4.947e+02 6.195e+02 7.724e+02 2.015e+03, threshold=1.239e+03, percent-clipped=3.0 +2023-04-02 07:17:33,621 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105406.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:17:42,529 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3392, 1.3607, 1.7331, 1.5072, 2.4864, 2.2188, 2.6449, 0.9692], + device='cuda:3'), covar=tensor([0.2369, 0.4073, 0.2451, 0.1878, 0.1477, 0.1964, 0.1431, 0.4184], + device='cuda:3'), in_proj_covar=tensor([0.0513, 0.0612, 0.0666, 0.0464, 0.0611, 0.0515, 0.0655, 0.0524], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 07:17:50,878 INFO [train.py:903] (3/4) Epoch 16, batch 3000, loss[loss=0.2482, simple_loss=0.3233, pruned_loss=0.0865, over 18721.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.2964, pruned_loss=0.07117, over 3812062.70 frames. ], batch size: 74, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:17:50,878 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 07:18:04,139 INFO [train.py:937] (3/4) Epoch 16, validation: loss=0.1725, simple_loss=0.273, pruned_loss=0.03604, over 944034.00 frames. +2023-04-02 07:18:04,140 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 07:18:07,763 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 07:19:07,115 INFO [train.py:903] (3/4) Epoch 16, batch 3050, loss[loss=0.2013, simple_loss=0.2858, pruned_loss=0.05846, over 19574.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2957, pruned_loss=0.07072, over 3814502.01 frames. ], batch size: 52, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:19:24,531 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3747, 2.1361, 1.9474, 1.8242, 1.6818, 1.8620, 0.3885, 1.2018], + device='cuda:3'), covar=tensor([0.0457, 0.0529, 0.0413, 0.0701, 0.0972, 0.0835, 0.1217, 0.0915], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0344, 0.0341, 0.0370, 0.0443, 0.0371, 0.0321, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 07:19:41,624 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.577e+02 4.913e+02 6.190e+02 8.953e+02 2.496e+03, threshold=1.238e+03, percent-clipped=7.0 +2023-04-02 07:19:49,939 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105504.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:20:11,582 INFO [train.py:903] (3/4) Epoch 16, batch 3100, loss[loss=0.195, simple_loss=0.2707, pruned_loss=0.05961, over 19778.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2952, pruned_loss=0.07037, over 3811767.48 frames. ], batch size: 46, lr: 5.19e-03, grad_scale: 8.0 +2023-04-02 07:20:11,898 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105520.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:20:47,395 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9870, 2.6698, 2.7473, 3.1358, 2.8827, 2.6319, 2.6869, 3.0220], + device='cuda:3'), covar=tensor([0.0760, 0.1510, 0.1170, 0.0885, 0.1212, 0.0438, 0.1016, 0.0507], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0357, 0.0302, 0.0249, 0.0301, 0.0252, 0.0298, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:21:13,220 INFO [train.py:903] (3/4) Epoch 16, batch 3150, loss[loss=0.1807, simple_loss=0.2541, pruned_loss=0.05364, over 18162.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2959, pruned_loss=0.07057, over 3809884.77 frames. ], batch size: 40, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:21:41,364 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 07:21:46,679 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.213e+02 5.014e+02 6.069e+02 7.545e+02 2.509e+03, threshold=1.214e+03, percent-clipped=2.0 +2023-04-02 07:22:12,921 INFO [train.py:903] (3/4) Epoch 16, batch 3200, loss[loss=0.2408, simple_loss=0.3234, pruned_loss=0.07913, over 19489.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2957, pruned_loss=0.07074, over 3810975.27 frames. ], batch size: 64, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:23:15,482 INFO [train.py:903] (3/4) Epoch 16, batch 3250, loss[loss=0.1848, simple_loss=0.2588, pruned_loss=0.05543, over 18647.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2963, pruned_loss=0.07071, over 3807197.41 frames. ], batch size: 41, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:23:50,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.698e+02 5.141e+02 5.939e+02 7.859e+02 1.653e+03, threshold=1.188e+03, percent-clipped=2.0 +2023-04-02 07:24:19,059 INFO [train.py:903] (3/4) Epoch 16, batch 3300, loss[loss=0.2588, simple_loss=0.3317, pruned_loss=0.09293, over 19491.00 frames. ], tot_loss[loss=0.2215, simple_loss=0.2984, pruned_loss=0.07231, over 3792098.84 frames. ], batch size: 64, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:24:22,587 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 07:24:55,822 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=105750.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:25:18,078 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3888, 4.0128, 2.4572, 3.5690, 0.8898, 3.7855, 3.7885, 3.8674], + device='cuda:3'), covar=tensor([0.0724, 0.1025, 0.2071, 0.0865, 0.4008, 0.0754, 0.0928, 0.1025], + device='cuda:3'), in_proj_covar=tensor([0.0471, 0.0387, 0.0463, 0.0327, 0.0393, 0.0398, 0.0395, 0.0427], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:25:21,448 INFO [train.py:903] (3/4) Epoch 16, batch 3350, loss[loss=0.1859, simple_loss=0.265, pruned_loss=0.05343, over 19484.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2954, pruned_loss=0.07065, over 3807900.36 frames. ], batch size: 49, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:25:57,715 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.211e+02 5.066e+02 6.242e+02 8.095e+02 2.652e+03, threshold=1.248e+03, percent-clipped=5.0 +2023-04-02 07:26:12,282 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105810.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:26:20,550 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105817.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:26:23,648 INFO [train.py:903] (3/4) Epoch 16, batch 3400, loss[loss=0.2508, simple_loss=0.3206, pruned_loss=0.0905, over 17446.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2952, pruned_loss=0.07078, over 3819499.25 frames. ], batch size: 101, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:27:00,024 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=105848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:27:18,424 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=105864.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:27:19,648 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=105865.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:27:26,110 INFO [train.py:903] (3/4) Epoch 16, batch 3450, loss[loss=0.2018, simple_loss=0.2764, pruned_loss=0.06358, over 19384.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2953, pruned_loss=0.07068, over 3819010.53 frames. ], batch size: 48, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:27:29,354 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 07:28:00,500 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.376e+02 5.355e+02 6.157e+02 7.690e+02 1.854e+03, threshold=1.231e+03, percent-clipped=4.0 +2023-04-02 07:28:00,875 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8795, 1.2622, 1.5501, 1.4777, 3.4226, 1.0382, 2.4189, 3.8440], + device='cuda:3'), covar=tensor([0.0469, 0.2750, 0.2637, 0.1927, 0.0731, 0.2574, 0.1183, 0.0235], + device='cuda:3'), in_proj_covar=tensor([0.0384, 0.0351, 0.0369, 0.0335, 0.0357, 0.0340, 0.0353, 0.0375], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:28:29,106 INFO [train.py:903] (3/4) Epoch 16, batch 3500, loss[loss=0.2241, simple_loss=0.3018, pruned_loss=0.07321, over 19530.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2951, pruned_loss=0.0704, over 3822218.27 frames. ], batch size: 54, lr: 5.18e-03, grad_scale: 8.0 +2023-04-02 07:29:23,595 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=105963.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:29:31,025 INFO [train.py:903] (3/4) Epoch 16, batch 3550, loss[loss=0.191, simple_loss=0.2682, pruned_loss=0.05689, over 19760.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2962, pruned_loss=0.07072, over 3821528.19 frames. ], batch size: 47, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:29:32,020 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.74 vs. limit=5.0 +2023-04-02 07:29:42,152 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=105979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:30:06,597 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.318e+02 5.018e+02 6.219e+02 7.272e+02 1.453e+03, threshold=1.244e+03, percent-clipped=3.0 +2023-04-02 07:30:21,859 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5872, 1.1490, 1.4461, 1.2350, 2.2703, 1.0065, 2.1161, 2.4941], + device='cuda:3'), covar=tensor([0.0665, 0.2604, 0.2518, 0.1537, 0.0827, 0.1911, 0.0889, 0.0415], + device='cuda:3'), in_proj_covar=tensor([0.0387, 0.0353, 0.0371, 0.0337, 0.0360, 0.0342, 0.0356, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:30:34,289 INFO [train.py:903] (3/4) Epoch 16, batch 3600, loss[loss=0.208, simple_loss=0.2897, pruned_loss=0.06317, over 18137.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2961, pruned_loss=0.07105, over 3802408.28 frames. ], batch size: 83, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:31:16,237 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.29 vs. limit=5.0 +2023-04-02 07:31:37,393 INFO [train.py:903] (3/4) Epoch 16, batch 3650, loss[loss=0.1851, simple_loss=0.2766, pruned_loss=0.04683, over 19732.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2961, pruned_loss=0.07105, over 3782048.74 frames. ], batch size: 63, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:32:09,805 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106095.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:32:13,045 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 5.371e+02 6.801e+02 8.277e+02 1.518e+03, threshold=1.360e+03, percent-clipped=5.0 +2023-04-02 07:32:42,529 INFO [train.py:903] (3/4) Epoch 16, batch 3700, loss[loss=0.2056, simple_loss=0.279, pruned_loss=0.0661, over 19409.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2957, pruned_loss=0.07052, over 3802649.44 frames. ], batch size: 48, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:32:44,125 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106121.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:33:13,573 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106146.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:33:23,614 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106154.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:33:34,065 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106161.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:33:44,495 INFO [train.py:903] (3/4) Epoch 16, batch 3750, loss[loss=0.2243, simple_loss=0.3021, pruned_loss=0.07331, over 19688.00 frames. ], tot_loss[loss=0.2181, simple_loss=0.2957, pruned_loss=0.07032, over 3798408.18 frames. ], batch size: 53, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:34:19,089 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.576e+02 4.705e+02 5.517e+02 6.973e+02 1.532e+03, threshold=1.103e+03, percent-clipped=3.0 +2023-04-02 07:34:45,466 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106219.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:34:46,184 INFO [train.py:903] (3/4) Epoch 16, batch 3800, loss[loss=0.2004, simple_loss=0.2688, pruned_loss=0.06601, over 19268.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2963, pruned_loss=0.07068, over 3803012.02 frames. ], batch size: 44, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:35:06,477 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106235.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:35:18,166 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 07:35:18,535 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106244.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:35:36,771 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106260.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:35:48,045 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106269.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:35:48,819 INFO [train.py:903] (3/4) Epoch 16, batch 3850, loss[loss=0.2026, simple_loss=0.2783, pruned_loss=0.06341, over 19840.00 frames. ], tot_loss[loss=0.2181, simple_loss=0.2956, pruned_loss=0.07034, over 3812091.29 frames. ], batch size: 52, lr: 5.17e-03, grad_scale: 8.0 +2023-04-02 07:35:56,876 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106276.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:36:09,604 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106286.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:36:23,154 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.040e+02 5.267e+02 6.364e+02 7.734e+02 1.610e+03, threshold=1.273e+03, percent-clipped=5.0 +2023-04-02 07:36:50,364 INFO [train.py:903] (3/4) Epoch 16, batch 3900, loss[loss=0.2465, simple_loss=0.3018, pruned_loss=0.09562, over 19100.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2951, pruned_loss=0.06978, over 3818889.18 frames. ], batch size: 42, lr: 5.17e-03, grad_scale: 16.0 +2023-04-02 07:37:32,215 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106354.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:37:53,351 INFO [train.py:903] (3/4) Epoch 16, batch 3950, loss[loss=0.2102, simple_loss=0.2758, pruned_loss=0.07227, over 18999.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2959, pruned_loss=0.07041, over 3815811.57 frames. ], batch size: 42, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:37:58,031 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 07:38:28,357 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.667e+02 4.931e+02 6.019e+02 7.839e+02 1.392e+03, threshold=1.204e+03, percent-clipped=3.0 +2023-04-02 07:38:54,928 INFO [train.py:903] (3/4) Epoch 16, batch 4000, loss[loss=0.1671, simple_loss=0.2439, pruned_loss=0.04516, over 19370.00 frames. ], tot_loss[loss=0.2181, simple_loss=0.2956, pruned_loss=0.07033, over 3822016.70 frames. ], batch size: 47, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:39:19,380 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106439.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:39:19,552 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5668, 1.2679, 1.5290, 1.1406, 2.1663, 0.9396, 1.9794, 2.5029], + device='cuda:3'), covar=tensor([0.0694, 0.2564, 0.2423, 0.1744, 0.0886, 0.2146, 0.1134, 0.0431], + device='cuda:3'), in_proj_covar=tensor([0.0386, 0.0350, 0.0370, 0.0337, 0.0360, 0.0340, 0.0354, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:39:45,494 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 07:39:56,769 INFO [train.py:903] (3/4) Epoch 16, batch 4050, loss[loss=0.2057, simple_loss=0.2956, pruned_loss=0.05792, over 19584.00 frames. ], tot_loss[loss=0.2175, simple_loss=0.2952, pruned_loss=0.06996, over 3825927.03 frames. ], batch size: 61, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:39:59,422 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106472.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:40:34,181 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.509e+02 4.912e+02 6.033e+02 8.007e+02 1.551e+03, threshold=1.207e+03, percent-clipped=4.0 +2023-04-02 07:40:59,898 INFO [train.py:903] (3/4) Epoch 16, batch 4100, loss[loss=0.1918, simple_loss=0.2785, pruned_loss=0.05258, over 19412.00 frames. ], tot_loss[loss=0.2176, simple_loss=0.2953, pruned_loss=0.0699, over 3826940.57 frames. ], batch size: 48, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:41:07,043 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106525.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:41:15,923 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106532.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:41:34,839 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 07:41:36,488 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106550.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:41:42,005 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106554.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:41:45,563 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106557.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:42:02,455 INFO [train.py:903] (3/4) Epoch 16, batch 4150, loss[loss=0.2314, simple_loss=0.3, pruned_loss=0.08143, over 19575.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2965, pruned_loss=0.07076, over 3809271.47 frames. ], batch size: 52, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:42:36,771 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.179e+02 4.926e+02 6.073e+02 7.216e+02 1.422e+03, threshold=1.215e+03, percent-clipped=1.0 +2023-04-02 07:43:03,953 INFO [train.py:903] (3/4) Epoch 16, batch 4200, loss[loss=0.2579, simple_loss=0.3229, pruned_loss=0.09649, over 13237.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2965, pruned_loss=0.07102, over 3804809.48 frames. ], batch size: 136, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:43:11,041 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 07:43:15,924 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:44:05,149 INFO [train.py:903] (3/4) Epoch 16, batch 4250, loss[loss=0.2535, simple_loss=0.3272, pruned_loss=0.08993, over 19690.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2953, pruned_loss=0.07027, over 3824109.60 frames. ], batch size: 58, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:44:20,464 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 07:44:32,722 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 07:44:33,381 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-02 07:44:41,261 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106698.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:44:42,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.502e+02 4.922e+02 6.114e+02 7.451e+02 1.808e+03, threshold=1.223e+03, percent-clipped=2.0 +2023-04-02 07:45:08,533 INFO [train.py:903] (3/4) Epoch 16, batch 4300, loss[loss=0.1743, simple_loss=0.2533, pruned_loss=0.04768, over 19393.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2956, pruned_loss=0.07071, over 3799805.69 frames. ], batch size: 48, lr: 5.16e-03, grad_scale: 8.0 +2023-04-02 07:45:39,051 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106745.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:45:58,136 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 07:46:11,648 INFO [train.py:903] (3/4) Epoch 16, batch 4350, loss[loss=0.217, simple_loss=0.3001, pruned_loss=0.06691, over 19517.00 frames. ], tot_loss[loss=0.2197, simple_loss=0.2964, pruned_loss=0.07154, over 3785218.59 frames. ], batch size: 54, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:46:28,690 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6350, 4.2208, 2.7274, 3.7112, 1.1105, 4.0570, 4.0056, 4.0949], + device='cuda:3'), covar=tensor([0.0636, 0.1052, 0.1971, 0.0841, 0.3875, 0.0727, 0.0835, 0.0862], + device='cuda:3'), in_proj_covar=tensor([0.0468, 0.0385, 0.0463, 0.0328, 0.0390, 0.0397, 0.0395, 0.0427], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:46:30,457 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-02 07:46:46,925 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.450e+02 5.102e+02 6.153e+02 8.101e+02 2.041e+03, threshold=1.231e+03, percent-clipped=8.0 +2023-04-02 07:47:03,099 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106810.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:47:05,391 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 07:47:06,265 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106813.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:47:09,657 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=106816.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:47:14,208 INFO [train.py:903] (3/4) Epoch 16, batch 4400, loss[loss=0.1944, simple_loss=0.2716, pruned_loss=0.05865, over 19405.00 frames. ], tot_loss[loss=0.2197, simple_loss=0.2961, pruned_loss=0.07165, over 3798789.28 frames. ], batch size: 48, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:47:33,164 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106835.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:47:36,427 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 07:47:46,749 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 07:47:48,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0206, 2.1067, 2.3764, 2.8188, 2.1202, 2.7313, 2.4551, 2.1304], + device='cuda:3'), covar=tensor([0.4032, 0.3767, 0.1725, 0.2159, 0.3803, 0.1884, 0.4273, 0.3226], + device='cuda:3'), in_proj_covar=tensor([0.0839, 0.0887, 0.0678, 0.0899, 0.0820, 0.0757, 0.0808, 0.0742], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 07:47:57,687 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6671, 1.7611, 1.9525, 2.0933, 1.6254, 2.0451, 2.0246, 1.8626], + device='cuda:3'), covar=tensor([0.3520, 0.3059, 0.1619, 0.1803, 0.2920, 0.1670, 0.4200, 0.2881], + device='cuda:3'), in_proj_covar=tensor([0.0839, 0.0887, 0.0678, 0.0899, 0.0820, 0.0757, 0.0807, 0.0741], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 07:48:17,268 INFO [train.py:903] (3/4) Epoch 16, batch 4450, loss[loss=0.2212, simple_loss=0.3018, pruned_loss=0.07034, over 19769.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2962, pruned_loss=0.07116, over 3802873.73 frames. ], batch size: 56, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:48:53,921 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.305e+02 4.966e+02 6.259e+02 8.420e+02 1.632e+03, threshold=1.252e+03, percent-clipped=6.0 +2023-04-02 07:49:17,344 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-04-02 07:49:18,989 INFO [train.py:903] (3/4) Epoch 16, batch 4500, loss[loss=0.221, simple_loss=0.303, pruned_loss=0.06953, over 19530.00 frames. ], tot_loss[loss=0.219, simple_loss=0.296, pruned_loss=0.07099, over 3815572.63 frames. ], batch size: 56, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:49:34,435 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106931.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:50:23,578 INFO [train.py:903] (3/4) Epoch 16, batch 4550, loss[loss=0.18, simple_loss=0.2609, pruned_loss=0.04955, over 19374.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2951, pruned_loss=0.07042, over 3817700.07 frames. ], batch size: 47, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:50:31,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 07:50:54,367 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 07:50:59,933 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.264e+02 4.883e+02 5.860e+02 7.136e+02 1.225e+03, threshold=1.172e+03, percent-clipped=0.0 +2023-04-02 07:51:02,643 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107001.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:51:27,832 INFO [train.py:903] (3/4) Epoch 16, batch 4600, loss[loss=0.2128, simple_loss=0.2974, pruned_loss=0.06416, over 19489.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2945, pruned_loss=0.07001, over 3821975.58 frames. ], batch size: 64, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:51:34,883 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107026.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:52:06,241 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 07:52:29,092 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107069.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:52:29,802 INFO [train.py:903] (3/4) Epoch 16, batch 4650, loss[loss=0.1993, simple_loss=0.2747, pruned_loss=0.062, over 19370.00 frames. ], tot_loss[loss=0.2189, simple_loss=0.2959, pruned_loss=0.07089, over 3799759.14 frames. ], batch size: 47, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:52:47,217 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 07:52:59,792 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 07:53:01,362 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:53:02,739 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 07:53:07,462 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.220e+02 5.459e+02 6.581e+02 8.910e+02 1.601e+03, threshold=1.316e+03, percent-clipped=6.0 +2023-04-02 07:53:31,770 INFO [train.py:903] (3/4) Epoch 16, batch 4700, loss[loss=0.2173, simple_loss=0.3073, pruned_loss=0.06367, over 19661.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2963, pruned_loss=0.07116, over 3791532.59 frames. ], batch size: 60, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:53:55,964 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 07:54:20,584 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107158.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 07:54:36,812 INFO [train.py:903] (3/4) Epoch 16, batch 4750, loss[loss=0.2043, simple_loss=0.2851, pruned_loss=0.06176, over 19574.00 frames. ], tot_loss[loss=0.219, simple_loss=0.296, pruned_loss=0.07101, over 3800885.59 frames. ], batch size: 61, lr: 5.15e-03, grad_scale: 8.0 +2023-04-02 07:54:37,123 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107170.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:54:58,937 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107187.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:55:03,366 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8949, 1.9466, 1.8826, 1.7611, 1.6166, 1.8001, 1.0484, 1.3539], + device='cuda:3'), covar=tensor([0.0452, 0.0481, 0.0329, 0.0496, 0.0726, 0.0573, 0.0876, 0.0682], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0346, 0.0344, 0.0370, 0.0446, 0.0377, 0.0325, 0.0332], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 07:55:11,308 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2119, 2.2383, 2.4286, 2.9439, 2.2305, 2.7617, 2.5511, 2.1384], + device='cuda:3'), covar=tensor([0.4225, 0.4013, 0.1795, 0.2558, 0.4279, 0.2172, 0.4321, 0.3384], + device='cuda:3'), in_proj_covar=tensor([0.0843, 0.0891, 0.0680, 0.0906, 0.0826, 0.0764, 0.0810, 0.0745], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 07:55:11,905 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.171e+02 5.545e+02 6.621e+02 8.650e+02 1.971e+03, threshold=1.324e+03, percent-clipped=6.0 +2023-04-02 07:55:30,505 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107212.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 07:55:40,203 INFO [train.py:903] (3/4) Epoch 16, batch 4800, loss[loss=0.2074, simple_loss=0.2819, pruned_loss=0.0665, over 19608.00 frames. ], tot_loss[loss=0.2198, simple_loss=0.2967, pruned_loss=0.07146, over 3804533.95 frames. ], batch size: 50, lr: 5.14e-03, grad_scale: 8.0 +2023-04-02 07:56:41,878 INFO [train.py:903] (3/4) Epoch 16, batch 4850, loss[loss=0.2458, simple_loss=0.3156, pruned_loss=0.08797, over 18131.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2972, pruned_loss=0.07172, over 3804683.52 frames. ], batch size: 83, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 07:57:07,076 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 07:57:19,837 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.329e+02 4.930e+02 6.295e+02 8.478e+02 1.665e+03, threshold=1.259e+03, percent-clipped=1.0 +2023-04-02 07:57:27,492 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 07:57:32,767 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 07:57:32,791 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 07:57:43,095 INFO [train.py:903] (3/4) Epoch 16, batch 4900, loss[loss=0.2209, simple_loss=0.3043, pruned_loss=0.06876, over 19392.00 frames. ], tot_loss[loss=0.2204, simple_loss=0.2972, pruned_loss=0.07186, over 3794437.04 frames. ], batch size: 70, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 07:57:43,111 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 07:57:49,619 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.01 vs. limit=5.0 +2023-04-02 07:58:04,169 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 07:58:46,431 INFO [train.py:903] (3/4) Epoch 16, batch 4950, loss[loss=0.2124, simple_loss=0.3, pruned_loss=0.06242, over 19794.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.2962, pruned_loss=0.07148, over 3816340.44 frames. ], batch size: 56, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 07:59:04,243 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 07:59:17,589 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4311, 1.0194, 1.3133, 1.3923, 2.7717, 0.9902, 2.2399, 3.2493], + device='cuda:3'), covar=tensor([0.0682, 0.3614, 0.3557, 0.2225, 0.1218, 0.3045, 0.1419, 0.0456], + device='cuda:3'), in_proj_covar=tensor([0.0385, 0.0351, 0.0374, 0.0334, 0.0359, 0.0340, 0.0356, 0.0376], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 07:59:22,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.794e+02 5.680e+02 6.678e+02 8.404e+02 2.020e+03, threshold=1.336e+03, percent-clipped=4.0 +2023-04-02 07:59:27,620 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 07:59:48,957 INFO [train.py:903] (3/4) Epoch 16, batch 5000, loss[loss=0.2413, simple_loss=0.3142, pruned_loss=0.08419, over 19070.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2955, pruned_loss=0.07088, over 3823413.79 frames. ], batch size: 69, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 07:59:58,954 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 08:00:09,006 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 08:00:50,465 INFO [train.py:903] (3/4) Epoch 16, batch 5050, loss[loss=0.2252, simple_loss=0.3064, pruned_loss=0.07204, over 19528.00 frames. ], tot_loss[loss=0.219, simple_loss=0.296, pruned_loss=0.07101, over 3830092.85 frames. ], batch size: 54, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 08:01:24,839 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9360, 1.1961, 1.5867, 0.5928, 2.1930, 2.4558, 2.1800, 2.6552], + device='cuda:3'), covar=tensor([0.1565, 0.3606, 0.3048, 0.2524, 0.0539, 0.0279, 0.0321, 0.0321], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0306, 0.0335, 0.0257, 0.0229, 0.0174, 0.0210, 0.0230], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:01:27,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.607e+02 5.470e+02 6.454e+02 8.047e+02 2.188e+03, threshold=1.291e+03, percent-clipped=2.0 +2023-04-02 08:01:27,918 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 08:01:30,407 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=107502.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 08:01:44,972 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=107514.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:01:51,915 INFO [train.py:903] (3/4) Epoch 16, batch 5100, loss[loss=0.2236, simple_loss=0.3046, pruned_loss=0.07129, over 18716.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2959, pruned_loss=0.07136, over 3829925.79 frames. ], batch size: 74, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 08:02:02,196 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107528.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:02:04,286 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 08:02:08,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 08:02:13,328 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 08:02:54,727 INFO [train.py:903] (3/4) Epoch 16, batch 5150, loss[loss=0.2011, simple_loss=0.2818, pruned_loss=0.06026, over 19784.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2957, pruned_loss=0.07132, over 3827703.08 frames. ], batch size: 56, lr: 5.14e-03, grad_scale: 4.0 +2023-04-02 08:03:08,976 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 08:03:31,801 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.461e+02 4.950e+02 6.087e+02 7.766e+02 1.818e+03, threshold=1.217e+03, percent-clipped=6.0 +2023-04-02 08:03:43,250 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 08:03:54,459 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107617.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:03:57,185 INFO [train.py:903] (3/4) Epoch 16, batch 5200, loss[loss=0.2503, simple_loss=0.3223, pruned_loss=0.08916, over 19567.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2956, pruned_loss=0.07097, over 3829602.16 frames. ], batch size: 61, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:04:08,674 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107629.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:04:09,569 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 08:04:09,732 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.2803, 3.9507, 3.0946, 3.4435, 1.7042, 3.7387, 3.6936, 3.8835], + device='cuda:3'), covar=tensor([0.0731, 0.1039, 0.1832, 0.0952, 0.3224, 0.0886, 0.0934, 0.1333], + device='cuda:3'), in_proj_covar=tensor([0.0470, 0.0388, 0.0466, 0.0331, 0.0393, 0.0402, 0.0400, 0.0429], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:04:25,017 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.61 vs. limit=2.0 +2023-04-02 08:04:55,318 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 08:04:59,577 INFO [train.py:903] (3/4) Epoch 16, batch 5250, loss[loss=0.204, simple_loss=0.2732, pruned_loss=0.06742, over 19741.00 frames. ], tot_loss[loss=0.2195, simple_loss=0.2962, pruned_loss=0.0714, over 3828538.55 frames. ], batch size: 51, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:05:07,732 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107677.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:05:24,224 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.55 vs. limit=2.0 +2023-04-02 08:05:36,449 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.729e+02 5.464e+02 6.488e+02 8.647e+02 1.622e+03, threshold=1.298e+03, percent-clipped=8.0 +2023-04-02 08:05:37,904 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1736, 2.1154, 1.9801, 1.7454, 1.6563, 1.8180, 0.7707, 1.2900], + device='cuda:3'), covar=tensor([0.0548, 0.0532, 0.0362, 0.0693, 0.0917, 0.0760, 0.1031, 0.0846], + device='cuda:3'), in_proj_covar=tensor([0.0346, 0.0340, 0.0340, 0.0366, 0.0440, 0.0371, 0.0320, 0.0328], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:05:50,343 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4367, 2.2460, 2.2292, 2.5900, 2.3756, 2.0950, 1.9856, 2.4081], + device='cuda:3'), covar=tensor([0.0869, 0.1428, 0.1156, 0.0850, 0.1174, 0.0480, 0.1199, 0.0586], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0352, 0.0301, 0.0243, 0.0298, 0.0248, 0.0290, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:06:00,484 INFO [train.py:903] (3/4) Epoch 16, batch 5300, loss[loss=0.2062, simple_loss=0.2822, pruned_loss=0.06514, over 19609.00 frames. ], tot_loss[loss=0.2202, simple_loss=0.2966, pruned_loss=0.07188, over 3818799.27 frames. ], batch size: 50, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:06:19,332 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 08:06:45,194 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107755.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:07:03,291 INFO [train.py:903] (3/4) Epoch 16, batch 5350, loss[loss=0.2075, simple_loss=0.2751, pruned_loss=0.06999, over 19711.00 frames. ], tot_loss[loss=0.2197, simple_loss=0.2962, pruned_loss=0.07164, over 3818556.98 frames. ], batch size: 45, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:07:37,249 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 08:07:40,607 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.374e+02 4.506e+02 5.884e+02 6.805e+02 1.610e+03, threshold=1.177e+03, percent-clipped=2.0 +2023-04-02 08:08:06,496 INFO [train.py:903] (3/4) Epoch 16, batch 5400, loss[loss=0.2116, simple_loss=0.2838, pruned_loss=0.06976, over 19385.00 frames. ], tot_loss[loss=0.2197, simple_loss=0.2962, pruned_loss=0.07161, over 3812833.43 frames. ], batch size: 47, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:08:41,526 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3981, 1.4777, 1.7537, 1.6369, 3.2366, 2.7424, 3.5154, 1.6862], + device='cuda:3'), covar=tensor([0.2488, 0.4252, 0.2723, 0.1923, 0.1426, 0.1844, 0.1524, 0.3739], + device='cuda:3'), in_proj_covar=tensor([0.0510, 0.0611, 0.0663, 0.0463, 0.0609, 0.0512, 0.0651, 0.0520], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:09:08,397 INFO [train.py:903] (3/4) Epoch 16, batch 5450, loss[loss=0.2513, simple_loss=0.328, pruned_loss=0.08733, over 18926.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2949, pruned_loss=0.07091, over 3818428.90 frames. ], batch size: 75, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:09:10,589 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=107872.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:09:11,984 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107873.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:09:26,840 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107885.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:09:44,605 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107898.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 08:09:46,448 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.026e+02 5.228e+02 6.592e+02 8.752e+02 1.860e+03, threshold=1.318e+03, percent-clipped=11.0 +2023-04-02 08:09:58,385 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107910.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:10:10,032 INFO [train.py:903] (3/4) Epoch 16, batch 5500, loss[loss=0.1867, simple_loss=0.2538, pruned_loss=0.05975, over 19102.00 frames. ], tot_loss[loss=0.2189, simple_loss=0.2949, pruned_loss=0.07141, over 3800420.80 frames. ], batch size: 42, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:10:18,750 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4444, 1.3496, 1.3083, 1.6989, 1.4490, 1.7034, 1.7557, 1.5461], + device='cuda:3'), covar=tensor([0.0857, 0.0984, 0.1129, 0.0799, 0.0826, 0.0755, 0.0781, 0.0719], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0221, 0.0223, 0.0245, 0.0226, 0.0207, 0.0189, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 08:10:34,930 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 08:11:13,118 INFO [train.py:903] (3/4) Epoch 16, batch 5550, loss[loss=0.2192, simple_loss=0.2881, pruned_loss=0.07509, over 19405.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2953, pruned_loss=0.07146, over 3799343.45 frames. ], batch size: 48, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:11:18,629 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 08:11:33,206 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6143, 4.1524, 2.7000, 3.7215, 0.9695, 4.0385, 3.9730, 4.1025], + device='cuda:3'), covar=tensor([0.0667, 0.1119, 0.1915, 0.0785, 0.4084, 0.0737, 0.0859, 0.1084], + device='cuda:3'), in_proj_covar=tensor([0.0470, 0.0385, 0.0463, 0.0329, 0.0392, 0.0402, 0.0399, 0.0429], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:11:34,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107987.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:11:50,037 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.970e+02 4.933e+02 6.287e+02 7.608e+02 2.106e+03, threshold=1.257e+03, percent-clipped=3.0 +2023-04-02 08:12:10,700 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 08:12:17,773 INFO [train.py:903] (3/4) Epoch 16, batch 5600, loss[loss=0.1949, simple_loss=0.2775, pruned_loss=0.05619, over 19685.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2954, pruned_loss=0.07091, over 3826374.33 frames. ], batch size: 60, lr: 5.13e-03, grad_scale: 8.0 +2023-04-02 08:12:19,214 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=108021.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:13:00,815 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0400, 1.7676, 1.6163, 1.9069, 1.8050, 1.5370, 1.5226, 1.8578], + device='cuda:3'), covar=tensor([0.0936, 0.1558, 0.1524, 0.0999, 0.1294, 0.0710, 0.1469, 0.0753], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0352, 0.0300, 0.0243, 0.0297, 0.0247, 0.0290, 0.0245], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:13:17,748 INFO [train.py:903] (3/4) Epoch 16, batch 5650, loss[loss=0.2319, simple_loss=0.3072, pruned_loss=0.07834, over 19675.00 frames. ], tot_loss[loss=0.2198, simple_loss=0.2962, pruned_loss=0.07167, over 3829059.42 frames. ], batch size: 55, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:13:31,909 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9528, 4.4562, 2.7739, 3.9360, 0.9501, 4.3371, 4.2738, 4.3974], + device='cuda:3'), covar=tensor([0.0565, 0.1032, 0.1979, 0.0776, 0.4305, 0.0699, 0.0879, 0.0950], + device='cuda:3'), in_proj_covar=tensor([0.0470, 0.0387, 0.0464, 0.0330, 0.0394, 0.0403, 0.0402, 0.0430], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:13:55,637 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=108099.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:13:56,647 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.333e+02 5.688e+02 6.420e+02 8.034e+02 1.662e+03, threshold=1.284e+03, percent-clipped=4.0 +2023-04-02 08:14:03,634 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 08:14:20,998 INFO [train.py:903] (3/4) Epoch 16, batch 5700, loss[loss=0.2252, simple_loss=0.299, pruned_loss=0.0757, over 19412.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2948, pruned_loss=0.07053, over 3831643.34 frames. ], batch size: 70, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:14:42,905 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108136.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:15:22,685 INFO [train.py:903] (3/4) Epoch 16, batch 5750, loss[loss=0.2307, simple_loss=0.3169, pruned_loss=0.07224, over 18185.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2959, pruned_loss=0.0708, over 3836484.86 frames. ], batch size: 83, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:15:22,716 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 08:15:33,015 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 08:15:36,726 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 08:16:00,581 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.754e+02 4.860e+02 6.382e+02 7.922e+02 1.732e+03, threshold=1.276e+03, percent-clipped=4.0 +2023-04-02 08:16:19,217 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108214.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:16:26,751 INFO [train.py:903] (3/4) Epoch 16, batch 5800, loss[loss=0.2377, simple_loss=0.3141, pruned_loss=0.0807, over 12728.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2958, pruned_loss=0.07071, over 3810770.37 frames. ], batch size: 136, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:16:54,106 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108243.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:17:25,879 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108268.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:17:27,725 INFO [train.py:903] (3/4) Epoch 16, batch 5850, loss[loss=0.212, simple_loss=0.2898, pruned_loss=0.06715, over 19465.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2953, pruned_loss=0.07089, over 3802224.09 frames. ], batch size: 49, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:18:05,295 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.738e+02 5.463e+02 6.888e+02 8.802e+02 1.964e+03, threshold=1.378e+03, percent-clipped=5.0 +2023-04-02 08:18:28,343 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 08:18:29,527 INFO [train.py:903] (3/4) Epoch 16, batch 5900, loss[loss=0.2111, simple_loss=0.3033, pruned_loss=0.05944, over 19665.00 frames. ], tot_loss[loss=0.2176, simple_loss=0.2944, pruned_loss=0.07044, over 3810463.50 frames. ], batch size: 58, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:18:52,012 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 08:18:59,411 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108343.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:19:30,982 INFO [train.py:903] (3/4) Epoch 16, batch 5950, loss[loss=0.215, simple_loss=0.2974, pruned_loss=0.06632, over 19348.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2954, pruned_loss=0.07084, over 3815864.88 frames. ], batch size: 70, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:19:59,913 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108392.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:20:09,622 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.801e+02 5.248e+02 6.439e+02 7.965e+02 2.252e+03, threshold=1.288e+03, percent-clipped=3.0 +2023-04-02 08:20:12,673 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.03 vs. limit=5.0 +2023-04-02 08:20:30,475 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108417.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:20:35,475 INFO [train.py:903] (3/4) Epoch 16, batch 6000, loss[loss=0.2429, simple_loss=0.316, pruned_loss=0.08488, over 19477.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.2939, pruned_loss=0.06993, over 3833895.78 frames. ], batch size: 49, lr: 5.12e-03, grad_scale: 8.0 +2023-04-02 08:20:35,476 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 08:20:47,897 INFO [train.py:937] (3/4) Epoch 16, validation: loss=0.1716, simple_loss=0.2723, pruned_loss=0.03545, over 944034.00 frames. +2023-04-02 08:20:47,899 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 08:20:49,621 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3880, 1.4431, 1.7724, 1.6034, 2.8687, 2.3417, 3.0745, 1.2509], + device='cuda:3'), covar=tensor([0.2268, 0.3962, 0.2549, 0.1768, 0.1286, 0.1842, 0.1224, 0.3846], + device='cuda:3'), in_proj_covar=tensor([0.0507, 0.0607, 0.0660, 0.0460, 0.0603, 0.0509, 0.0645, 0.0517], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:21:25,619 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108450.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:21:51,712 INFO [train.py:903] (3/4) Epoch 16, batch 6050, loss[loss=0.2173, simple_loss=0.3002, pruned_loss=0.06724, over 19688.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2953, pruned_loss=0.07076, over 3826944.53 frames. ], batch size: 59, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:21:52,195 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108470.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:21:52,470 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 08:22:22,217 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108495.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:22:28,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.284e+02 5.272e+02 6.402e+02 8.353e+02 1.883e+03, threshold=1.280e+03, percent-clipped=4.0 +2023-04-02 08:22:41,573 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9590, 1.9279, 1.7391, 1.5816, 1.4742, 1.6034, 0.3827, 0.8764], + device='cuda:3'), covar=tensor([0.0524, 0.0553, 0.0391, 0.0650, 0.1141, 0.0671, 0.1137, 0.0978], + device='cuda:3'), in_proj_covar=tensor([0.0348, 0.0345, 0.0343, 0.0369, 0.0445, 0.0374, 0.0323, 0.0332], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:22:53,808 INFO [train.py:903] (3/4) Epoch 16, batch 6100, loss[loss=0.1889, simple_loss=0.2724, pruned_loss=0.05272, over 19419.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2959, pruned_loss=0.07137, over 3817672.57 frames. ], batch size: 48, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:23:56,095 INFO [train.py:903] (3/4) Epoch 16, batch 6150, loss[loss=0.2203, simple_loss=0.2946, pruned_loss=0.07299, over 18089.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.2969, pruned_loss=0.07145, over 3829716.67 frames. ], batch size: 83, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:24:00,399 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-02 08:24:26,236 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 08:24:28,846 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108595.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:24:35,759 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.970e+02 4.845e+02 6.012e+02 7.583e+02 1.796e+03, threshold=1.202e+03, percent-clipped=3.0 +2023-04-02 08:24:58,817 INFO [train.py:903] (3/4) Epoch 16, batch 6200, loss[loss=0.2518, simple_loss=0.3248, pruned_loss=0.08937, over 17261.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2964, pruned_loss=0.07064, over 3829513.80 frames. ], batch size: 101, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:25:10,033 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4361, 2.2137, 2.3264, 2.5574, 2.3352, 2.1958, 2.2376, 2.4490], + device='cuda:3'), covar=tensor([0.0747, 0.1226, 0.0910, 0.0644, 0.0996, 0.0423, 0.0979, 0.0485], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0353, 0.0301, 0.0243, 0.0299, 0.0249, 0.0293, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:26:02,207 INFO [train.py:903] (3/4) Epoch 16, batch 6250, loss[loss=0.202, simple_loss=0.2812, pruned_loss=0.06137, over 19578.00 frames. ], tot_loss[loss=0.2167, simple_loss=0.2946, pruned_loss=0.06938, over 3817691.85 frames. ], batch size: 52, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:26:22,919 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=108687.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:26:34,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 08:26:40,072 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.877e+02 4.966e+02 6.026e+02 7.805e+02 1.726e+03, threshold=1.205e+03, percent-clipped=3.0 +2023-04-02 08:27:04,711 INFO [train.py:903] (3/4) Epoch 16, batch 6300, loss[loss=0.2112, simple_loss=0.2965, pruned_loss=0.06296, over 19731.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.2945, pruned_loss=0.06961, over 3831505.00 frames. ], batch size: 63, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:27:07,315 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4748, 1.4965, 1.6609, 1.5967, 2.3086, 1.9870, 2.3869, 1.3252], + device='cuda:3'), covar=tensor([0.1850, 0.3303, 0.2134, 0.1581, 0.1235, 0.1760, 0.1074, 0.3712], + device='cuda:3'), in_proj_covar=tensor([0.0506, 0.0608, 0.0661, 0.0460, 0.0604, 0.0511, 0.0648, 0.0519], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:27:09,336 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8199, 1.6239, 1.5762, 2.3589, 1.8377, 2.1359, 2.1687, 1.8864], + device='cuda:3'), covar=tensor([0.0763, 0.0972, 0.1043, 0.0694, 0.0789, 0.0698, 0.0787, 0.0656], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0222, 0.0225, 0.0246, 0.0227, 0.0207, 0.0189, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 08:27:56,260 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7922, 1.6519, 1.5819, 2.2764, 1.6347, 2.0451, 2.1143, 1.8904], + device='cuda:3'), covar=tensor([0.0795, 0.0946, 0.1018, 0.0817, 0.0885, 0.0762, 0.0839, 0.0666], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0221, 0.0224, 0.0245, 0.0226, 0.0206, 0.0189, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 08:27:59,208 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.51 vs. limit=5.0 +2023-04-02 08:28:06,338 INFO [train.py:903] (3/4) Epoch 16, batch 6350, loss[loss=0.2801, simple_loss=0.3514, pruned_loss=0.1044, over 17361.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.2936, pruned_loss=0.06947, over 3833746.13 frames. ], batch size: 101, lr: 5.11e-03, grad_scale: 4.0 +2023-04-02 08:28:38,959 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=108794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:28:47,227 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.196e+02 5.125e+02 6.267e+02 8.166e+02 1.468e+03, threshold=1.253e+03, percent-clipped=8.0 +2023-04-02 08:28:48,979 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108802.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:28:56,975 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108809.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:29:09,790 INFO [train.py:903] (3/4) Epoch 16, batch 6400, loss[loss=0.218, simple_loss=0.3011, pruned_loss=0.06747, over 19181.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.2944, pruned_loss=0.06974, over 3839984.43 frames. ], batch size: 69, lr: 5.11e-03, grad_scale: 8.0 +2023-04-02 08:29:17,119 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3094, 1.4054, 1.5145, 1.4747, 1.6662, 1.8530, 1.6616, 0.4867], + device='cuda:3'), covar=tensor([0.2331, 0.3951, 0.2497, 0.1853, 0.1642, 0.2110, 0.1443, 0.4365], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0612, 0.0664, 0.0464, 0.0606, 0.0513, 0.0651, 0.0521], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:29:56,413 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3167, 1.3596, 1.5085, 1.4826, 1.6759, 1.8306, 1.6232, 0.5106], + device='cuda:3'), covar=tensor([0.2301, 0.3985, 0.2504, 0.1877, 0.1625, 0.2198, 0.1517, 0.4410], + device='cuda:3'), in_proj_covar=tensor([0.0507, 0.0611, 0.0663, 0.0463, 0.0605, 0.0512, 0.0650, 0.0521], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:30:14,205 INFO [train.py:903] (3/4) Epoch 16, batch 6450, loss[loss=0.2503, simple_loss=0.3213, pruned_loss=0.08966, over 19737.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2949, pruned_loss=0.06976, over 3844957.50 frames. ], batch size: 63, lr: 5.11e-03, grad_scale: 8.0 +2023-04-02 08:30:40,421 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-04-02 08:30:52,172 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.276e+02 5.109e+02 6.250e+02 7.655e+02 1.750e+03, threshold=1.250e+03, percent-clipped=4.0 +2023-04-02 08:30:59,839 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 08:31:00,314 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 08:31:04,101 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108909.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:31:18,889 INFO [train.py:903] (3/4) Epoch 16, batch 6500, loss[loss=0.1989, simple_loss=0.2721, pruned_loss=0.06287, over 14798.00 frames. ], tot_loss[loss=0.216, simple_loss=0.294, pruned_loss=0.06899, over 3842030.27 frames. ], batch size: 32, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:31:24,478 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 08:31:37,676 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8987, 0.8664, 0.8761, 0.9562, 0.7905, 0.9694, 0.9765, 0.9065], + device='cuda:3'), covar=tensor([0.0647, 0.0712, 0.0775, 0.0550, 0.0747, 0.0586, 0.0667, 0.0570], + device='cuda:3'), in_proj_covar=tensor([0.0206, 0.0219, 0.0222, 0.0243, 0.0224, 0.0205, 0.0188, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 08:31:40,959 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=108939.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:32:20,051 INFO [train.py:903] (3/4) Epoch 16, batch 6550, loss[loss=0.2774, simple_loss=0.3402, pruned_loss=0.1073, over 19332.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2937, pruned_loss=0.06875, over 3845788.43 frames. ], batch size: 66, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:32:27,823 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-02 08:32:58,929 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.314e+02 4.624e+02 5.967e+02 6.821e+02 1.232e+03, threshold=1.193e+03, percent-clipped=0.0 +2023-04-02 08:33:21,265 INFO [train.py:903] (3/4) Epoch 16, batch 6600, loss[loss=0.204, simple_loss=0.2872, pruned_loss=0.06047, over 17321.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2932, pruned_loss=0.06888, over 3835873.43 frames. ], batch size: 101, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:33:53,960 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9901, 3.4194, 1.9666, 1.9912, 3.0114, 1.6045, 1.2161, 2.1248], + device='cuda:3'), covar=tensor([0.1410, 0.0548, 0.1121, 0.0846, 0.0516, 0.1262, 0.1057, 0.0730], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0305, 0.0329, 0.0254, 0.0240, 0.0329, 0.0287, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:34:03,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109054.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:34:05,065 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109055.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:34:08,550 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109058.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:34:22,639 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.61 vs. limit=5.0 +2023-04-02 08:34:22,860 INFO [train.py:903] (3/4) Epoch 16, batch 6650, loss[loss=0.2375, simple_loss=0.313, pruned_loss=0.08102, over 19678.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.2939, pruned_loss=0.0695, over 3834295.59 frames. ], batch size: 59, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:34:41,492 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109083.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:35:01,848 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.091e+02 5.420e+02 6.464e+02 8.289e+02 2.034e+03, threshold=1.293e+03, percent-clipped=5.0 +2023-04-02 08:35:18,230 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 08:35:27,581 INFO [train.py:903] (3/4) Epoch 16, batch 6700, loss[loss=0.2211, simple_loss=0.3023, pruned_loss=0.06988, over 18186.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2933, pruned_loss=0.06929, over 3831398.12 frames. ], batch size: 83, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:36:06,063 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109153.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:36:19,862 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:36:25,877 INFO [train.py:903] (3/4) Epoch 16, batch 6750, loss[loss=0.2137, simple_loss=0.292, pruned_loss=0.06772, over 19671.00 frames. ], tot_loss[loss=0.217, simple_loss=0.2941, pruned_loss=0.06992, over 3818039.10 frames. ], batch size: 58, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:36:29,588 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6066, 1.4633, 1.4317, 2.2091, 1.7240, 1.8672, 2.1304, 1.7142], + device='cuda:3'), covar=tensor([0.0839, 0.1001, 0.1097, 0.0794, 0.0882, 0.0782, 0.0783, 0.0744], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0221, 0.0224, 0.0244, 0.0225, 0.0206, 0.0189, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 08:36:50,461 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109190.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:37:03,547 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.450e+02 4.897e+02 5.778e+02 7.062e+02 1.289e+03, threshold=1.156e+03, percent-clipped=0.0 +2023-04-02 08:37:15,727 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109212.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:37:22,665 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-02 08:37:24,165 INFO [train.py:903] (3/4) Epoch 16, batch 6800, loss[loss=0.2043, simple_loss=0.2804, pruned_loss=0.06407, over 19666.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2953, pruned_loss=0.07039, over 3821052.82 frames. ], batch size: 53, lr: 5.10e-03, grad_scale: 8.0 +2023-04-02 08:37:42,841 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-04-02 08:38:09,493 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 08:38:09,930 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 08:38:13,448 INFO [train.py:903] (3/4) Epoch 17, batch 0, loss[loss=0.2405, simple_loss=0.3136, pruned_loss=0.08367, over 19665.00 frames. ], tot_loss[loss=0.2405, simple_loss=0.3136, pruned_loss=0.08367, over 19665.00 frames. ], batch size: 60, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:38:13,449 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 08:38:23,516 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2889, 1.2866, 1.5915, 1.4598, 2.2446, 1.8675, 2.2664, 1.1734], + device='cuda:3'), covar=tensor([0.2412, 0.4356, 0.2684, 0.1904, 0.1434, 0.2242, 0.1238, 0.4418], + device='cuda:3'), in_proj_covar=tensor([0.0509, 0.0612, 0.0664, 0.0460, 0.0606, 0.0512, 0.0650, 0.0520], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:38:26,004 INFO [train.py:937] (3/4) Epoch 17, validation: loss=0.1721, simple_loss=0.2728, pruned_loss=0.03571, over 944034.00 frames. +2023-04-02 08:38:26,005 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 08:38:39,464 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 08:38:51,296 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109268.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:39:29,170 INFO [train.py:903] (3/4) Epoch 17, batch 50, loss[loss=0.2389, simple_loss=0.3166, pruned_loss=0.08061, over 19789.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2959, pruned_loss=0.07235, over 840860.20 frames. ], batch size: 56, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:39:32,717 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.530e+02 5.319e+02 6.338e+02 7.961e+02 1.981e+03, threshold=1.268e+03, percent-clipped=4.0 +2023-04-02 08:39:43,545 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109310.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:39:45,680 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109312.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:39:59,842 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 08:40:13,445 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109335.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:40:14,435 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109336.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:40:28,818 INFO [train.py:903] (3/4) Epoch 17, batch 100, loss[loss=0.1965, simple_loss=0.2755, pruned_loss=0.05877, over 19390.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2953, pruned_loss=0.0714, over 1491684.65 frames. ], batch size: 48, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:40:36,771 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 08:40:53,546 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7501, 1.7779, 1.6199, 1.4172, 1.4238, 1.4088, 0.1840, 0.6698], + device='cuda:3'), covar=tensor([0.0565, 0.0555, 0.0337, 0.0493, 0.1023, 0.0637, 0.1028, 0.0892], + device='cuda:3'), in_proj_covar=tensor([0.0345, 0.0346, 0.0343, 0.0368, 0.0444, 0.0374, 0.0323, 0.0330], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:41:29,300 INFO [train.py:903] (3/4) Epoch 17, batch 150, loss[loss=0.207, simple_loss=0.295, pruned_loss=0.05949, over 18362.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.2962, pruned_loss=0.07129, over 2011726.65 frames. ], batch size: 83, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:41:30,559 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109399.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:41:32,676 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 5.062e+02 6.078e+02 8.331e+02 1.364e+03, threshold=1.216e+03, percent-clipped=3.0 +2023-04-02 08:41:34,552 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.06 vs. limit=5.0 +2023-04-02 08:41:44,717 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.10 vs. limit=5.0 +2023-04-02 08:42:07,844 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4195, 1.5601, 2.1348, 1.7280, 3.2074, 2.5080, 3.4540, 1.5121], + device='cuda:3'), covar=tensor([0.2385, 0.4210, 0.2528, 0.1843, 0.1439, 0.2001, 0.1546, 0.4074], + device='cuda:3'), in_proj_covar=tensor([0.0505, 0.0607, 0.0659, 0.0458, 0.0601, 0.0508, 0.0646, 0.0517], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:42:22,302 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 08:42:29,333 INFO [train.py:903] (3/4) Epoch 17, batch 200, loss[loss=0.2464, simple_loss=0.321, pruned_loss=0.08587, over 19556.00 frames. ], tot_loss[loss=0.2199, simple_loss=0.297, pruned_loss=0.07144, over 2421661.01 frames. ], batch size: 61, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:43:25,310 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3844, 2.1596, 1.5776, 1.3970, 1.9882, 1.1953, 1.2039, 1.8082], + device='cuda:3'), covar=tensor([0.0897, 0.0667, 0.0969, 0.0735, 0.0482, 0.1221, 0.0715, 0.0421], + device='cuda:3'), in_proj_covar=tensor([0.0293, 0.0305, 0.0326, 0.0254, 0.0240, 0.0326, 0.0285, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:43:32,827 INFO [train.py:903] (3/4) Epoch 17, batch 250, loss[loss=0.2123, simple_loss=0.2955, pruned_loss=0.06451, over 19528.00 frames. ], tot_loss[loss=0.2209, simple_loss=0.2976, pruned_loss=0.07209, over 2744636.88 frames. ], batch size: 54, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:43:36,242 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.403e+02 5.064e+02 6.047e+02 7.271e+02 1.663e+03, threshold=1.209e+03, percent-clipped=2.0 +2023-04-02 08:43:52,402 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109514.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:44:04,251 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109524.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:44:35,520 INFO [train.py:903] (3/4) Epoch 17, batch 300, loss[loss=0.1816, simple_loss=0.2566, pruned_loss=0.05328, over 16006.00 frames. ], tot_loss[loss=0.2197, simple_loss=0.2966, pruned_loss=0.07135, over 2974020.31 frames. ], batch size: 35, lr: 4.94e-03, grad_scale: 8.0 +2023-04-02 08:44:37,171 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109549.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:44:45,081 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109556.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:44:45,555 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 08:45:36,080 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109597.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:45:36,900 INFO [train.py:903] (3/4) Epoch 17, batch 350, loss[loss=0.2452, simple_loss=0.3161, pruned_loss=0.08719, over 19533.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2954, pruned_loss=0.07047, over 3172159.17 frames. ], batch size: 56, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:45:38,097 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 08:45:40,567 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.272e+02 4.783e+02 5.858e+02 7.548e+02 1.929e+03, threshold=1.172e+03, percent-clipped=3.0 +2023-04-02 08:46:33,322 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0044, 4.3327, 4.7402, 4.7171, 1.6532, 4.3775, 3.7090, 4.4158], + device='cuda:3'), covar=tensor([0.1557, 0.0898, 0.0536, 0.0635, 0.6069, 0.0819, 0.0715, 0.1047], + device='cuda:3'), in_proj_covar=tensor([0.0731, 0.0677, 0.0876, 0.0761, 0.0785, 0.0625, 0.0525, 0.0805], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 08:46:38,843 INFO [train.py:903] (3/4) Epoch 17, batch 400, loss[loss=0.2189, simple_loss=0.3023, pruned_loss=0.06769, over 19521.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2955, pruned_loss=0.07015, over 3323804.31 frames. ], batch size: 54, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:46:49,103 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109656.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:47:09,937 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109671.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:47:16,706 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109677.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:47:20,165 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109680.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:47:40,317 INFO [train.py:903] (3/4) Epoch 17, batch 450, loss[loss=0.1806, simple_loss=0.2634, pruned_loss=0.04893, over 19619.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2952, pruned_loss=0.0696, over 3432268.11 frames. ], batch size: 50, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:47:44,680 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.768e+02 5.202e+02 6.539e+02 8.058e+02 1.631e+03, threshold=1.308e+03, percent-clipped=6.0 +2023-04-02 08:48:12,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 08:48:13,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 08:48:44,992 INFO [train.py:903] (3/4) Epoch 17, batch 500, loss[loss=0.1966, simple_loss=0.281, pruned_loss=0.0561, over 19527.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2963, pruned_loss=0.07039, over 3518353.70 frames. ], batch size: 54, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:48:53,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7221, 4.0418, 4.3287, 4.3377, 1.8468, 4.0472, 3.5313, 4.0424], + device='cuda:3'), covar=tensor([0.1346, 0.1451, 0.0586, 0.0662, 0.5337, 0.0965, 0.0657, 0.1072], + device='cuda:3'), in_proj_covar=tensor([0.0734, 0.0676, 0.0876, 0.0761, 0.0784, 0.0625, 0.0525, 0.0809], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 08:48:59,863 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-04-02 08:49:11,994 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109770.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 08:49:12,930 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109771.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:49:16,865 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.74 vs. limit=5.0 +2023-04-02 08:49:44,307 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109795.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:49:44,354 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109795.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 08:49:47,270 INFO [train.py:903] (3/4) Epoch 17, batch 550, loss[loss=0.2422, simple_loss=0.3229, pruned_loss=0.08072, over 19466.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2959, pruned_loss=0.07077, over 3583418.65 frames. ], batch size: 64, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:49:50,690 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.960e+02 5.345e+02 7.483e+02 9.617e+02 2.288e+03, threshold=1.497e+03, percent-clipped=10.0 +2023-04-02 08:50:48,373 INFO [train.py:903] (3/4) Epoch 17, batch 600, loss[loss=0.2181, simple_loss=0.2953, pruned_loss=0.0705, over 19285.00 frames. ], tot_loss[loss=0.2202, simple_loss=0.2972, pruned_loss=0.07164, over 3621789.10 frames. ], batch size: 66, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:51:27,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 08:51:35,306 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.14 vs. limit=5.0 +2023-04-02 08:51:49,477 INFO [train.py:903] (3/4) Epoch 17, batch 650, loss[loss=0.2078, simple_loss=0.2759, pruned_loss=0.06986, over 18706.00 frames. ], tot_loss[loss=0.2209, simple_loss=0.2975, pruned_loss=0.07214, over 3676674.96 frames. ], batch size: 41, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:51:53,023 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.996e+02 5.642e+02 6.698e+02 8.791e+02 1.815e+03, threshold=1.340e+03, percent-clipped=2.0 +2023-04-02 08:52:15,015 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3549, 1.4785, 1.7554, 1.5609, 2.6138, 2.1084, 2.7448, 1.2574], + device='cuda:3'), covar=tensor([0.2305, 0.3917, 0.2402, 0.1867, 0.1436, 0.2087, 0.1407, 0.3863], + device='cuda:3'), in_proj_covar=tensor([0.0505, 0.0609, 0.0661, 0.0457, 0.0601, 0.0509, 0.0645, 0.0515], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 08:52:27,871 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:52:31,273 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9773, 1.8180, 1.6982, 2.0975, 1.8124, 1.7927, 1.6574, 1.9090], + device='cuda:3'), covar=tensor([0.1050, 0.1505, 0.1383, 0.0943, 0.1231, 0.0511, 0.1336, 0.0691], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0353, 0.0300, 0.0243, 0.0295, 0.0248, 0.0295, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:52:43,999 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=109941.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:52:51,843 INFO [train.py:903] (3/4) Epoch 17, batch 700, loss[loss=0.2077, simple_loss=0.2926, pruned_loss=0.06137, over 19789.00 frames. ], tot_loss[loss=0.2195, simple_loss=0.2966, pruned_loss=0.07122, over 3708895.95 frames. ], batch size: 56, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:52:58,969 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109952.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:53:57,844 INFO [train.py:903] (3/4) Epoch 17, batch 750, loss[loss=0.261, simple_loss=0.3315, pruned_loss=0.09525, over 19729.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2949, pruned_loss=0.06988, over 3746525.40 frames. ], batch size: 63, lr: 4.93e-03, grad_scale: 8.0 +2023-04-02 08:54:02,554 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.920e+02 4.697e+02 5.712e+02 7.217e+02 1.165e+03, threshold=1.142e+03, percent-clipped=0.0 +2023-04-02 08:54:25,959 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110021.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:54:33,253 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110027.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:54:44,229 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110035.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:55:00,573 INFO [train.py:903] (3/4) Epoch 17, batch 800, loss[loss=0.219, simple_loss=0.3031, pruned_loss=0.06744, over 19660.00 frames. ], tot_loss[loss=0.2168, simple_loss=0.2949, pruned_loss=0.06936, over 3770559.50 frames. ], batch size: 55, lr: 4.92e-03, grad_scale: 8.0 +2023-04-02 08:55:04,469 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110051.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:55:05,703 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110052.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:55:10,530 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110056.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:55:14,886 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 08:55:17,675 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110062.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:55:27,182 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.86 vs. limit=5.0 +2023-04-02 08:55:35,612 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110076.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:56:03,217 INFO [train.py:903] (3/4) Epoch 17, batch 850, loss[loss=0.1826, simple_loss=0.2647, pruned_loss=0.05029, over 19837.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2953, pruned_loss=0.06965, over 3778359.56 frames. ], batch size: 52, lr: 4.92e-03, grad_scale: 8.0 +2023-04-02 08:56:06,200 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.347e+02 5.213e+02 6.563e+02 8.363e+02 2.159e+03, threshold=1.313e+03, percent-clipped=10.0 +2023-04-02 08:56:42,126 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1935, 3.8374, 2.6674, 3.4369, 1.2445, 3.7077, 3.6641, 3.7385], + device='cuda:3'), covar=tensor([0.0845, 0.1079, 0.1968, 0.0791, 0.3527, 0.0837, 0.0912, 0.1295], + device='cuda:3'), in_proj_covar=tensor([0.0470, 0.0382, 0.0463, 0.0330, 0.0387, 0.0399, 0.0398, 0.0429], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 08:56:42,143 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110128.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:56:47,720 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110133.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:56:51,232 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110136.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 08:56:56,482 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 08:57:04,358 INFO [train.py:903] (3/4) Epoch 17, batch 900, loss[loss=0.245, simple_loss=0.3224, pruned_loss=0.08381, over 19384.00 frames. ], tot_loss[loss=0.2176, simple_loss=0.2954, pruned_loss=0.06984, over 3796304.37 frames. ], batch size: 70, lr: 4.92e-03, grad_scale: 8.0 +2023-04-02 08:58:05,928 INFO [train.py:903] (3/4) Epoch 17, batch 950, loss[loss=0.2481, simple_loss=0.3208, pruned_loss=0.08768, over 19725.00 frames. ], tot_loss[loss=0.2193, simple_loss=0.2969, pruned_loss=0.07079, over 3795622.26 frames. ], batch size: 63, lr: 4.92e-03, grad_scale: 8.0 +2023-04-02 08:58:08,342 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 08:58:09,604 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.667e+02 4.867e+02 6.255e+02 8.229e+02 2.250e+03, threshold=1.251e+03, percent-clipped=5.0 +2023-04-02 08:59:09,004 INFO [train.py:903] (3/4) Epoch 17, batch 1000, loss[loss=0.2158, simple_loss=0.2998, pruned_loss=0.06585, over 19583.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2962, pruned_loss=0.07059, over 3808415.89 frames. ], batch size: 61, lr: 4.92e-03, grad_scale: 4.0 +2023-04-02 09:00:05,684 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 09:00:11,629 INFO [train.py:903] (3/4) Epoch 17, batch 1050, loss[loss=0.193, simple_loss=0.2815, pruned_loss=0.05224, over 19673.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2956, pruned_loss=0.07038, over 3809160.52 frames. ], batch size: 58, lr: 4.92e-03, grad_scale: 4.0 +2023-04-02 09:00:15,433 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110301.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:00:16,231 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.124e+02 4.918e+02 6.228e+02 8.050e+02 1.872e+03, threshold=1.246e+03, percent-clipped=2.0 +2023-04-02 09:00:28,181 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110312.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:00:40,523 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110321.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:00:45,000 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 09:01:00,212 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110336.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:01:01,463 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110337.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:01:13,841 INFO [train.py:903] (3/4) Epoch 17, batch 1100, loss[loss=0.1982, simple_loss=0.2715, pruned_loss=0.06243, over 19502.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2954, pruned_loss=0.07032, over 3805217.59 frames. ], batch size: 49, lr: 4.92e-03, grad_scale: 4.0 +2023-04-02 09:01:36,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1191, 1.1224, 1.4645, 1.3268, 2.7361, 1.0007, 2.1498, 2.9903], + device='cuda:3'), covar=tensor([0.0598, 0.2867, 0.2829, 0.1792, 0.0803, 0.2371, 0.1053, 0.0360], + device='cuda:3'), in_proj_covar=tensor([0.0389, 0.0355, 0.0375, 0.0335, 0.0360, 0.0342, 0.0357, 0.0381], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:01:54,278 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110379.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:02:09,871 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110392.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:02:16,582 INFO [train.py:903] (3/4) Epoch 17, batch 1150, loss[loss=0.2163, simple_loss=0.2952, pruned_loss=0.06871, over 19553.00 frames. ], tot_loss[loss=0.2175, simple_loss=0.2951, pruned_loss=0.06991, over 3810432.81 frames. ], batch size: 64, lr: 4.92e-03, grad_scale: 4.0 +2023-04-02 09:02:21,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.021e+02 4.729e+02 5.705e+02 7.310e+02 1.426e+03, threshold=1.141e+03, percent-clipped=3.0 +2023-04-02 09:02:28,349 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110406.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:02:43,205 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110417.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:03:20,434 INFO [train.py:903] (3/4) Epoch 17, batch 1200, loss[loss=0.213, simple_loss=0.2902, pruned_loss=0.06795, over 19761.00 frames. ], tot_loss[loss=0.2175, simple_loss=0.2953, pruned_loss=0.06985, over 3818569.93 frames. ], batch size: 54, lr: 4.92e-03, grad_scale: 8.0 +2023-04-02 09:03:22,075 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.12 vs. limit=5.0 +2023-04-02 09:03:49,916 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110472.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:03:53,238 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 09:03:55,659 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:04:19,557 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110494.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:04:23,826 INFO [train.py:903] (3/4) Epoch 17, batch 1250, loss[loss=0.2092, simple_loss=0.2848, pruned_loss=0.06678, over 19736.00 frames. ], tot_loss[loss=0.2171, simple_loss=0.2949, pruned_loss=0.06969, over 3814535.85 frames. ], batch size: 51, lr: 4.91e-03, grad_scale: 8.0 +2023-04-02 09:04:28,259 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.549e+02 5.334e+02 6.614e+02 7.905e+02 1.343e+03, threshold=1.323e+03, percent-clipped=1.0 +2023-04-02 09:04:48,482 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3471, 3.0319, 2.2852, 2.7663, 0.8621, 2.9578, 2.8846, 2.9726], + device='cuda:3'), covar=tensor([0.1091, 0.1460, 0.2030, 0.1164, 0.3842, 0.1077, 0.1137, 0.1340], + device='cuda:3'), in_proj_covar=tensor([0.0471, 0.0385, 0.0465, 0.0331, 0.0389, 0.0402, 0.0400, 0.0429], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:04:52,178 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110521.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:05:25,877 INFO [train.py:903] (3/4) Epoch 17, batch 1300, loss[loss=0.2429, simple_loss=0.3242, pruned_loss=0.08076, over 19479.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.295, pruned_loss=0.06986, over 3820622.06 frames. ], batch size: 64, lr: 4.91e-03, grad_scale: 8.0 +2023-04-02 09:06:02,874 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.37 vs. limit=5.0 +2023-04-02 09:06:14,221 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110587.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:06:20,240 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110592.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:06:26,791 INFO [train.py:903] (3/4) Epoch 17, batch 1350, loss[loss=0.2185, simple_loss=0.3018, pruned_loss=0.06757, over 18881.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2962, pruned_loss=0.07056, over 3822492.44 frames. ], batch size: 75, lr: 4.91e-03, grad_scale: 8.0 +2023-04-02 09:06:31,267 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.362e+02 4.458e+02 6.078e+02 8.226e+02 1.667e+03, threshold=1.216e+03, percent-clipped=3.0 +2023-04-02 09:06:35,113 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2734, 1.4052, 1.7953, 1.5503, 2.4408, 2.0519, 2.4733, 1.0380], + device='cuda:3'), covar=tensor([0.2732, 0.4608, 0.2758, 0.2122, 0.1619, 0.2391, 0.1692, 0.4645], + device='cuda:3'), in_proj_covar=tensor([0.0515, 0.0621, 0.0675, 0.0467, 0.0612, 0.0518, 0.0658, 0.0525], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 09:07:07,157 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110629.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:07:26,715 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110645.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:07:28,037 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0490, 1.4906, 1.8745, 1.3382, 2.8959, 4.4923, 4.4336, 4.8986], + device='cuda:3'), covar=tensor([0.1738, 0.3521, 0.3197, 0.2282, 0.0632, 0.0228, 0.0156, 0.0170], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0308, 0.0337, 0.0258, 0.0232, 0.0175, 0.0209, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:07:29,985 INFO [train.py:903] (3/4) Epoch 17, batch 1400, loss[loss=0.2099, simple_loss=0.285, pruned_loss=0.06742, over 19622.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2954, pruned_loss=0.07019, over 3836229.36 frames. ], batch size: 50, lr: 4.91e-03, grad_scale: 4.0 +2023-04-02 09:07:51,413 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110665.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:08:08,773 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110680.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:08:32,871 INFO [train.py:903] (3/4) Epoch 17, batch 1450, loss[loss=0.2446, simple_loss=0.3117, pruned_loss=0.08873, over 19768.00 frames. ], tot_loss[loss=0.2187, simple_loss=0.2961, pruned_loss=0.07071, over 3841146.91 frames. ], batch size: 54, lr: 4.91e-03, grad_scale: 4.0 +2023-04-02 09:08:32,997 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 09:08:38,609 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.085e+02 5.047e+02 5.721e+02 7.117e+02 1.861e+03, threshold=1.144e+03, percent-clipped=3.0 +2023-04-02 09:09:35,399 INFO [train.py:903] (3/4) Epoch 17, batch 1500, loss[loss=0.1899, simple_loss=0.2819, pruned_loss=0.04899, over 19668.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2955, pruned_loss=0.06965, over 3849819.02 frames. ], batch size: 58, lr: 4.91e-03, grad_scale: 4.0 +2023-04-02 09:09:38,214 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110750.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:09:42,819 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110754.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:09:50,088 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110760.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:11,865 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110775.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:14,254 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110777.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:17,726 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110780.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:27,709 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-02 09:10:35,383 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110795.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:38,184 INFO [train.py:903] (3/4) Epoch 17, batch 1550, loss[loss=0.2479, simple_loss=0.3215, pruned_loss=0.08717, over 17916.00 frames. ], tot_loss[loss=0.217, simple_loss=0.2951, pruned_loss=0.06949, over 3830193.17 frames. ], batch size: 83, lr: 4.91e-03, grad_scale: 4.0 +2023-04-02 09:10:44,270 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110802.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:10:44,893 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.880e+02 4.421e+02 5.256e+02 6.667e+02 1.625e+03, threshold=1.051e+03, percent-clipped=2.0 +2023-04-02 09:11:35,791 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110843.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:11:42,192 INFO [train.py:903] (3/4) Epoch 17, batch 1600, loss[loss=0.2419, simple_loss=0.3174, pruned_loss=0.08322, over 19535.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2943, pruned_loss=0.06927, over 3828145.09 frames. ], batch size: 54, lr: 4.91e-03, grad_scale: 8.0 +2023-04-02 09:11:42,643 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:12:05,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 09:12:07,122 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110868.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:12:12,612 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110873.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:12:29,585 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110886.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:12:45,475 INFO [train.py:903] (3/4) Epoch 17, batch 1650, loss[loss=0.1911, simple_loss=0.2672, pruned_loss=0.05746, over 19725.00 frames. ], tot_loss[loss=0.217, simple_loss=0.2942, pruned_loss=0.06985, over 3823325.44 frames. ], batch size: 46, lr: 4.91e-03, grad_scale: 8.0 +2023-04-02 09:12:51,298 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.530e+02 5.600e+02 6.791e+02 9.379e+02 3.114e+03, threshold=1.358e+03, percent-clipped=15.0 +2023-04-02 09:13:47,390 INFO [train.py:903] (3/4) Epoch 17, batch 1700, loss[loss=0.2239, simple_loss=0.3037, pruned_loss=0.07202, over 19689.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2935, pruned_loss=0.06897, over 3830127.40 frames. ], batch size: 60, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:14:04,774 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-02 09:14:19,118 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=110973.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:14:24,098 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110976.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:14:29,664 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 09:14:49,216 INFO [train.py:903] (3/4) Epoch 17, batch 1750, loss[loss=0.2209, simple_loss=0.3031, pruned_loss=0.0693, over 18363.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2933, pruned_loss=0.069, over 3834701.89 frames. ], batch size: 84, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:14:55,345 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.686e+02 4.870e+02 5.792e+02 7.151e+02 1.845e+03, threshold=1.158e+03, percent-clipped=1.0 +2023-04-02 09:15:13,784 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3302, 1.3296, 1.9447, 1.5482, 3.0318, 4.6573, 4.5390, 5.0223], + device='cuda:3'), covar=tensor([0.1560, 0.3801, 0.3152, 0.2214, 0.0569, 0.0176, 0.0160, 0.0163], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0309, 0.0338, 0.0259, 0.0233, 0.0176, 0.0210, 0.0234], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 09:15:15,078 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111016.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:15:37,646 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111036.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:15:43,447 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111041.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:15:51,886 INFO [train.py:903] (3/4) Epoch 17, batch 1800, loss[loss=0.275, simple_loss=0.3442, pruned_loss=0.1029, over 19272.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2935, pruned_loss=0.06928, over 3831993.26 frames. ], batch size: 66, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:15:57,974 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111051.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:16:10,614 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111061.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:16:28,047 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111076.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:16:44,399 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111088.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:16:51,950 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 09:16:56,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0516, 2.1179, 2.3101, 2.8200, 2.1046, 2.7570, 2.4765, 2.1569], + device='cuda:3'), covar=tensor([0.3872, 0.3680, 0.1728, 0.2109, 0.3853, 0.1745, 0.4049, 0.3087], + device='cuda:3'), in_proj_covar=tensor([0.0853, 0.0902, 0.0682, 0.0909, 0.0833, 0.0770, 0.0812, 0.0752], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 09:16:57,620 INFO [train.py:903] (3/4) Epoch 17, batch 1850, loss[loss=0.1829, simple_loss=0.2554, pruned_loss=0.05519, over 19719.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.2936, pruned_loss=0.06899, over 3829429.49 frames. ], batch size: 45, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:16:57,841 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=111098.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:17:03,773 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.487e+02 5.241e+02 6.549e+02 7.790e+02 1.838e+03, threshold=1.310e+03, percent-clipped=3.0 +2023-04-02 09:17:29,438 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 09:18:00,991 INFO [train.py:903] (3/4) Epoch 17, batch 1900, loss[loss=0.2216, simple_loss=0.2889, pruned_loss=0.07715, over 19404.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.2929, pruned_loss=0.06834, over 3833935.83 frames. ], batch size: 48, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:18:17,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 09:18:24,354 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 09:18:34,781 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3820, 1.4011, 1.7199, 1.5966, 2.4967, 2.1860, 2.5325, 1.1107], + device='cuda:3'), covar=tensor([0.2384, 0.4282, 0.2664, 0.1906, 0.1476, 0.2079, 0.1421, 0.4127], + device='cuda:3'), in_proj_covar=tensor([0.0506, 0.0613, 0.0670, 0.0462, 0.0608, 0.0513, 0.0651, 0.0521], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:18:49,758 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 09:19:02,961 INFO [train.py:903] (3/4) Epoch 17, batch 1950, loss[loss=0.2171, simple_loss=0.2829, pruned_loss=0.07561, over 19410.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2928, pruned_loss=0.06852, over 3841852.11 frames. ], batch size: 48, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:19:08,707 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.803e+02 5.118e+02 6.155e+02 7.161e+02 1.490e+03, threshold=1.231e+03, percent-clipped=2.0 +2023-04-02 09:19:10,547 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.59 vs. limit=2.0 +2023-04-02 09:19:23,921 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111213.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:19:24,967 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5809, 1.1910, 1.4869, 1.6092, 3.1318, 1.0405, 2.3337, 3.4894], + device='cuda:3'), covar=tensor([0.0495, 0.2929, 0.2818, 0.1701, 0.0737, 0.2564, 0.1223, 0.0293], + device='cuda:3'), in_proj_covar=tensor([0.0389, 0.0352, 0.0374, 0.0335, 0.0360, 0.0342, 0.0358, 0.0381], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:19:44,081 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=111230.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:19:59,225 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111243.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:20:05,434 INFO [train.py:903] (3/4) Epoch 17, batch 2000, loss[loss=0.1767, simple_loss=0.2574, pruned_loss=0.04804, over 19071.00 frames. ], tot_loss[loss=0.215, simple_loss=0.2932, pruned_loss=0.06838, over 3839342.63 frames. ], batch size: 42, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:21:04,232 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 09:21:07,754 INFO [train.py:903] (3/4) Epoch 17, batch 2050, loss[loss=0.2064, simple_loss=0.2726, pruned_loss=0.07013, over 19755.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2924, pruned_loss=0.068, over 3847673.66 frames. ], batch size: 47, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:21:14,880 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.549e+02 5.413e+02 6.604e+02 7.706e+02 1.674e+03, threshold=1.321e+03, percent-clipped=5.0 +2023-04-02 09:21:22,818 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 09:21:24,037 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 09:21:34,592 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=111320.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:21:43,899 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 09:22:05,012 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111344.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:22:06,132 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111345.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:22:09,327 INFO [train.py:903] (3/4) Epoch 17, batch 2100, loss[loss=0.226, simple_loss=0.3034, pruned_loss=0.07425, over 19334.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2928, pruned_loss=0.06848, over 3854590.39 frames. ], batch size: 70, lr: 4.90e-03, grad_scale: 8.0 +2023-04-02 09:22:34,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111369.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:22:38,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 09:22:55,888 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3460, 2.1563, 2.0140, 1.8362, 1.5600, 1.7937, 0.7120, 1.2391], + device='cuda:3'), covar=tensor([0.0525, 0.0564, 0.0423, 0.0712, 0.1137, 0.0837, 0.1171, 0.0984], + device='cuda:3'), in_proj_covar=tensor([0.0351, 0.0350, 0.0348, 0.0374, 0.0450, 0.0382, 0.0326, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:23:01,386 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 09:23:10,553 INFO [train.py:903] (3/4) Epoch 17, batch 2150, loss[loss=0.2231, simple_loss=0.2931, pruned_loss=0.07655, over 19845.00 frames. ], tot_loss[loss=0.2153, simple_loss=0.2925, pruned_loss=0.06907, over 3843083.16 frames. ], batch size: 52, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:23:15,802 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4888, 1.6289, 2.1217, 1.7920, 3.3484, 2.7162, 3.7313, 1.7579], + device='cuda:3'), covar=tensor([0.2359, 0.4196, 0.2604, 0.1788, 0.1451, 0.1911, 0.1435, 0.3764], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0615, 0.0671, 0.0463, 0.0611, 0.0514, 0.0653, 0.0522], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:23:16,478 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.230e+02 4.943e+02 5.975e+02 7.359e+02 1.553e+03, threshold=1.195e+03, percent-clipped=3.0 +2023-04-02 09:23:56,918 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111435.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:24:11,372 INFO [train.py:903] (3/4) Epoch 17, batch 2200, loss[loss=0.2784, simple_loss=0.3492, pruned_loss=0.1038, over 19449.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2928, pruned_loss=0.06917, over 3832611.60 frames. ], batch size: 64, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:24:39,059 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111469.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:24:58,479 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111486.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:25:11,183 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111494.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:25:15,257 INFO [train.py:903] (3/4) Epoch 17, batch 2250, loss[loss=0.2273, simple_loss=0.3103, pruned_loss=0.07212, over 19664.00 frames. ], tot_loss[loss=0.2162, simple_loss=0.2934, pruned_loss=0.06952, over 3820813.46 frames. ], batch size: 60, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:25:22,029 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.234e+02 5.257e+02 6.827e+02 8.687e+02 2.303e+03, threshold=1.365e+03, percent-clipped=8.0 +2023-04-02 09:26:16,850 INFO [train.py:903] (3/4) Epoch 17, batch 2300, loss[loss=0.198, simple_loss=0.2933, pruned_loss=0.05132, over 19785.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2947, pruned_loss=0.06998, over 3818532.24 frames. ], batch size: 56, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:26:27,079 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 09:26:38,588 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-04-02 09:27:05,581 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=111587.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:27:18,435 INFO [train.py:903] (3/4) Epoch 17, batch 2350, loss[loss=0.1721, simple_loss=0.2565, pruned_loss=0.04384, over 19396.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.293, pruned_loss=0.06925, over 3829235.65 frames. ], batch size: 48, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:27:22,382 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111601.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:27:24,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.974e+02 5.149e+02 5.946e+02 7.803e+02 1.982e+03, threshold=1.189e+03, percent-clipped=4.0 +2023-04-02 09:27:54,516 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111626.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:27:58,756 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 09:28:06,315 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.35 vs. limit=5.0 +2023-04-02 09:28:14,873 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 09:28:19,177 INFO [train.py:903] (3/4) Epoch 17, batch 2400, loss[loss=0.2664, simple_loss=0.3316, pruned_loss=0.1006, over 19611.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2947, pruned_loss=0.0705, over 3827702.35 frames. ], batch size: 57, lr: 4.89e-03, grad_scale: 8.0 +2023-04-02 09:28:23,952 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5363, 1.6388, 1.7268, 1.9772, 1.5146, 1.8779, 1.8019, 1.6679], + device='cuda:3'), covar=tensor([0.3262, 0.2667, 0.1460, 0.1651, 0.2814, 0.1486, 0.3373, 0.2416], + device='cuda:3'), in_proj_covar=tensor([0.0854, 0.0902, 0.0683, 0.0909, 0.0833, 0.0768, 0.0816, 0.0752], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 09:29:15,449 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111691.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:29:24,913 INFO [train.py:903] (3/4) Epoch 17, batch 2450, loss[loss=0.2025, simple_loss=0.2865, pruned_loss=0.05925, over 19761.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2941, pruned_loss=0.07014, over 3818891.55 frames. ], batch size: 54, lr: 4.89e-03, grad_scale: 4.0 +2023-04-02 09:29:29,898 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111702.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:29:30,392 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-04-02 09:29:32,591 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.040e+02 4.976e+02 6.292e+02 8.323e+02 1.636e+03, threshold=1.258e+03, percent-clipped=0.0 +2023-04-02 09:29:33,040 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9750, 1.5761, 1.5164, 1.8398, 1.5660, 1.7000, 1.4174, 1.8219], + device='cuda:3'), covar=tensor([0.0981, 0.1260, 0.1510, 0.1000, 0.1197, 0.0535, 0.1423, 0.0718], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0354, 0.0302, 0.0245, 0.0299, 0.0248, 0.0297, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:29:47,095 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111716.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:30:18,828 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9363, 1.9763, 2.2042, 2.6753, 1.9722, 2.5591, 2.3132, 2.0004], + device='cuda:3'), covar=tensor([0.4022, 0.3713, 0.1734, 0.2168, 0.3675, 0.1871, 0.4371, 0.3121], + device='cuda:3'), in_proj_covar=tensor([0.0857, 0.0907, 0.0687, 0.0914, 0.0836, 0.0773, 0.0818, 0.0753], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 09:30:27,987 INFO [train.py:903] (3/4) Epoch 17, batch 2500, loss[loss=0.1805, simple_loss=0.2563, pruned_loss=0.05234, over 19801.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2946, pruned_loss=0.07002, over 3819750.96 frames. ], batch size: 47, lr: 4.89e-03, grad_scale: 4.0 +2023-04-02 09:31:28,662 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-02 09:31:31,103 INFO [train.py:903] (3/4) Epoch 17, batch 2550, loss[loss=0.2634, simple_loss=0.3351, pruned_loss=0.09585, over 19620.00 frames. ], tot_loss[loss=0.2167, simple_loss=0.2945, pruned_loss=0.06944, over 3821835.23 frames. ], batch size: 57, lr: 4.89e-03, grad_scale: 4.0 +2023-04-02 09:31:38,617 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.039e+02 5.268e+02 6.331e+02 7.722e+02 1.710e+03, threshold=1.266e+03, percent-clipped=3.0 +2023-04-02 09:31:44,786 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111809.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 09:32:13,247 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=111830.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:32:28,427 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 09:32:34,317 INFO [train.py:903] (3/4) Epoch 17, batch 2600, loss[loss=0.2444, simple_loss=0.3192, pruned_loss=0.08483, over 19598.00 frames. ], tot_loss[loss=0.217, simple_loss=0.2947, pruned_loss=0.06959, over 3827764.58 frames. ], batch size: 61, lr: 4.88e-03, grad_scale: 4.0 +2023-04-02 09:33:38,519 INFO [train.py:903] (3/4) Epoch 17, batch 2650, loss[loss=0.2617, simple_loss=0.3267, pruned_loss=0.09834, over 17364.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2942, pruned_loss=0.06931, over 3825601.54 frames. ], batch size: 101, lr: 4.88e-03, grad_scale: 4.0 +2023-04-02 09:33:46,345 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.387e+02 5.132e+02 6.430e+02 7.842e+02 1.964e+03, threshold=1.286e+03, percent-clipped=4.0 +2023-04-02 09:33:47,844 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9216, 4.3581, 4.6348, 4.6307, 1.6203, 4.3762, 3.7620, 4.3172], + device='cuda:3'), covar=tensor([0.1450, 0.0788, 0.0537, 0.0583, 0.5666, 0.0744, 0.0615, 0.1075], + device='cuda:3'), in_proj_covar=tensor([0.0745, 0.0681, 0.0887, 0.0773, 0.0792, 0.0640, 0.0532, 0.0818], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 09:33:58,627 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 09:34:38,392 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111945.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:34:41,289 INFO [train.py:903] (3/4) Epoch 17, batch 2700, loss[loss=0.1948, simple_loss=0.2806, pruned_loss=0.05448, over 19675.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.2936, pruned_loss=0.06927, over 3825148.58 frames. ], batch size: 53, lr: 4.88e-03, grad_scale: 4.0 +2023-04-02 09:34:54,245 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111958.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:35:26,072 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111983.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:35:43,876 INFO [train.py:903] (3/4) Epoch 17, batch 2750, loss[loss=0.2045, simple_loss=0.2947, pruned_loss=0.05711, over 19500.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.2939, pruned_loss=0.06915, over 3836289.77 frames. ], batch size: 64, lr: 4.88e-03, grad_scale: 4.0 +2023-04-02 09:35:52,116 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.760e+02 5.468e+02 6.814e+02 8.726e+02 1.544e+03, threshold=1.363e+03, percent-clipped=3.0 +2023-04-02 09:36:45,032 INFO [train.py:903] (3/4) Epoch 17, batch 2800, loss[loss=0.2493, simple_loss=0.3206, pruned_loss=0.08905, over 19733.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2957, pruned_loss=0.0703, over 3824097.52 frames. ], batch size: 63, lr: 4.88e-03, grad_scale: 8.0 +2023-04-02 09:37:26,163 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0020, 1.2741, 1.6302, 1.1119, 2.4619, 3.3621, 3.0704, 3.5913], + device='cuda:3'), covar=tensor([0.1722, 0.3602, 0.3198, 0.2464, 0.0572, 0.0188, 0.0222, 0.0237], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0307, 0.0337, 0.0256, 0.0229, 0.0174, 0.0208, 0.0233], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:37:29,661 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112084.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:37:48,060 INFO [train.py:903] (3/4) Epoch 17, batch 2850, loss[loss=0.2132, simple_loss=0.3003, pruned_loss=0.06303, over 19694.00 frames. ], tot_loss[loss=0.2177, simple_loss=0.2955, pruned_loss=0.06997, over 3837279.31 frames. ], batch size: 59, lr: 4.88e-03, grad_scale: 8.0 +2023-04-02 09:37:54,817 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.003e+02 5.207e+02 6.661e+02 8.674e+02 1.797e+03, threshold=1.332e+03, percent-clipped=6.0 +2023-04-02 09:37:57,498 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5877, 1.0872, 1.4157, 1.1812, 2.2395, 0.9950, 2.0169, 2.4371], + device='cuda:3'), covar=tensor([0.0655, 0.2711, 0.2710, 0.1625, 0.0848, 0.2062, 0.1031, 0.0458], + device='cuda:3'), in_proj_covar=tensor([0.0386, 0.0349, 0.0371, 0.0332, 0.0358, 0.0340, 0.0357, 0.0378], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:38:46,061 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 09:38:49,622 INFO [train.py:903] (3/4) Epoch 17, batch 2900, loss[loss=0.2325, simple_loss=0.3083, pruned_loss=0.07832, over 19755.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2961, pruned_loss=0.07035, over 3839018.31 frames. ], batch size: 63, lr: 4.88e-03, grad_scale: 8.0 +2023-04-02 09:38:56,333 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112153.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 09:39:51,810 INFO [train.py:903] (3/4) Epoch 17, batch 2950, loss[loss=0.2017, simple_loss=0.275, pruned_loss=0.06426, over 19803.00 frames. ], tot_loss[loss=0.218, simple_loss=0.296, pruned_loss=0.07, over 3839226.28 frames. ], batch size: 49, lr: 4.88e-03, grad_scale: 8.0 +2023-04-02 09:39:55,966 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112201.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:39:58,769 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.039e+02 4.879e+02 6.137e+02 7.850e+02 1.399e+03, threshold=1.227e+03, percent-clipped=1.0 +2023-04-02 09:40:27,856 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112226.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:40:36,006 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4137, 1.5309, 1.9608, 1.7233, 3.1138, 2.5214, 3.4693, 1.6771], + device='cuda:3'), covar=tensor([0.2449, 0.4105, 0.2658, 0.1901, 0.1510, 0.2158, 0.1637, 0.3929], + device='cuda:3'), in_proj_covar=tensor([0.0507, 0.0610, 0.0664, 0.0461, 0.0603, 0.0513, 0.0649, 0.0522], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 09:40:54,421 INFO [train.py:903] (3/4) Epoch 17, batch 3000, loss[loss=0.2298, simple_loss=0.3065, pruned_loss=0.07648, over 19681.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2948, pruned_loss=0.06905, over 3850338.00 frames. ], batch size: 58, lr: 4.88e-03, grad_scale: 8.0 +2023-04-02 09:40:54,422 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 09:41:09,003 INFO [train.py:937] (3/4) Epoch 17, validation: loss=0.1717, simple_loss=0.272, pruned_loss=0.03576, over 944034.00 frames. +2023-04-02 09:41:09,005 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 09:41:13,737 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 09:41:33,207 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112268.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 09:41:33,241 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2873, 1.9234, 1.5308, 1.0811, 1.8942, 0.9983, 1.1651, 1.7353], + device='cuda:3'), covar=tensor([0.0886, 0.0736, 0.1036, 0.0961, 0.0450, 0.1368, 0.0680, 0.0398], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0307, 0.0330, 0.0255, 0.0242, 0.0329, 0.0290, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:42:09,768 INFO [train.py:903] (3/4) Epoch 17, batch 3050, loss[loss=0.2281, simple_loss=0.3079, pruned_loss=0.07413, over 19674.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2956, pruned_loss=0.06935, over 3836769.69 frames. ], batch size: 60, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:42:16,486 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.397e+02 5.197e+02 6.217e+02 9.038e+02 1.667e+03, threshold=1.243e+03, percent-clipped=8.0 +2023-04-02 09:42:16,904 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1020, 1.7408, 1.7264, 2.0219, 1.6584, 1.8396, 1.6575, 1.9576], + device='cuda:3'), covar=tensor([0.0946, 0.1586, 0.1367, 0.1071, 0.1424, 0.0504, 0.1325, 0.0685], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0358, 0.0305, 0.0246, 0.0299, 0.0250, 0.0298, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:43:10,145 INFO [train.py:903] (3/4) Epoch 17, batch 3100, loss[loss=0.1934, simple_loss=0.2789, pruned_loss=0.05393, over 18762.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2967, pruned_loss=0.07075, over 3823969.40 frames. ], batch size: 74, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:44:14,537 INFO [train.py:903] (3/4) Epoch 17, batch 3150, loss[loss=0.2518, simple_loss=0.3241, pruned_loss=0.0898, over 19692.00 frames. ], tot_loss[loss=0.2203, simple_loss=0.2976, pruned_loss=0.07152, over 3816084.50 frames. ], batch size: 59, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:44:21,815 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.470e+02 5.016e+02 6.190e+02 7.660e+02 1.883e+03, threshold=1.238e+03, percent-clipped=9.0 +2023-04-02 09:44:25,575 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112407.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:44:42,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 09:44:52,122 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112428.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:45:17,097 INFO [train.py:903] (3/4) Epoch 17, batch 3200, loss[loss=0.229, simple_loss=0.294, pruned_loss=0.08195, over 19729.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2965, pruned_loss=0.07094, over 3811414.23 frames. ], batch size: 51, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:45:28,589 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4306, 2.0984, 1.6911, 1.3460, 2.0077, 1.2686, 1.1622, 1.8240], + device='cuda:3'), covar=tensor([0.0962, 0.0728, 0.1032, 0.0894, 0.0538, 0.1302, 0.0781, 0.0440], + device='cuda:3'), in_proj_covar=tensor([0.0295, 0.0308, 0.0330, 0.0256, 0.0243, 0.0329, 0.0290, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:46:19,038 INFO [train.py:903] (3/4) Epoch 17, batch 3250, loss[loss=0.2068, simple_loss=0.276, pruned_loss=0.06882, over 19746.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2954, pruned_loss=0.07028, over 3813283.06 frames. ], batch size: 46, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:46:26,172 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.294e+02 4.958e+02 6.274e+02 7.840e+02 2.025e+03, threshold=1.255e+03, percent-clipped=2.0 +2023-04-02 09:46:50,003 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1341, 1.3504, 1.5014, 1.4248, 2.7496, 1.0361, 2.2323, 3.0734], + device='cuda:3'), covar=tensor([0.0513, 0.2587, 0.2665, 0.1705, 0.0700, 0.2324, 0.1034, 0.0305], + device='cuda:3'), in_proj_covar=tensor([0.0385, 0.0350, 0.0369, 0.0330, 0.0356, 0.0338, 0.0355, 0.0377], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:46:51,236 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112524.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 09:47:13,539 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112543.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:47:18,737 INFO [train.py:903] (3/4) Epoch 17, batch 3300, loss[loss=0.2419, simple_loss=0.3204, pruned_loss=0.08173, over 19646.00 frames. ], tot_loss[loss=0.2194, simple_loss=0.2966, pruned_loss=0.07109, over 3816112.70 frames. ], batch size: 60, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:47:21,208 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112549.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 09:47:25,362 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 09:48:00,378 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112580.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:48:20,898 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112595.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:48:24,061 INFO [train.py:903] (3/4) Epoch 17, batch 3350, loss[loss=0.2405, simple_loss=0.3159, pruned_loss=0.08253, over 19269.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2956, pruned_loss=0.07063, over 3811287.27 frames. ], batch size: 66, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:48:31,311 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.087e+02 5.436e+02 6.846e+02 8.612e+02 1.565e+03, threshold=1.369e+03, percent-clipped=5.0 +2023-04-02 09:48:45,600 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8585, 1.9412, 2.1711, 2.5201, 1.7981, 2.4029, 2.2793, 1.9836], + device='cuda:3'), covar=tensor([0.4184, 0.3785, 0.1866, 0.2193, 0.4057, 0.1969, 0.4573, 0.3300], + device='cuda:3'), in_proj_covar=tensor([0.0857, 0.0907, 0.0687, 0.0913, 0.0838, 0.0775, 0.0821, 0.0755], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 09:49:09,948 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112635.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:49:24,124 INFO [train.py:903] (3/4) Epoch 17, batch 3400, loss[loss=0.204, simple_loss=0.2761, pruned_loss=0.06594, over 19788.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2959, pruned_loss=0.07081, over 3803739.34 frames. ], batch size: 47, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:49:27,243 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 09:49:31,437 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5982, 4.1445, 2.5187, 3.6686, 1.1122, 4.0399, 3.9435, 4.0728], + device='cuda:3'), covar=tensor([0.0600, 0.1060, 0.2145, 0.0861, 0.3718, 0.0759, 0.0891, 0.1142], + device='cuda:3'), in_proj_covar=tensor([0.0474, 0.0389, 0.0472, 0.0332, 0.0391, 0.0407, 0.0403, 0.0432], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:49:45,252 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-02 09:50:25,918 INFO [train.py:903] (3/4) Epoch 17, batch 3450, loss[loss=0.2553, simple_loss=0.3348, pruned_loss=0.08789, over 19744.00 frames. ], tot_loss[loss=0.2184, simple_loss=0.2959, pruned_loss=0.07045, over 3806586.45 frames. ], batch size: 63, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:50:28,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 09:50:33,006 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.425e+02 4.932e+02 6.092e+02 9.481e+02 2.200e+03, threshold=1.218e+03, percent-clipped=6.0 +2023-04-02 09:50:55,138 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.86 vs. limit=5.0 +2023-04-02 09:51:04,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8718, 1.7662, 1.4930, 1.8213, 1.7318, 1.4414, 1.4078, 1.7696], + device='cuda:3'), covar=tensor([0.1057, 0.1351, 0.1629, 0.1039, 0.1251, 0.0789, 0.1657, 0.0807], + device='cuda:3'), in_proj_covar=tensor([0.0261, 0.0353, 0.0299, 0.0242, 0.0295, 0.0246, 0.0293, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:51:27,355 INFO [train.py:903] (3/4) Epoch 17, batch 3500, loss[loss=0.1797, simple_loss=0.2588, pruned_loss=0.05036, over 19739.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.2969, pruned_loss=0.07115, over 3796408.79 frames. ], batch size: 45, lr: 4.87e-03, grad_scale: 8.0 +2023-04-02 09:51:32,076 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112751.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:52:31,109 INFO [train.py:903] (3/4) Epoch 17, batch 3550, loss[loss=0.2229, simple_loss=0.2936, pruned_loss=0.07606, over 19737.00 frames. ], tot_loss[loss=0.219, simple_loss=0.2967, pruned_loss=0.07058, over 3817591.56 frames. ], batch size: 51, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:52:32,837 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112799.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:52:38,380 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.345e+02 4.759e+02 5.980e+02 7.566e+02 1.638e+03, threshold=1.196e+03, percent-clipped=2.0 +2023-04-02 09:52:57,235 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8960, 4.3187, 4.6175, 4.6189, 1.7213, 4.3302, 3.7027, 4.3152], + device='cuda:3'), covar=tensor([0.1511, 0.0836, 0.0584, 0.0583, 0.5716, 0.0797, 0.0665, 0.1076], + device='cuda:3'), in_proj_covar=tensor([0.0752, 0.0691, 0.0900, 0.0777, 0.0797, 0.0647, 0.0538, 0.0823], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 09:53:03,354 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112824.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:53:33,303 INFO [train.py:903] (3/4) Epoch 17, batch 3600, loss[loss=0.2116, simple_loss=0.2955, pruned_loss=0.06383, over 19528.00 frames. ], tot_loss[loss=0.2168, simple_loss=0.2951, pruned_loss=0.06931, over 3821786.16 frames. ], batch size: 56, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:53:55,565 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112866.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:54:35,710 INFO [train.py:903] (3/4) Epoch 17, batch 3650, loss[loss=0.2171, simple_loss=0.2975, pruned_loss=0.06837, over 19654.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2948, pruned_loss=0.06902, over 3819826.44 frames. ], batch size: 53, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:54:43,571 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.951e+02 4.960e+02 5.826e+02 7.647e+02 1.614e+03, threshold=1.165e+03, percent-clipped=2.0 +2023-04-02 09:55:09,909 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112924.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:55:28,595 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112939.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:55:28,727 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112939.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:55:38,877 INFO [train.py:903] (3/4) Epoch 17, batch 3700, loss[loss=0.2587, simple_loss=0.3256, pruned_loss=0.09587, over 13713.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.2943, pruned_loss=0.06872, over 3821589.39 frames. ], batch size: 139, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:56:05,001 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7154, 1.4705, 1.4438, 2.0566, 1.5400, 2.0801, 2.0388, 1.8848], + device='cuda:3'), covar=tensor([0.0882, 0.0993, 0.1103, 0.0869, 0.0967, 0.0723, 0.0878, 0.0683], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0222, 0.0244, 0.0228, 0.0208, 0.0189, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 09:56:17,481 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=112979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:56:42,238 INFO [train.py:903] (3/4) Epoch 17, batch 3750, loss[loss=0.1911, simple_loss=0.271, pruned_loss=0.0556, over 19739.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.2947, pruned_loss=0.06846, over 3822820.71 frames. ], batch size: 51, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:56:49,255 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.286e+02 4.723e+02 6.001e+02 7.947e+02 1.345e+03, threshold=1.200e+03, percent-clipped=4.0 +2023-04-02 09:56:58,841 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5518, 2.3818, 1.7594, 1.5821, 2.2390, 1.4915, 1.4845, 1.9008], + device='cuda:3'), covar=tensor([0.1119, 0.0713, 0.0966, 0.0815, 0.0501, 0.1163, 0.0723, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0298, 0.0313, 0.0334, 0.0258, 0.0245, 0.0333, 0.0294, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 09:57:32,730 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113039.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:57:42,417 INFO [train.py:903] (3/4) Epoch 17, batch 3800, loss[loss=0.1966, simple_loss=0.2784, pruned_loss=0.05737, over 19590.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.295, pruned_loss=0.06902, over 3822334.27 frames. ], batch size: 52, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:57:49,591 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113054.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:58:14,097 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 09:58:39,388 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:58:43,323 INFO [train.py:903] (3/4) Epoch 17, batch 3850, loss[loss=0.249, simple_loss=0.3288, pruned_loss=0.08462, over 19677.00 frames. ], tot_loss[loss=0.2185, simple_loss=0.2966, pruned_loss=0.07018, over 3818011.53 frames. ], batch size: 60, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 09:58:51,557 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.627e+02 5.316e+02 6.326e+02 9.097e+02 1.552e+03, threshold=1.265e+03, percent-clipped=8.0 +2023-04-02 09:59:14,523 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113122.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:59:44,611 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113147.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 09:59:45,350 INFO [train.py:903] (3/4) Epoch 17, batch 3900, loss[loss=0.1982, simple_loss=0.2724, pruned_loss=0.06196, over 19746.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.296, pruned_loss=0.06992, over 3824050.35 frames. ], batch size: 47, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 10:00:48,758 INFO [train.py:903] (3/4) Epoch 17, batch 3950, loss[loss=0.2066, simple_loss=0.2828, pruned_loss=0.06523, over 19748.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2934, pruned_loss=0.06851, over 3833109.38 frames. ], batch size: 51, lr: 4.86e-03, grad_scale: 8.0 +2023-04-02 10:00:56,116 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 10:00:57,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.297e+02 4.545e+02 5.288e+02 6.585e+02 1.560e+03, threshold=1.058e+03, percent-clipped=1.0 +2023-04-02 10:01:51,451 INFO [train.py:903] (3/4) Epoch 17, batch 4000, loss[loss=0.2363, simple_loss=0.3121, pruned_loss=0.08025, over 19731.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2934, pruned_loss=0.06847, over 3829801.48 frames. ], batch size: 63, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:01:56,615 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113252.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:02:35,233 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=113283.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:02:39,797 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 10:02:49,725 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113295.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:02:52,758 INFO [train.py:903] (3/4) Epoch 17, batch 4050, loss[loss=0.2036, simple_loss=0.2801, pruned_loss=0.06359, over 19679.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2924, pruned_loss=0.06801, over 3830215.96 frames. ], batch size: 53, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:03:00,898 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.601e+02 4.703e+02 5.716e+02 7.594e+02 1.568e+03, threshold=1.143e+03, percent-clipped=5.0 +2023-04-02 10:03:08,184 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113310.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:03:21,623 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113320.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:03:38,694 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113335.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:03:54,663 INFO [train.py:903] (3/4) Epoch 17, batch 4100, loss[loss=0.2236, simple_loss=0.3037, pruned_loss=0.07179, over 19600.00 frames. ], tot_loss[loss=0.2141, simple_loss=0.2924, pruned_loss=0.06791, over 3836028.31 frames. ], batch size: 61, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:03:57,603 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113350.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:04:28,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113375.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:04:31,522 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 10:04:56,310 INFO [train.py:903] (3/4) Epoch 17, batch 4150, loss[loss=0.1555, simple_loss=0.2362, pruned_loss=0.03737, over 19776.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2917, pruned_loss=0.06751, over 3842218.86 frames. ], batch size: 47, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:04:56,637 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113398.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:05:03,828 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.156e+02 5.375e+02 6.520e+02 8.152e+02 2.133e+03, threshold=1.304e+03, percent-clipped=6.0 +2023-04-02 10:05:57,676 INFO [train.py:903] (3/4) Epoch 17, batch 4200, loss[loss=0.1916, simple_loss=0.2707, pruned_loss=0.05628, over 19593.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2926, pruned_loss=0.0684, over 3831759.98 frames. ], batch size: 52, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:06:02,338 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 10:06:41,260 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7362, 2.0669, 2.3571, 2.1962, 3.1465, 3.7057, 3.6477, 3.9956], + device='cuda:3'), covar=tensor([0.1445, 0.2830, 0.2507, 0.1880, 0.0801, 0.0305, 0.0184, 0.0252], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0310, 0.0339, 0.0258, 0.0232, 0.0177, 0.0211, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:06:59,531 INFO [train.py:903] (3/4) Epoch 17, batch 4250, loss[loss=0.2181, simple_loss=0.2854, pruned_loss=0.07541, over 19388.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2939, pruned_loss=0.06945, over 3813046.29 frames. ], batch size: 47, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:07:06,464 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.137e+02 4.808e+02 5.898e+02 7.585e+02 1.571e+03, threshold=1.180e+03, percent-clipped=5.0 +2023-04-02 10:07:13,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 10:07:15,390 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-02 10:07:21,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5779, 2.3655, 1.7451, 1.4957, 2.2170, 1.4383, 1.3113, 1.8913], + device='cuda:3'), covar=tensor([0.1055, 0.0690, 0.1008, 0.0837, 0.0533, 0.1185, 0.0820, 0.0522], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0310, 0.0329, 0.0255, 0.0244, 0.0329, 0.0291, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:07:24,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 10:08:02,054 INFO [train.py:903] (3/4) Epoch 17, batch 4300, loss[loss=0.2094, simple_loss=0.2891, pruned_loss=0.06484, over 17982.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2939, pruned_loss=0.06909, over 3813875.13 frames. ], batch size: 83, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:08:03,898 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 10:08:55,443 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 10:09:02,225 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=113596.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:09:04,377 INFO [train.py:903] (3/4) Epoch 17, batch 4350, loss[loss=0.2142, simple_loss=0.2897, pruned_loss=0.06934, over 19595.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.2931, pruned_loss=0.06828, over 3828059.75 frames. ], batch size: 50, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:09:12,442 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.419e+02 4.847e+02 6.118e+02 7.738e+02 1.753e+03, threshold=1.224e+03, percent-clipped=4.0 +2023-04-02 10:09:49,677 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113635.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:10:07,377 INFO [train.py:903] (3/4) Epoch 17, batch 4400, loss[loss=0.2143, simple_loss=0.2952, pruned_loss=0.06667, over 19672.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.292, pruned_loss=0.06819, over 3829503.36 frames. ], batch size: 59, lr: 4.85e-03, grad_scale: 8.0 +2023-04-02 10:10:14,945 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113654.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:10:26,336 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113664.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:10:33,150 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 10:10:43,117 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 10:10:45,285 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113679.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:11:07,792 INFO [train.py:903] (3/4) Epoch 17, batch 4450, loss[loss=0.2352, simple_loss=0.3227, pruned_loss=0.07387, over 19704.00 frames. ], tot_loss[loss=0.2154, simple_loss=0.2928, pruned_loss=0.06899, over 3838281.46 frames. ], batch size: 63, lr: 4.84e-03, grad_scale: 16.0 +2023-04-02 10:11:14,459 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.974e+02 5.100e+02 6.811e+02 8.906e+02 1.680e+03, threshold=1.362e+03, percent-clipped=7.0 +2023-04-02 10:11:22,995 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113711.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:12:07,712 INFO [train.py:903] (3/4) Epoch 17, batch 4500, loss[loss=0.2135, simple_loss=0.309, pruned_loss=0.059, over 19704.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.294, pruned_loss=0.06988, over 3840657.49 frames. ], batch size: 59, lr: 4.84e-03, grad_scale: 8.0 +2023-04-02 10:13:09,295 INFO [train.py:903] (3/4) Epoch 17, batch 4550, loss[loss=0.2182, simple_loss=0.2997, pruned_loss=0.06839, over 19492.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2945, pruned_loss=0.07018, over 3831670.93 frames. ], batch size: 64, lr: 4.84e-03, grad_scale: 4.0 +2023-04-02 10:13:19,359 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 10:13:20,452 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.619e+02 5.090e+02 6.214e+02 7.749e+02 1.433e+03, threshold=1.243e+03, percent-clipped=2.0 +2023-04-02 10:13:42,464 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 10:14:12,747 INFO [train.py:903] (3/4) Epoch 17, batch 4600, loss[loss=0.2069, simple_loss=0.2842, pruned_loss=0.06475, over 19739.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2944, pruned_loss=0.07025, over 3823485.35 frames. ], batch size: 51, lr: 4.84e-03, grad_scale: 4.0 +2023-04-02 10:15:14,385 INFO [train.py:903] (3/4) Epoch 17, batch 4650, loss[loss=0.2096, simple_loss=0.2874, pruned_loss=0.06593, over 19849.00 frames. ], tot_loss[loss=0.2177, simple_loss=0.295, pruned_loss=0.07017, over 3828456.38 frames. ], batch size: 52, lr: 4.84e-03, grad_scale: 4.0 +2023-04-02 10:15:23,609 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.367e+02 5.276e+02 6.482e+02 7.907e+02 1.823e+03, threshold=1.296e+03, percent-clipped=2.0 +2023-04-02 10:15:31,987 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 10:15:44,642 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 10:15:57,554 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9884, 3.4274, 1.9581, 2.0001, 3.0921, 1.8755, 1.4075, 2.1430], + device='cuda:3'), covar=tensor([0.1441, 0.0526, 0.1127, 0.0870, 0.0529, 0.1236, 0.1015, 0.0789], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0311, 0.0330, 0.0255, 0.0243, 0.0329, 0.0291, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:16:16,826 INFO [train.py:903] (3/4) Epoch 17, batch 4700, loss[loss=0.2258, simple_loss=0.3036, pruned_loss=0.07401, over 19371.00 frames. ], tot_loss[loss=0.2174, simple_loss=0.2949, pruned_loss=0.06994, over 3834346.01 frames. ], batch size: 70, lr: 4.84e-03, grad_scale: 4.0 +2023-04-02 10:16:43,000 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113967.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:16:43,744 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 10:16:56,543 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=113979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:17:12,237 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113992.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:17:19,856 INFO [train.py:903] (3/4) Epoch 17, batch 4750, loss[loss=0.1852, simple_loss=0.2631, pruned_loss=0.05364, over 19770.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2934, pruned_loss=0.06881, over 3842479.72 frames. ], batch size: 47, lr: 4.84e-03, grad_scale: 4.0 +2023-04-02 10:17:32,731 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.044e+02 4.836e+02 6.122e+02 7.624e+02 1.576e+03, threshold=1.224e+03, percent-clipped=1.0 +2023-04-02 10:17:35,088 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=114008.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:18:24,454 INFO [train.py:903] (3/4) Epoch 17, batch 4800, loss[loss=0.2122, simple_loss=0.2851, pruned_loss=0.06968, over 19577.00 frames. ], tot_loss[loss=0.2167, simple_loss=0.2942, pruned_loss=0.06966, over 3825042.51 frames. ], batch size: 52, lr: 4.84e-03, grad_scale: 8.0 +2023-04-02 10:19:22,151 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:19:26,560 INFO [train.py:903] (3/4) Epoch 17, batch 4850, loss[loss=0.202, simple_loss=0.2756, pruned_loss=0.06421, over 18603.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2946, pruned_loss=0.06996, over 3815854.16 frames. ], batch size: 41, lr: 4.84e-03, grad_scale: 8.0 +2023-04-02 10:19:35,430 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.185e+02 5.130e+02 6.675e+02 8.728e+02 1.864e+03, threshold=1.335e+03, percent-clipped=11.0 +2023-04-02 10:19:38,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1895, 2.2019, 2.4366, 3.1766, 2.1721, 2.8411, 2.6136, 2.1173], + device='cuda:3'), covar=tensor([0.4096, 0.4003, 0.1753, 0.2228, 0.4385, 0.2020, 0.4154, 0.3289], + device='cuda:3'), in_proj_covar=tensor([0.0857, 0.0906, 0.0688, 0.0913, 0.0837, 0.0774, 0.0819, 0.0756], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 10:19:52,811 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 10:19:57,430 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114123.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:20:09,285 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 10:20:14,655 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 10:20:19,134 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 10:20:20,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 10:20:25,234 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=114145.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:20:28,430 INFO [train.py:903] (3/4) Epoch 17, batch 4900, loss[loss=0.2041, simple_loss=0.2923, pruned_loss=0.05795, over 19712.00 frames. ], tot_loss[loss=0.2162, simple_loss=0.2936, pruned_loss=0.06941, over 3808005.06 frames. ], batch size: 51, lr: 4.84e-03, grad_scale: 8.0 +2023-04-02 10:20:28,479 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 10:20:48,168 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 10:21:23,045 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0154, 1.3007, 1.7351, 1.3120, 2.7992, 3.7797, 3.4847, 4.0105], + device='cuda:3'), covar=tensor([0.1790, 0.3780, 0.3228, 0.2366, 0.0565, 0.0161, 0.0212, 0.0201], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0312, 0.0341, 0.0259, 0.0233, 0.0178, 0.0211, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:21:29,557 INFO [train.py:903] (3/4) Epoch 17, batch 4950, loss[loss=0.234, simple_loss=0.3133, pruned_loss=0.07738, over 19670.00 frames. ], tot_loss[loss=0.2166, simple_loss=0.2938, pruned_loss=0.06968, over 3795530.24 frames. ], batch size: 60, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:21:41,941 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.237e+02 5.073e+02 6.090e+02 7.599e+02 1.461e+03, threshold=1.218e+03, percent-clipped=1.0 +2023-04-02 10:21:48,510 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 10:22:09,564 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 10:22:31,811 INFO [train.py:903] (3/4) Epoch 17, batch 5000, loss[loss=0.2067, simple_loss=0.2946, pruned_loss=0.05938, over 17976.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2944, pruned_loss=0.07015, over 3801224.40 frames. ], batch size: 83, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:22:39,615 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 10:22:50,097 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 10:23:32,992 INFO [train.py:903] (3/4) Epoch 17, batch 5050, loss[loss=0.1703, simple_loss=0.2511, pruned_loss=0.04471, over 18677.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2933, pruned_loss=0.0693, over 3796725.23 frames. ], batch size: 41, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:23:42,365 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.900e+02 5.068e+02 6.244e+02 7.899e+02 1.430e+03, threshold=1.249e+03, percent-clipped=5.0 +2023-04-02 10:24:10,966 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 10:24:34,966 INFO [train.py:903] (3/4) Epoch 17, batch 5100, loss[loss=0.1927, simple_loss=0.2685, pruned_loss=0.05844, over 19770.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2935, pruned_loss=0.06911, over 3797772.90 frames. ], batch size: 47, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:24:37,674 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114350.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:24:44,365 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 10:24:46,825 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 10:24:52,458 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 10:25:10,325 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114375.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:25:14,877 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114379.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:25:20,129 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.36 vs. limit=5.0 +2023-04-02 10:25:36,283 INFO [train.py:903] (3/4) Epoch 17, batch 5150, loss[loss=0.2281, simple_loss=0.3215, pruned_loss=0.0673, over 19665.00 frames. ], tot_loss[loss=0.2171, simple_loss=0.2949, pruned_loss=0.06966, over 3800570.19 frames. ], batch size: 58, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:25:44,134 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6995, 4.1532, 4.3926, 4.4463, 1.7305, 4.1096, 3.6111, 4.0727], + device='cuda:3'), covar=tensor([0.1631, 0.1042, 0.0617, 0.0627, 0.5681, 0.0882, 0.0690, 0.1178], + device='cuda:3'), in_proj_covar=tensor([0.0744, 0.0694, 0.0892, 0.0780, 0.0798, 0.0644, 0.0537, 0.0821], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 10:25:47,613 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114404.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:25:49,639 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.419e+02 5.288e+02 6.652e+02 7.839e+02 1.735e+03, threshold=1.330e+03, percent-clipped=3.0 +2023-04-02 10:25:50,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 10:26:24,429 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 10:26:41,804 INFO [train.py:903] (3/4) Epoch 17, batch 5200, loss[loss=0.2616, simple_loss=0.327, pruned_loss=0.09808, over 12789.00 frames. ], tot_loss[loss=0.2178, simple_loss=0.2951, pruned_loss=0.07027, over 3796946.66 frames. ], batch size: 136, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:26:54,840 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 10:27:10,257 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8660, 2.4311, 2.2252, 2.6730, 2.5923, 2.3497, 2.0763, 2.9703], + device='cuda:3'), covar=tensor([0.0844, 0.1604, 0.1376, 0.1057, 0.1322, 0.0498, 0.1302, 0.0517], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0356, 0.0301, 0.0245, 0.0300, 0.0248, 0.0295, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:27:19,897 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-02 10:27:33,131 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=114489.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:27:37,829 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 10:27:43,639 INFO [train.py:903] (3/4) Epoch 17, batch 5250, loss[loss=0.1932, simple_loss=0.2816, pruned_loss=0.05239, over 19543.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.2939, pruned_loss=0.06938, over 3797151.01 frames. ], batch size: 56, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:27:53,069 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.560e+02 4.802e+02 5.852e+02 7.465e+02 1.395e+03, threshold=1.170e+03, percent-clipped=1.0 +2023-04-02 10:28:44,585 INFO [train.py:903] (3/4) Epoch 17, batch 5300, loss[loss=0.2041, simple_loss=0.2754, pruned_loss=0.06637, over 19733.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2921, pruned_loss=0.06827, over 3809062.82 frames. ], batch size: 46, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:28:59,071 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 10:29:31,221 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3357, 1.3736, 1.6974, 1.5517, 2.4887, 2.1449, 2.6439, 1.2034], + device='cuda:3'), covar=tensor([0.2393, 0.4321, 0.2648, 0.1874, 0.1569, 0.2083, 0.1437, 0.4188], + device='cuda:3'), in_proj_covar=tensor([0.0513, 0.0614, 0.0668, 0.0461, 0.0607, 0.0515, 0.0649, 0.0524], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 10:29:44,122 INFO [train.py:903] (3/4) Epoch 17, batch 5350, loss[loss=0.1976, simple_loss=0.2715, pruned_loss=0.06179, over 19764.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2926, pruned_loss=0.06797, over 3823678.47 frames. ], batch size: 47, lr: 4.83e-03, grad_scale: 8.0 +2023-04-02 10:29:44,501 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1789, 1.4493, 1.9477, 1.4405, 3.0441, 4.3506, 4.2486, 4.8571], + device='cuda:3'), covar=tensor([0.1689, 0.3583, 0.3096, 0.2222, 0.0625, 0.0228, 0.0179, 0.0162], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0311, 0.0340, 0.0259, 0.0232, 0.0178, 0.0212, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:29:45,505 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4573, 2.2491, 2.0470, 2.4622, 2.3385, 2.0474, 2.1020, 2.4743], + device='cuda:3'), covar=tensor([0.0887, 0.1448, 0.1307, 0.0881, 0.1227, 0.0523, 0.1146, 0.0568], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0354, 0.0301, 0.0245, 0.0300, 0.0248, 0.0295, 0.0248], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:29:51,254 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114604.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:29:54,896 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.137e+02 5.171e+02 6.688e+02 9.089e+02 2.274e+03, threshold=1.338e+03, percent-clipped=9.0 +2023-04-02 10:30:19,023 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 10:30:46,226 INFO [train.py:903] (3/4) Epoch 17, batch 5400, loss[loss=0.2339, simple_loss=0.3123, pruned_loss=0.07772, over 19594.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2928, pruned_loss=0.06812, over 3833112.66 frames. ], batch size: 61, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:31:29,351 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4768, 1.5652, 1.7571, 1.8703, 1.3738, 1.7285, 1.8213, 1.6820], + device='cuda:3'), covar=tensor([0.3647, 0.3048, 0.1687, 0.1865, 0.3199, 0.1769, 0.4083, 0.2859], + device='cuda:3'), in_proj_covar=tensor([0.0865, 0.0915, 0.0692, 0.0922, 0.0844, 0.0779, 0.0821, 0.0761], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 10:31:29,566 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-02 10:31:47,169 INFO [train.py:903] (3/4) Epoch 17, batch 5450, loss[loss=0.2332, simple_loss=0.3118, pruned_loss=0.07732, over 19602.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2936, pruned_loss=0.06881, over 3823483.22 frames. ], batch size: 61, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:31:56,199 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.018e+02 4.503e+02 5.761e+02 7.243e+02 1.420e+03, threshold=1.152e+03, percent-clipped=1.0 +2023-04-02 10:32:31,004 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=114734.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:32:47,134 INFO [train.py:903] (3/4) Epoch 17, batch 5500, loss[loss=0.2036, simple_loss=0.2758, pruned_loss=0.06569, over 19403.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2937, pruned_loss=0.06834, over 3821897.68 frames. ], batch size: 48, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:32:47,643 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3625, 1.4971, 1.6557, 1.6218, 2.2768, 2.1209, 2.3841, 0.9171], + device='cuda:3'), covar=tensor([0.2353, 0.4012, 0.2519, 0.1814, 0.1450, 0.2092, 0.1267, 0.4225], + device='cuda:3'), in_proj_covar=tensor([0.0511, 0.0612, 0.0668, 0.0461, 0.0607, 0.0515, 0.0648, 0.0523], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 10:33:10,793 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 10:33:46,750 INFO [train.py:903] (3/4) Epoch 17, batch 5550, loss[loss=0.2251, simple_loss=0.302, pruned_loss=0.0741, over 19611.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2929, pruned_loss=0.06794, over 3841507.18 frames. ], batch size: 61, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:33:54,795 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 10:33:55,924 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.429e+02 4.950e+02 6.230e+02 7.289e+02 1.704e+03, threshold=1.246e+03, percent-clipped=5.0 +2023-04-02 10:34:05,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1612, 2.2060, 2.3994, 3.0320, 2.1928, 2.8781, 2.5553, 2.1371], + device='cuda:3'), covar=tensor([0.4212, 0.4076, 0.1906, 0.2486, 0.4406, 0.2046, 0.4355, 0.3314], + device='cuda:3'), in_proj_covar=tensor([0.0861, 0.0911, 0.0688, 0.0917, 0.0837, 0.0775, 0.0817, 0.0758], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 10:34:39,090 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-04-02 10:34:41,648 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 10:34:48,921 INFO [train.py:903] (3/4) Epoch 17, batch 5600, loss[loss=0.2162, simple_loss=0.2962, pruned_loss=0.0681, over 18801.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2929, pruned_loss=0.06792, over 3826065.41 frames. ], batch size: 74, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:35:03,678 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114860.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:35:33,269 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114885.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:35:39,359 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=114889.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:35:50,181 INFO [train.py:903] (3/4) Epoch 17, batch 5650, loss[loss=0.2138, simple_loss=0.2906, pruned_loss=0.06851, over 19673.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2937, pruned_loss=0.06878, over 3817657.56 frames. ], batch size: 60, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:35:59,360 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 5.160e+02 6.498e+02 8.575e+02 1.504e+03, threshold=1.300e+03, percent-clipped=5.0 +2023-04-02 10:36:35,344 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2099, 1.1586, 1.6192, 1.3059, 2.4332, 3.3145, 3.1003, 3.6010], + device='cuda:3'), covar=tensor([0.1734, 0.4803, 0.4041, 0.2200, 0.0652, 0.0219, 0.0286, 0.0294], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0312, 0.0340, 0.0259, 0.0233, 0.0178, 0.0212, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:36:36,156 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 10:36:51,140 INFO [train.py:903] (3/4) Epoch 17, batch 5700, loss[loss=0.1853, simple_loss=0.265, pruned_loss=0.0528, over 19502.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2932, pruned_loss=0.06887, over 3822713.16 frames. ], batch size: 49, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:37:08,346 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4242, 1.5153, 1.7557, 1.6703, 2.7122, 2.2426, 2.7364, 1.1671], + device='cuda:3'), covar=tensor([0.2391, 0.4180, 0.2672, 0.1850, 0.1373, 0.2064, 0.1370, 0.4146], + device='cuda:3'), in_proj_covar=tensor([0.0513, 0.0615, 0.0673, 0.0463, 0.0611, 0.0516, 0.0650, 0.0526], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 10:37:43,462 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4292, 1.2925, 1.2533, 1.8182, 1.4572, 1.7340, 1.8022, 1.5292], + device='cuda:3'), covar=tensor([0.0877, 0.0994, 0.1145, 0.0845, 0.0897, 0.0779, 0.0869, 0.0769], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0222, 0.0223, 0.0243, 0.0227, 0.0209, 0.0188, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 10:37:50,250 INFO [train.py:903] (3/4) Epoch 17, batch 5750, loss[loss=0.2071, simple_loss=0.2693, pruned_loss=0.07244, over 19766.00 frames. ], tot_loss[loss=0.2162, simple_loss=0.2939, pruned_loss=0.06922, over 3811068.94 frames. ], batch size: 46, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:37:50,264 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 10:37:57,211 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 10:37:59,472 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.823e+02 5.221e+02 6.429e+02 7.572e+02 1.818e+03, threshold=1.286e+03, percent-clipped=4.0 +2023-04-02 10:38:04,569 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 10:38:50,578 INFO [train.py:903] (3/4) Epoch 17, batch 5800, loss[loss=0.2117, simple_loss=0.278, pruned_loss=0.07268, over 19772.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.2929, pruned_loss=0.06836, over 3825043.08 frames. ], batch size: 47, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:38:52,860 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6874, 4.2220, 4.4566, 4.4738, 1.6371, 4.1853, 3.5986, 4.1664], + device='cuda:3'), covar=tensor([0.1604, 0.0780, 0.0602, 0.0633, 0.5887, 0.0884, 0.0698, 0.1076], + device='cuda:3'), in_proj_covar=tensor([0.0741, 0.0691, 0.0891, 0.0775, 0.0794, 0.0645, 0.0536, 0.0819], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 10:39:27,225 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=115078.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:39:28,644 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1963, 1.7890, 1.6870, 1.9662, 1.7803, 1.8124, 1.5127, 2.0334], + device='cuda:3'), covar=tensor([0.0862, 0.1255, 0.1488, 0.1035, 0.1291, 0.0529, 0.1447, 0.0682], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0352, 0.0302, 0.0245, 0.0298, 0.0247, 0.0294, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:39:35,785 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.08 vs. limit=2.0 +2023-04-02 10:39:52,203 INFO [train.py:903] (3/4) Epoch 17, batch 5850, loss[loss=0.2222, simple_loss=0.3016, pruned_loss=0.07144, over 18253.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2932, pruned_loss=0.06862, over 3799694.74 frames. ], batch size: 83, lr: 4.82e-03, grad_scale: 8.0 +2023-04-02 10:39:59,358 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115104.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:40:01,406 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.460e+02 4.663e+02 6.050e+02 7.882e+02 1.454e+03, threshold=1.210e+03, percent-clipped=2.0 +2023-04-02 10:40:51,591 INFO [train.py:903] (3/4) Epoch 17, batch 5900, loss[loss=0.1707, simple_loss=0.2519, pruned_loss=0.0448, over 19760.00 frames. ], tot_loss[loss=0.2153, simple_loss=0.2931, pruned_loss=0.06881, over 3808883.32 frames. ], batch size: 47, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:40:55,187 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 10:40:55,630 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0753, 1.8032, 1.4878, 1.1681, 1.6314, 1.1863, 1.0604, 1.5851], + device='cuda:3'), covar=tensor([0.0832, 0.0742, 0.1010, 0.0824, 0.0504, 0.1208, 0.0606, 0.0364], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0309, 0.0328, 0.0257, 0.0245, 0.0327, 0.0290, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:41:13,992 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 10:41:45,877 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115193.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:41:51,141 INFO [train.py:903] (3/4) Epoch 17, batch 5950, loss[loss=0.2466, simple_loss=0.3171, pruned_loss=0.08803, over 17234.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2922, pruned_loss=0.06867, over 3809082.54 frames. ], batch size: 101, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:42:00,464 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.984e+02 4.938e+02 6.318e+02 8.201e+02 2.090e+03, threshold=1.264e+03, percent-clipped=7.0 +2023-04-02 10:42:20,345 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3468, 1.3954, 1.6785, 1.5725, 2.5771, 2.2767, 2.7553, 1.0513], + device='cuda:3'), covar=tensor([0.2413, 0.4268, 0.2663, 0.1928, 0.1455, 0.2060, 0.1381, 0.4216], + device='cuda:3'), in_proj_covar=tensor([0.0514, 0.0615, 0.0674, 0.0464, 0.0611, 0.0518, 0.0652, 0.0527], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 10:42:35,073 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=115233.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:42:43,153 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6068, 4.1842, 2.7720, 3.7066, 1.0449, 4.0481, 3.9714, 4.0895], + device='cuda:3'), covar=tensor([0.0652, 0.0981, 0.1760, 0.0787, 0.3748, 0.0685, 0.0902, 0.1112], + device='cuda:3'), in_proj_covar=tensor([0.0471, 0.0388, 0.0469, 0.0336, 0.0391, 0.0406, 0.0404, 0.0433], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:42:51,742 INFO [train.py:903] (3/4) Epoch 17, batch 6000, loss[loss=0.2263, simple_loss=0.2975, pruned_loss=0.07755, over 19538.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2921, pruned_loss=0.06843, over 3798084.52 frames. ], batch size: 54, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:42:51,742 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 10:43:04,253 INFO [train.py:937] (3/4) Epoch 17, validation: loss=0.1707, simple_loss=0.2712, pruned_loss=0.03505, over 944034.00 frames. +2023-04-02 10:43:04,254 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 10:44:04,140 INFO [train.py:903] (3/4) Epoch 17, batch 6050, loss[loss=0.1957, simple_loss=0.2775, pruned_loss=0.05695, over 19769.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2933, pruned_loss=0.06909, over 3792101.85 frames. ], batch size: 54, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:44:15,951 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.299e+02 5.156e+02 6.136e+02 7.598e+02 1.906e+03, threshold=1.227e+03, percent-clipped=4.0 +2023-04-02 10:44:20,981 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-02 10:44:22,793 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3425, 1.7734, 2.2867, 1.8456, 3.3683, 4.9105, 4.7862, 5.2611], + device='cuda:3'), covar=tensor([0.1521, 0.3250, 0.2738, 0.1926, 0.0489, 0.0146, 0.0156, 0.0158], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0314, 0.0341, 0.0261, 0.0235, 0.0179, 0.0213, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:45:03,331 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3625, 3.8174, 3.9381, 3.9404, 1.5924, 3.7270, 3.2560, 3.6907], + device='cuda:3'), covar=tensor([0.1665, 0.0790, 0.0668, 0.0781, 0.5274, 0.0853, 0.0718, 0.1105], + device='cuda:3'), in_proj_covar=tensor([0.0747, 0.0697, 0.0898, 0.0778, 0.0799, 0.0648, 0.0537, 0.0824], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 10:45:06,512 INFO [train.py:903] (3/4) Epoch 17, batch 6100, loss[loss=0.2122, simple_loss=0.2868, pruned_loss=0.06879, over 19753.00 frames. ], tot_loss[loss=0.2151, simple_loss=0.2928, pruned_loss=0.06873, over 3808746.51 frames. ], batch size: 51, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:45:06,913 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115348.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:45:34,913 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6441, 4.1751, 2.8409, 3.7691, 1.0329, 4.0881, 4.0680, 4.1462], + device='cuda:3'), covar=tensor([0.0579, 0.1187, 0.1812, 0.0803, 0.3970, 0.0732, 0.0793, 0.1116], + device='cuda:3'), in_proj_covar=tensor([0.0473, 0.0389, 0.0472, 0.0335, 0.0393, 0.0409, 0.0405, 0.0435], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:45:54,927 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115388.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:45:56,156 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115389.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:46:06,534 INFO [train.py:903] (3/4) Epoch 17, batch 6150, loss[loss=0.1935, simple_loss=0.2712, pruned_loss=0.05786, over 19411.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2932, pruned_loss=0.06907, over 3813276.16 frames. ], batch size: 48, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:46:15,597 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.636e+02 5.209e+02 6.440e+02 8.380e+02 1.538e+03, threshold=1.288e+03, percent-clipped=5.0 +2023-04-02 10:46:33,786 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 10:47:07,332 INFO [train.py:903] (3/4) Epoch 17, batch 6200, loss[loss=0.2015, simple_loss=0.2918, pruned_loss=0.05558, over 19648.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2935, pruned_loss=0.06928, over 3802161.53 frames. ], batch size: 55, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:47:07,487 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=115448.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:47:08,928 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115449.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:47:12,269 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8593, 0.9193, 0.8621, 0.7252, 0.7725, 0.7747, 0.0753, 0.2875], + device='cuda:3'), covar=tensor([0.0501, 0.0488, 0.0319, 0.0450, 0.0780, 0.0525, 0.1044, 0.0747], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0345, 0.0346, 0.0375, 0.0445, 0.0380, 0.0326, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 10:47:39,870 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115474.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:48:07,507 INFO [train.py:903] (3/4) Epoch 17, batch 6250, loss[loss=0.2054, simple_loss=0.2873, pruned_loss=0.06171, over 19688.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2938, pruned_loss=0.06913, over 3806971.75 frames. ], batch size: 53, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:48:16,577 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.081e+02 4.690e+02 5.769e+02 7.890e+02 2.007e+03, threshold=1.154e+03, percent-clipped=3.0 +2023-04-02 10:48:37,586 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 10:49:09,165 INFO [train.py:903] (3/4) Epoch 17, batch 6300, loss[loss=0.2077, simple_loss=0.2891, pruned_loss=0.06316, over 19593.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2942, pruned_loss=0.06885, over 3818495.01 frames. ], batch size: 61, lr: 4.81e-03, grad_scale: 8.0 +2023-04-02 10:49:27,741 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115563.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:50:12,479 INFO [train.py:903] (3/4) Epoch 17, batch 6350, loss[loss=0.1976, simple_loss=0.2821, pruned_loss=0.05653, over 19661.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2923, pruned_loss=0.0673, over 3834652.82 frames. ], batch size: 53, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:50:19,289 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 10:50:20,012 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115604.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:50:21,935 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.064e+02 4.833e+02 6.077e+02 8.091e+02 1.466e+03, threshold=1.215e+03, percent-clipped=5.0 +2023-04-02 10:50:50,668 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115629.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:51:13,882 INFO [train.py:903] (3/4) Epoch 17, batch 6400, loss[loss=0.2545, simple_loss=0.332, pruned_loss=0.08848, over 19510.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2917, pruned_loss=0.06723, over 3836576.94 frames. ], batch size: 64, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:52:15,080 INFO [train.py:903] (3/4) Epoch 17, batch 6450, loss[loss=0.1878, simple_loss=0.2683, pruned_loss=0.05362, over 19488.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2911, pruned_loss=0.06687, over 3833207.66 frames. ], batch size: 49, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:52:25,113 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.211e+02 4.829e+02 5.862e+02 7.962e+02 1.327e+03, threshold=1.172e+03, percent-clipped=3.0 +2023-04-02 10:52:41,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0630, 1.2869, 1.7347, 0.9265, 2.3331, 3.0892, 2.7281, 3.2525], + device='cuda:3'), covar=tensor([0.1568, 0.3658, 0.3122, 0.2569, 0.0588, 0.0198, 0.0258, 0.0269], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0313, 0.0341, 0.0260, 0.0235, 0.0179, 0.0211, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:52:58,106 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=115732.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:52:59,235 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=115733.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:53:01,411 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 10:53:16,380 INFO [train.py:903] (3/4) Epoch 17, batch 6500, loss[loss=0.2096, simple_loss=0.277, pruned_loss=0.07104, over 19733.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2916, pruned_loss=0.06724, over 3835672.30 frames. ], batch size: 45, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:53:24,003 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 10:54:18,490 INFO [train.py:903] (3/4) Epoch 17, batch 6550, loss[loss=0.2777, simple_loss=0.3447, pruned_loss=0.1054, over 19655.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2929, pruned_loss=0.06781, over 3837230.08 frames. ], batch size: 58, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:54:28,764 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.723e+02 5.150e+02 6.522e+02 8.804e+02 2.234e+03, threshold=1.304e+03, percent-clipped=7.0 +2023-04-02 10:54:43,123 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115819.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:55:08,075 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9248, 4.5305, 3.3494, 4.0063, 2.0222, 4.3394, 4.3285, 4.4676], + device='cuda:3'), covar=tensor([0.0491, 0.0822, 0.1655, 0.0759, 0.2900, 0.0707, 0.0838, 0.1064], + device='cuda:3'), in_proj_covar=tensor([0.0477, 0.0393, 0.0475, 0.0335, 0.0397, 0.0412, 0.0406, 0.0435], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 10:55:15,231 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115844.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:55:18,691 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115847.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:55:19,546 INFO [train.py:903] (3/4) Epoch 17, batch 6600, loss[loss=0.2532, simple_loss=0.3247, pruned_loss=0.09089, over 19569.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2936, pruned_loss=0.06838, over 3815223.75 frames. ], batch size: 61, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:55:19,907 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:56:19,810 INFO [train.py:903] (3/4) Epoch 17, batch 6650, loss[loss=0.2161, simple_loss=0.2989, pruned_loss=0.06669, over 19626.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2922, pruned_loss=0.06759, over 3819177.44 frames. ], batch size: 57, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:56:30,892 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.015e+02 4.833e+02 5.946e+02 8.225e+02 1.682e+03, threshold=1.189e+03, percent-clipped=7.0 +2023-04-02 10:56:35,576 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115910.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 10:57:08,368 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1261, 1.2057, 1.7169, 1.1638, 2.4130, 3.4249, 3.0930, 3.5839], + device='cuda:3'), covar=tensor([0.1617, 0.3779, 0.3134, 0.2430, 0.0593, 0.0179, 0.0213, 0.0246], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0314, 0.0341, 0.0259, 0.0235, 0.0178, 0.0210, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 10:57:21,972 INFO [train.py:903] (3/4) Epoch 17, batch 6700, loss[loss=0.2115, simple_loss=0.2908, pruned_loss=0.06613, over 19727.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2913, pruned_loss=0.06716, over 3830578.95 frames. ], batch size: 51, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:58:05,025 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-04-02 10:58:20,393 INFO [train.py:903] (3/4) Epoch 17, batch 6750, loss[loss=0.2309, simple_loss=0.3128, pruned_loss=0.07451, over 19657.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.2916, pruned_loss=0.06771, over 3836480.28 frames. ], batch size: 55, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 10:58:31,538 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.004e+02 5.304e+02 6.320e+02 7.514e+02 1.971e+03, threshold=1.264e+03, percent-clipped=7.0 +2023-04-02 10:58:59,781 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116032.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 10:59:17,210 INFO [train.py:903] (3/4) Epoch 17, batch 6800, loss[loss=0.2193, simple_loss=0.2906, pruned_loss=0.07395, over 19675.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2918, pruned_loss=0.06768, over 3845642.50 frames. ], batch size: 53, lr: 4.80e-03, grad_scale: 8.0 +2023-04-02 11:00:02,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 11:00:03,440 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 11:00:07,138 INFO [train.py:903] (3/4) Epoch 18, batch 0, loss[loss=0.2097, simple_loss=0.2905, pruned_loss=0.06441, over 19695.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2905, pruned_loss=0.06441, over 19695.00 frames. ], batch size: 60, lr: 4.66e-03, grad_scale: 8.0 +2023-04-02 11:00:07,139 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 11:00:18,780 INFO [train.py:937] (3/4) Epoch 18, validation: loss=0.1712, simple_loss=0.2722, pruned_loss=0.03505, over 944034.00 frames. +2023-04-02 11:00:18,781 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 11:00:32,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 11:00:51,654 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116103.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:00:52,780 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116104.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:00:55,572 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.457e+02 4.972e+02 6.494e+02 8.085e+02 1.604e+03, threshold=1.299e+03, percent-clipped=1.0 +2023-04-02 11:01:14,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5104, 2.2743, 2.1413, 2.6772, 2.4853, 2.1238, 2.1492, 2.3346], + device='cuda:3'), covar=tensor([0.0970, 0.1560, 0.1475, 0.0994, 0.1282, 0.0518, 0.1214, 0.0741], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0353, 0.0303, 0.0245, 0.0297, 0.0246, 0.0294, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:01:15,459 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116123.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:01:18,718 INFO [train.py:903] (3/4) Epoch 18, batch 50, loss[loss=0.2218, simple_loss=0.3041, pruned_loss=0.06979, over 19604.00 frames. ], tot_loss[loss=0.2213, simple_loss=0.2989, pruned_loss=0.07185, over 872454.18 frames. ], batch size: 57, lr: 4.66e-03, grad_scale: 8.0 +2023-04-02 11:01:21,415 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116128.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:01:23,545 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116129.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:01:30,196 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116134.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:01:47,382 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.53 vs. limit=2.0 +2023-04-02 11:01:52,481 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 11:01:55,424 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.43 vs. limit=5.0 +2023-04-02 11:02:21,175 INFO [train.py:903] (3/4) Epoch 18, batch 100, loss[loss=0.2182, simple_loss=0.2965, pruned_loss=0.06993, over 19163.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2955, pruned_loss=0.07013, over 1524878.98 frames. ], batch size: 69, lr: 4.66e-03, grad_scale: 8.0 +2023-04-02 11:02:32,231 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 11:02:58,314 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.917e+02 4.838e+02 6.090e+02 7.458e+02 2.009e+03, threshold=1.218e+03, percent-clipped=2.0 +2023-04-02 11:03:21,606 INFO [train.py:903] (3/4) Epoch 18, batch 150, loss[loss=0.2825, simple_loss=0.3551, pruned_loss=0.105, over 18814.00 frames. ], tot_loss[loss=0.2168, simple_loss=0.2951, pruned_loss=0.06923, over 2041580.82 frames. ], batch size: 74, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:03:56,113 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116254.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 11:04:20,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 11:04:21,605 INFO [train.py:903] (3/4) Epoch 18, batch 200, loss[loss=0.2295, simple_loss=0.2982, pruned_loss=0.08039, over 12655.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2957, pruned_loss=0.07039, over 2413731.51 frames. ], batch size: 136, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:05:01,398 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.017e+02 4.750e+02 5.613e+02 7.153e+02 1.890e+03, threshold=1.123e+03, percent-clipped=2.0 +2023-04-02 11:05:24,092 INFO [train.py:903] (3/4) Epoch 18, batch 250, loss[loss=0.1654, simple_loss=0.2461, pruned_loss=0.04231, over 19738.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.2941, pruned_loss=0.06908, over 2726762.80 frames. ], batch size: 46, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:05:43,851 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116341.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:06:18,099 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116369.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 11:06:25,436 INFO [train.py:903] (3/4) Epoch 18, batch 300, loss[loss=0.2535, simple_loss=0.3342, pruned_loss=0.08635, over 19631.00 frames. ], tot_loss[loss=0.2178, simple_loss=0.2955, pruned_loss=0.07007, over 2969494.12 frames. ], batch size: 57, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:06:25,600 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116376.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:07:03,942 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.432e+02 5.241e+02 6.705e+02 8.261e+02 1.478e+03, threshold=1.341e+03, percent-clipped=3.0 +2023-04-02 11:07:28,467 INFO [train.py:903] (3/4) Epoch 18, batch 350, loss[loss=0.2118, simple_loss=0.2889, pruned_loss=0.06732, over 19579.00 frames. ], tot_loss[loss=0.215, simple_loss=0.293, pruned_loss=0.06847, over 3150694.13 frames. ], batch size: 52, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:07:31,286 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2330, 1.3299, 1.4170, 1.5321, 1.1602, 1.4347, 1.4836, 1.3363], + device='cuda:3'), covar=tensor([0.2764, 0.2153, 0.1308, 0.1478, 0.2375, 0.1289, 0.3058, 0.2181], + device='cuda:3'), in_proj_covar=tensor([0.0856, 0.0909, 0.0686, 0.0911, 0.0837, 0.0773, 0.0813, 0.0754], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 11:07:33,269 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 11:07:44,264 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-04-02 11:08:16,153 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116464.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:08:19,446 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116467.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:08:29,497 INFO [train.py:903] (3/4) Epoch 18, batch 400, loss[loss=0.2105, simple_loss=0.2875, pruned_loss=0.06671, over 19765.00 frames. ], tot_loss[loss=0.2151, simple_loss=0.293, pruned_loss=0.06855, over 3310424.77 frames. ], batch size: 54, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:08:31,857 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116478.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:08:48,759 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116491.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:09:08,770 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.099e+02 4.874e+02 5.859e+02 7.069e+02 1.370e+03, threshold=1.172e+03, percent-clipped=1.0 +2023-04-02 11:09:27,691 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116523.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:09:31,055 INFO [train.py:903] (3/4) Epoch 18, batch 450, loss[loss=0.2322, simple_loss=0.3137, pruned_loss=0.07537, over 17510.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2926, pruned_loss=0.06801, over 3422891.98 frames. ], batch size: 101, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:10:06,943 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 11:10:08,080 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 11:10:36,072 INFO [train.py:903] (3/4) Epoch 18, batch 500, loss[loss=0.214, simple_loss=0.2904, pruned_loss=0.06882, over 19784.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2921, pruned_loss=0.06756, over 3519583.78 frames. ], batch size: 54, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:10:43,278 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116582.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:10:57,199 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116593.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:11:13,493 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.162e+02 5.090e+02 6.291e+02 8.243e+02 1.843e+03, threshold=1.258e+03, percent-clipped=5.0 +2023-04-02 11:11:38,155 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116625.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 11:11:38,804 INFO [train.py:903] (3/4) Epoch 18, batch 550, loss[loss=0.2126, simple_loss=0.2895, pruned_loss=0.0679, over 19601.00 frames. ], tot_loss[loss=0.2146, simple_loss=0.2929, pruned_loss=0.06812, over 3590849.76 frames. ], batch size: 50, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:11:51,138 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116636.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:12:07,798 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116650.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 11:12:07,846 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2558, 1.3698, 1.6941, 1.5067, 2.1529, 1.8518, 2.0619, 0.8856], + device='cuda:3'), covar=tensor([0.2601, 0.4304, 0.2519, 0.1997, 0.1680, 0.2468, 0.1722, 0.4565], + device='cuda:3'), in_proj_covar=tensor([0.0512, 0.0612, 0.0671, 0.0463, 0.0611, 0.0516, 0.0648, 0.0522], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 11:12:41,002 INFO [train.py:903] (3/4) Epoch 18, batch 600, loss[loss=0.2282, simple_loss=0.3088, pruned_loss=0.07381, over 18827.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.2928, pruned_loss=0.06837, over 3642412.37 frames. ], batch size: 74, lr: 4.65e-03, grad_scale: 8.0 +2023-04-02 11:12:51,638 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116685.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:13:18,881 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.849e+02 4.880e+02 6.214e+02 8.095e+02 1.532e+03, threshold=1.243e+03, percent-clipped=4.0 +2023-04-02 11:13:21,159 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 11:13:28,763 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9763, 1.8145, 1.7139, 2.0750, 1.8136, 1.8348, 1.6809, 1.9456], + device='cuda:3'), covar=tensor([0.0969, 0.1425, 0.1357, 0.0938, 0.1264, 0.0503, 0.1215, 0.0704], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0356, 0.0302, 0.0247, 0.0299, 0.0247, 0.0296, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:13:37,825 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116722.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:13:42,029 INFO [train.py:903] (3/4) Epoch 18, batch 650, loss[loss=0.2859, simple_loss=0.3594, pruned_loss=0.1062, over 18844.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2934, pruned_loss=0.06882, over 3676036.56 frames. ], batch size: 74, lr: 4.64e-03, grad_scale: 8.0 +2023-04-02 11:13:55,007 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0222, 1.3185, 1.5055, 1.5107, 2.6524, 1.1536, 2.2275, 2.9821], + device='cuda:3'), covar=tensor([0.0615, 0.2670, 0.2682, 0.1765, 0.0724, 0.2298, 0.1275, 0.0368], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0354, 0.0374, 0.0340, 0.0362, 0.0345, 0.0359, 0.0384], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:14:08,750 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116747.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:14:34,416 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.35 vs. limit=5.0 +2023-04-02 11:14:38,823 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116772.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:14:43,006 INFO [train.py:903] (3/4) Epoch 18, batch 700, loss[loss=0.2216, simple_loss=0.2999, pruned_loss=0.07166, over 17387.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2931, pruned_loss=0.06892, over 3701861.59 frames. ], batch size: 101, lr: 4.64e-03, grad_scale: 8.0 +2023-04-02 11:15:15,683 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116800.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:15:23,659 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.148e+02 4.855e+02 5.777e+02 7.000e+02 1.472e+03, threshold=1.155e+03, percent-clipped=1.0 +2023-04-02 11:15:24,934 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116808.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:15:27,850 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 11:15:39,746 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116820.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:15:47,502 INFO [train.py:903] (3/4) Epoch 18, batch 750, loss[loss=0.1632, simple_loss=0.2479, pruned_loss=0.03925, over 19586.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2923, pruned_loss=0.0683, over 3733071.32 frames. ], batch size: 52, lr: 4.64e-03, grad_scale: 8.0 +2023-04-02 11:16:03,714 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116838.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:16:16,556 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116849.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:16:35,164 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116863.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:16:39,890 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116867.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:16:50,651 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116874.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:16:52,571 INFO [train.py:903] (3/4) Epoch 18, batch 800, loss[loss=0.1866, simple_loss=0.2716, pruned_loss=0.05081, over 19755.00 frames. ], tot_loss[loss=0.214, simple_loss=0.2922, pruned_loss=0.0679, over 3766659.75 frames. ], batch size: 51, lr: 4.64e-03, grad_scale: 8.0 +2023-04-02 11:17:06,580 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 11:17:32,326 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.570e+02 5.513e+02 6.326e+02 7.669e+02 1.889e+03, threshold=1.265e+03, percent-clipped=5.0 +2023-04-02 11:17:51,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116923.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:17:55,033 INFO [train.py:903] (3/4) Epoch 18, batch 850, loss[loss=0.2167, simple_loss=0.3008, pruned_loss=0.06633, over 19670.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.292, pruned_loss=0.0677, over 3783302.40 frames. ], batch size: 59, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:18:12,871 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0972, 1.2228, 1.5527, 0.6156, 2.0274, 2.4991, 2.1386, 2.6143], + device='cuda:3'), covar=tensor([0.1390, 0.3701, 0.3143, 0.2499, 0.0540, 0.0247, 0.0335, 0.0324], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0312, 0.0339, 0.0256, 0.0231, 0.0177, 0.0209, 0.0236], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 11:18:48,135 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 11:18:56,389 INFO [train.py:903] (3/4) Epoch 18, batch 900, loss[loss=0.2087, simple_loss=0.2932, pruned_loss=0.06214, over 19586.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2936, pruned_loss=0.06868, over 3795415.29 frames. ], batch size: 61, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:19:01,375 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=116980.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:19:04,004 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116982.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:19:38,590 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.269e+02 4.671e+02 5.637e+02 7.276e+02 1.422e+03, threshold=1.127e+03, percent-clipped=2.0 +2023-04-02 11:20:00,615 INFO [train.py:903] (3/4) Epoch 18, batch 950, loss[loss=0.2102, simple_loss=0.2943, pruned_loss=0.06308, over 19647.00 frames. ], tot_loss[loss=0.2138, simple_loss=0.2926, pruned_loss=0.06753, over 3802708.03 frames. ], batch size: 58, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:20:02,906 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 11:20:38,650 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117056.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:20:51,139 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=117066.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:21:03,205 INFO [train.py:903] (3/4) Epoch 18, batch 1000, loss[loss=0.216, simple_loss=0.2941, pruned_loss=0.06895, over 19771.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.293, pruned_loss=0.06804, over 3798917.32 frames. ], batch size: 54, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:21:11,176 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117081.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:21:27,370 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117095.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:21:43,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.648e+02 4.999e+02 6.181e+02 7.829e+02 2.221e+03, threshold=1.236e+03, percent-clipped=4.0 +2023-04-02 11:21:56,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 11:22:07,201 INFO [train.py:903] (3/4) Epoch 18, batch 1050, loss[loss=0.2715, simple_loss=0.3351, pruned_loss=0.104, over 19274.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.2949, pruned_loss=0.06888, over 3811332.49 frames. ], batch size: 66, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:22:40,249 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 11:22:55,006 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=117164.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:23:09,065 INFO [train.py:903] (3/4) Epoch 18, batch 1100, loss[loss=0.2758, simple_loss=0.3337, pruned_loss=0.109, over 13122.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.294, pruned_loss=0.06844, over 3821253.97 frames. ], batch size: 136, lr: 4.64e-03, grad_scale: 4.0 +2023-04-02 11:23:13,182 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117179.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:23:15,446 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117181.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:23:25,752 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2650, 2.1053, 1.9246, 1.8025, 1.6063, 1.7827, 0.7147, 1.3129], + device='cuda:3'), covar=tensor([0.0505, 0.0593, 0.0514, 0.0754, 0.1104, 0.0947, 0.1185, 0.0971], + device='cuda:3'), in_proj_covar=tensor([0.0351, 0.0347, 0.0347, 0.0376, 0.0449, 0.0382, 0.0328, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 11:23:44,528 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117204.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:23:49,510 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.961e+02 5.150e+02 6.225e+02 7.900e+02 1.283e+03, threshold=1.245e+03, percent-clipped=2.0 +2023-04-02 11:24:11,116 INFO [train.py:903] (3/4) Epoch 18, batch 1150, loss[loss=0.1998, simple_loss=0.2829, pruned_loss=0.05836, over 19674.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2933, pruned_loss=0.06809, over 3833508.98 frames. ], batch size: 53, lr: 4.63e-03, grad_scale: 4.0 +2023-04-02 11:24:27,286 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117238.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:24:28,297 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117239.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:24:39,746 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4665, 2.2335, 1.7190, 2.0533, 0.8758, 2.1950, 2.1322, 2.1867], + device='cuda:3'), covar=tensor([0.1201, 0.1234, 0.1771, 0.0846, 0.2673, 0.0966, 0.1044, 0.1278], + device='cuda:3'), in_proj_covar=tensor([0.0479, 0.0393, 0.0475, 0.0337, 0.0395, 0.0413, 0.0406, 0.0436], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:24:58,359 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117263.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:25:14,039 INFO [train.py:903] (3/4) Epoch 18, batch 1200, loss[loss=0.1939, simple_loss=0.2759, pruned_loss=0.05598, over 19596.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.293, pruned_loss=0.06814, over 3829878.41 frames. ], batch size: 52, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:25:19,010 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117279.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:25:31,276 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7674, 1.4948, 1.4207, 1.7881, 1.4746, 1.5709, 1.4569, 1.6860], + device='cuda:3'), covar=tensor([0.1038, 0.1310, 0.1453, 0.0937, 0.1213, 0.0575, 0.1313, 0.0780], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0355, 0.0302, 0.0248, 0.0299, 0.0247, 0.0295, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:25:50,718 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 11:25:54,166 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.294e+02 4.775e+02 5.862e+02 7.562e+02 1.280e+03, threshold=1.172e+03, percent-clipped=1.0 +2023-04-02 11:26:12,980 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-02 11:26:18,172 INFO [train.py:903] (3/4) Epoch 18, batch 1250, loss[loss=0.1915, simple_loss=0.2698, pruned_loss=0.05655, over 19578.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.293, pruned_loss=0.06829, over 3823232.22 frames. ], batch size: 52, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:26:43,761 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:26:49,497 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117351.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:26:58,440 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2446, 2.1017, 1.9567, 1.8392, 1.5604, 1.7899, 0.5834, 1.1401], + device='cuda:3'), covar=tensor([0.0522, 0.0576, 0.0402, 0.0671, 0.1104, 0.0798, 0.1155, 0.0933], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0347, 0.0348, 0.0376, 0.0449, 0.0381, 0.0328, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 11:27:20,894 INFO [train.py:903] (3/4) Epoch 18, batch 1300, loss[loss=0.2261, simple_loss=0.3049, pruned_loss=0.0736, over 17302.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.294, pruned_loss=0.06876, over 3812782.05 frames. ], batch size: 101, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:27:21,304 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117376.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:27:48,194 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6771, 1.6571, 1.6634, 2.3333, 1.5894, 2.1253, 2.1347, 1.9060], + device='cuda:3'), covar=tensor([0.0886, 0.0931, 0.1002, 0.0727, 0.0931, 0.0717, 0.0828, 0.0696], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0221, 0.0223, 0.0240, 0.0226, 0.0209, 0.0186, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 11:28:01,376 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.655e+02 5.418e+02 6.594e+02 8.357e+02 1.516e+03, threshold=1.319e+03, percent-clipped=5.0 +2023-04-02 11:28:22,249 INFO [train.py:903] (3/4) Epoch 18, batch 1350, loss[loss=0.2436, simple_loss=0.3105, pruned_loss=0.0884, over 19533.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2939, pruned_loss=0.06868, over 3823266.12 frames. ], batch size: 54, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:28:37,217 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117437.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:29:07,761 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117462.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:29:21,700 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.49 vs. limit=2.0 +2023-04-02 11:29:24,518 INFO [train.py:903] (3/4) Epoch 18, batch 1400, loss[loss=0.2152, simple_loss=0.292, pruned_loss=0.06925, over 19752.00 frames. ], tot_loss[loss=0.2166, simple_loss=0.2948, pruned_loss=0.0692, over 3830289.79 frames. ], batch size: 51, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:30:04,548 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.042e+02 5.443e+02 6.741e+02 8.791e+02 2.167e+03, threshold=1.348e+03, percent-clipped=5.0 +2023-04-02 11:30:28,246 INFO [train.py:903] (3/4) Epoch 18, batch 1450, loss[loss=0.1967, simple_loss=0.2852, pruned_loss=0.05414, over 19622.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.295, pruned_loss=0.06941, over 3809314.27 frames. ], batch size: 61, lr: 4.63e-03, grad_scale: 4.0 +2023-04-02 11:30:29,437 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 11:30:40,064 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117535.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:31:11,240 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117560.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:31:30,927 INFO [train.py:903] (3/4) Epoch 18, batch 1500, loss[loss=0.2486, simple_loss=0.3182, pruned_loss=0.0895, over 19636.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2951, pruned_loss=0.06968, over 3812567.88 frames. ], batch size: 58, lr: 4.63e-03, grad_scale: 4.0 +2023-04-02 11:31:39,057 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=117583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:31:53,561 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-04-02 11:32:11,893 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 4.657e+02 6.062e+02 7.787e+02 1.498e+03, threshold=1.212e+03, percent-clipped=2.0 +2023-04-02 11:32:32,137 INFO [train.py:903] (3/4) Epoch 18, batch 1550, loss[loss=0.2543, simple_loss=0.3305, pruned_loss=0.08906, over 19518.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2948, pruned_loss=0.06974, over 3820107.02 frames. ], batch size: 54, lr: 4.63e-03, grad_scale: 4.0 +2023-04-02 11:33:34,528 INFO [train.py:903] (3/4) Epoch 18, batch 1600, loss[loss=0.1749, simple_loss=0.2543, pruned_loss=0.04773, over 19773.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.294, pruned_loss=0.0691, over 3822028.29 frames. ], batch size: 49, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:33:54,974 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=117691.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:34:01,911 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 11:34:03,402 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117698.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:34:15,613 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.549e+02 4.964e+02 5.954e+02 7.051e+02 1.393e+03, threshold=1.191e+03, percent-clipped=4.0 +2023-04-02 11:34:37,826 INFO [train.py:903] (3/4) Epoch 18, batch 1650, loss[loss=0.2007, simple_loss=0.2644, pruned_loss=0.06852, over 19078.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2932, pruned_loss=0.06827, over 3834841.76 frames. ], batch size: 42, lr: 4.63e-03, grad_scale: 8.0 +2023-04-02 11:34:44,904 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9487, 1.1790, 1.5426, 0.6575, 1.9944, 2.3542, 2.0600, 2.4672], + device='cuda:3'), covar=tensor([0.1568, 0.3663, 0.3162, 0.2704, 0.0685, 0.0312, 0.0344, 0.0389], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0309, 0.0338, 0.0257, 0.0230, 0.0177, 0.0209, 0.0235], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 11:35:04,237 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3303, 2.3680, 2.5658, 3.2502, 2.3172, 3.0851, 2.6232, 2.4226], + device='cuda:3'), covar=tensor([0.4115, 0.4034, 0.1672, 0.2293, 0.4366, 0.1956, 0.4486, 0.3090], + device='cuda:3'), in_proj_covar=tensor([0.0858, 0.0911, 0.0689, 0.0910, 0.0837, 0.0775, 0.0818, 0.0757], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 11:35:39,482 INFO [train.py:903] (3/4) Epoch 18, batch 1700, loss[loss=0.2149, simple_loss=0.2947, pruned_loss=0.06757, over 19544.00 frames. ], tot_loss[loss=0.2146, simple_loss=0.293, pruned_loss=0.06814, over 3831284.50 frames. ], batch size: 56, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:36:16,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117806.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:36:19,945 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.954e+02 5.098e+02 6.258e+02 7.253e+02 1.524e+03, threshold=1.252e+03, percent-clipped=2.0 +2023-04-02 11:36:21,118 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 11:36:40,234 INFO [train.py:903] (3/4) Epoch 18, batch 1750, loss[loss=0.1952, simple_loss=0.279, pruned_loss=0.05575, over 19484.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.292, pruned_loss=0.06768, over 3837288.24 frames. ], batch size: 49, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:37:14,809 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-02 11:37:43,036 INFO [train.py:903] (3/4) Epoch 18, batch 1800, loss[loss=0.1755, simple_loss=0.2664, pruned_loss=0.04227, over 19737.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2911, pruned_loss=0.06729, over 3847498.86 frames. ], batch size: 51, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:38:23,379 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.984e+02 4.803e+02 6.041e+02 7.952e+02 1.877e+03, threshold=1.208e+03, percent-clipped=3.0 +2023-04-02 11:38:42,054 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 11:38:45,247 INFO [train.py:903] (3/4) Epoch 18, batch 1850, loss[loss=0.1716, simple_loss=0.2451, pruned_loss=0.04904, over 19369.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2911, pruned_loss=0.06752, over 3838044.90 frames. ], batch size: 47, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:39:18,895 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 11:39:19,318 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117954.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:39:47,632 INFO [train.py:903] (3/4) Epoch 18, batch 1900, loss[loss=0.2245, simple_loss=0.3069, pruned_loss=0.07103, over 19782.00 frames. ], tot_loss[loss=0.2133, simple_loss=0.2913, pruned_loss=0.06764, over 3820498.95 frames. ], batch size: 56, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:39:51,614 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117979.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:40:03,238 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 11:40:07,848 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 11:40:27,943 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.102e+02 5.120e+02 6.309e+02 7.973e+02 1.539e+03, threshold=1.262e+03, percent-clipped=6.0 +2023-04-02 11:40:34,564 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 11:40:48,202 INFO [train.py:903] (3/4) Epoch 18, batch 1950, loss[loss=0.2064, simple_loss=0.2792, pruned_loss=0.06687, over 19486.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2931, pruned_loss=0.06903, over 3809087.91 frames. ], batch size: 49, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:41:28,946 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118058.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:41:33,474 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118062.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:41:49,952 INFO [train.py:903] (3/4) Epoch 18, batch 2000, loss[loss=0.2455, simple_loss=0.316, pruned_loss=0.08751, over 17310.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.2935, pruned_loss=0.06957, over 3795869.43 frames. ], batch size: 101, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:42:05,400 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118087.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:42:11,297 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.61 vs. limit=5.0 +2023-04-02 11:42:31,795 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.263e+02 5.367e+02 6.632e+02 8.072e+02 1.503e+03, threshold=1.326e+03, percent-clipped=5.0 +2023-04-02 11:42:46,431 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 11:42:54,015 INFO [train.py:903] (3/4) Epoch 18, batch 2050, loss[loss=0.2072, simple_loss=0.2888, pruned_loss=0.06287, over 19777.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2933, pruned_loss=0.06906, over 3799177.45 frames. ], batch size: 54, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:43:06,254 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 11:43:07,433 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 11:43:26,101 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 11:43:55,356 INFO [train.py:903] (3/4) Epoch 18, batch 2100, loss[loss=0.2445, simple_loss=0.3219, pruned_loss=0.08358, over 19660.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.2939, pruned_loss=0.06913, over 3807551.81 frames. ], batch size: 60, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:44:07,021 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118186.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:44:21,260 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 11:44:36,112 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.283e+02 5.186e+02 6.689e+02 8.493e+02 1.656e+03, threshold=1.338e+03, percent-clipped=6.0 +2023-04-02 11:44:44,889 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 11:44:56,295 INFO [train.py:903] (3/4) Epoch 18, batch 2150, loss[loss=0.2311, simple_loss=0.3007, pruned_loss=0.08077, over 19578.00 frames. ], tot_loss[loss=0.215, simple_loss=0.2931, pruned_loss=0.0684, over 3821198.12 frames. ], batch size: 52, lr: 4.62e-03, grad_scale: 8.0 +2023-04-02 11:45:39,512 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6021, 1.2203, 1.2642, 1.5041, 1.1479, 1.4090, 1.2380, 1.4629], + device='cuda:3'), covar=tensor([0.1151, 0.1254, 0.1619, 0.1016, 0.1270, 0.0601, 0.1492, 0.0830], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0350, 0.0299, 0.0245, 0.0296, 0.0244, 0.0292, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:45:57,805 INFO [train.py:903] (3/4) Epoch 18, batch 2200, loss[loss=0.2163, simple_loss=0.2981, pruned_loss=0.0673, over 18678.00 frames. ], tot_loss[loss=0.2157, simple_loss=0.2938, pruned_loss=0.06884, over 3827987.17 frames. ], batch size: 74, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:46:13,121 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118287.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:46:37,341 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118307.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:46:39,292 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.599e+02 5.366e+02 7.280e+02 9.450e+02 2.114e+03, threshold=1.456e+03, percent-clipped=6.0 +2023-04-02 11:47:01,193 INFO [train.py:903] (3/4) Epoch 18, batch 2250, loss[loss=0.2244, simple_loss=0.3002, pruned_loss=0.0743, over 19685.00 frames. ], tot_loss[loss=0.2172, simple_loss=0.2953, pruned_loss=0.06949, over 3842357.27 frames. ], batch size: 59, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:47:27,313 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:47:47,809 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6502, 4.2304, 2.6781, 3.8198, 0.8779, 4.0788, 4.0628, 4.1140], + device='cuda:3'), covar=tensor([0.0609, 0.1115, 0.2037, 0.0785, 0.4225, 0.0701, 0.0844, 0.1012], + device='cuda:3'), in_proj_covar=tensor([0.0478, 0.0395, 0.0479, 0.0340, 0.0399, 0.0416, 0.0411, 0.0441], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:48:02,750 INFO [train.py:903] (3/4) Epoch 18, batch 2300, loss[loss=0.201, simple_loss=0.2759, pruned_loss=0.06307, over 19731.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2944, pruned_loss=0.06919, over 3840369.35 frames. ], batch size: 51, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:48:15,082 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 11:48:22,177 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8859, 1.4182, 1.6122, 1.5282, 2.6550, 1.2730, 2.2432, 2.8112], + device='cuda:3'), covar=tensor([0.0538, 0.2356, 0.2260, 0.1684, 0.0624, 0.2068, 0.1864, 0.0451], + device='cuda:3'), in_proj_covar=tensor([0.0398, 0.0359, 0.0377, 0.0344, 0.0367, 0.0347, 0.0367, 0.0388], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:48:35,792 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118402.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:48:44,940 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.375e+02 5.253e+02 6.243e+02 7.561e+02 1.558e+03, threshold=1.249e+03, percent-clipped=2.0 +2023-04-02 11:49:05,625 INFO [train.py:903] (3/4) Epoch 18, batch 2350, loss[loss=0.2172, simple_loss=0.2993, pruned_loss=0.06756, over 19673.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.2938, pruned_loss=0.06887, over 3833930.12 frames. ], batch size: 60, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:49:12,834 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4350, 2.2019, 1.7290, 1.4914, 2.0802, 1.3561, 1.3951, 1.8461], + device='cuda:3'), covar=tensor([0.0913, 0.0777, 0.0985, 0.0838, 0.0497, 0.1211, 0.0729, 0.0479], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0314, 0.0336, 0.0262, 0.0247, 0.0334, 0.0293, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:49:46,248 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 11:50:02,118 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 11:50:06,554 INFO [train.py:903] (3/4) Epoch 18, batch 2400, loss[loss=0.1813, simple_loss=0.2678, pruned_loss=0.04742, over 19757.00 frames. ], tot_loss[loss=0.2166, simple_loss=0.2942, pruned_loss=0.06944, over 3812009.42 frames. ], batch size: 54, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:50:48,686 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.320e+02 4.821e+02 5.593e+02 7.730e+02 1.750e+03, threshold=1.119e+03, percent-clipped=3.0 +2023-04-02 11:50:58,533 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118517.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:51:10,676 INFO [train.py:903] (3/4) Epoch 18, batch 2450, loss[loss=0.2559, simple_loss=0.3261, pruned_loss=0.09287, over 13530.00 frames. ], tot_loss[loss=0.215, simple_loss=0.2931, pruned_loss=0.06845, over 3812807.65 frames. ], batch size: 137, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:51:16,262 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118530.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:51:57,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4290, 1.5005, 1.9036, 1.5271, 2.8441, 3.6560, 3.4227, 3.8340], + device='cuda:3'), covar=tensor([0.1564, 0.3522, 0.3031, 0.2302, 0.0589, 0.0191, 0.0190, 0.0241], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0310, 0.0340, 0.0257, 0.0232, 0.0178, 0.0210, 0.0237], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 11:52:12,593 INFO [train.py:903] (3/4) Epoch 18, batch 2500, loss[loss=0.2231, simple_loss=0.3079, pruned_loss=0.06919, over 19700.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2928, pruned_loss=0.06789, over 3817554.71 frames. ], batch size: 59, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:52:53,545 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.320e+02 5.081e+02 6.040e+02 7.266e+02 1.380e+03, threshold=1.208e+03, percent-clipped=1.0 +2023-04-02 11:53:13,672 INFO [train.py:903] (3/4) Epoch 18, batch 2550, loss[loss=0.2385, simple_loss=0.316, pruned_loss=0.08048, over 19599.00 frames. ], tot_loss[loss=0.2153, simple_loss=0.2937, pruned_loss=0.06847, over 3821731.65 frames. ], batch size: 57, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:53:19,829 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118631.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:53:37,342 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.82 vs. limit=5.0 +2023-04-02 11:53:37,914 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118645.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:53:43,519 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118649.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:53:45,538 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118651.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:54:06,484 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 11:54:15,889 INFO [train.py:903] (3/4) Epoch 18, batch 2600, loss[loss=0.1986, simple_loss=0.2862, pruned_loss=0.05552, over 19774.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2932, pruned_loss=0.06862, over 3823353.63 frames. ], batch size: 54, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:54:35,487 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118691.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:54:56,094 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.395e+02 4.742e+02 5.749e+02 7.114e+02 1.367e+03, threshold=1.150e+03, percent-clipped=3.0 +2023-04-02 11:55:16,572 INFO [train.py:903] (3/4) Epoch 18, batch 2650, loss[loss=0.2191, simple_loss=0.3022, pruned_loss=0.06798, over 18173.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2941, pruned_loss=0.06898, over 3818760.30 frames. ], batch size: 83, lr: 4.61e-03, grad_scale: 8.0 +2023-04-02 11:55:33,742 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 11:55:42,205 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118746.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:56:06,207 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118766.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:56:15,267 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118773.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:56:18,245 INFO [train.py:903] (3/4) Epoch 18, batch 2700, loss[loss=0.1944, simple_loss=0.2682, pruned_loss=0.06033, over 19367.00 frames. ], tot_loss[loss=0.2146, simple_loss=0.2928, pruned_loss=0.06822, over 3823109.45 frames. ], batch size: 47, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 11:56:44,803 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118798.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:56:49,107 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.88 vs. limit=5.0 +2023-04-02 11:56:55,808 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118806.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:56:59,035 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.246e+02 5.095e+02 6.154e+02 8.195e+02 1.746e+03, threshold=1.231e+03, percent-clipped=5.0 +2023-04-02 11:57:05,253 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1960, 5.5787, 2.9670, 4.9145, 0.9661, 5.7115, 5.6211, 5.6426], + device='cuda:3'), covar=tensor([0.0378, 0.0749, 0.1831, 0.0652, 0.4105, 0.0490, 0.0724, 0.0913], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0394, 0.0480, 0.0341, 0.0402, 0.0419, 0.0413, 0.0445], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:57:06,522 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3475, 1.2800, 0.9647, 1.2125, 1.1327, 1.0633, 0.9478, 1.1648], + device='cuda:3'), covar=tensor([0.1217, 0.1229, 0.1861, 0.1095, 0.1283, 0.1130, 0.1844, 0.1104], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0353, 0.0299, 0.0248, 0.0298, 0.0246, 0.0295, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:57:20,473 INFO [train.py:903] (3/4) Epoch 18, batch 2750, loss[loss=0.2399, simple_loss=0.3222, pruned_loss=0.07885, over 18806.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.293, pruned_loss=0.06826, over 3817731.58 frames. ], batch size: 74, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 11:58:15,368 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4380, 1.4843, 1.8422, 1.6132, 3.1187, 2.3503, 3.3212, 1.5688], + device='cuda:3'), covar=tensor([0.2572, 0.4445, 0.2784, 0.2066, 0.1442, 0.2234, 0.1450, 0.4166], + device='cuda:3'), in_proj_covar=tensor([0.0519, 0.0622, 0.0679, 0.0468, 0.0615, 0.0521, 0.0655, 0.0530], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 11:58:23,105 INFO [train.py:903] (3/4) Epoch 18, batch 2800, loss[loss=0.2086, simple_loss=0.293, pruned_loss=0.06212, over 19520.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2934, pruned_loss=0.06854, over 3830056.24 frames. ], batch size: 54, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 11:58:41,681 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2218, 1.2845, 1.2738, 1.0889, 1.0748, 1.1091, 0.0442, 0.3411], + device='cuda:3'), covar=tensor([0.0634, 0.0613, 0.0384, 0.0521, 0.1258, 0.0594, 0.1226, 0.1039], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0348, 0.0347, 0.0374, 0.0448, 0.0380, 0.0329, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 11:58:44,038 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6751, 1.3760, 1.5281, 1.6172, 3.2724, 1.2923, 2.5897, 3.6616], + device='cuda:3'), covar=tensor([0.0426, 0.2676, 0.2684, 0.1790, 0.0649, 0.2278, 0.1037, 0.0232], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0357, 0.0378, 0.0342, 0.0365, 0.0345, 0.0366, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 11:58:54,250 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118901.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 11:59:03,985 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 4.804e+02 6.148e+02 8.494e+02 1.418e+03, threshold=1.230e+03, percent-clipped=5.0 +2023-04-02 11:59:24,481 INFO [train.py:903] (3/4) Epoch 18, batch 2850, loss[loss=0.2478, simple_loss=0.3235, pruned_loss=0.08604, over 19394.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2927, pruned_loss=0.06832, over 3825739.93 frames. ], batch size: 70, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 11:59:24,957 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118926.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:00:22,866 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 12:00:26,007 INFO [train.py:903] (3/4) Epoch 18, batch 2900, loss[loss=0.22, simple_loss=0.3075, pruned_loss=0.06625, over 19720.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2912, pruned_loss=0.06717, over 3820275.54 frames. ], batch size: 63, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:00:27,569 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1428, 2.8793, 2.1845, 2.2696, 2.0524, 2.4894, 1.0409, 2.0932], + device='cuda:3'), covar=tensor([0.0637, 0.0543, 0.0626, 0.0964, 0.0979, 0.0976, 0.1222, 0.0948], + device='cuda:3'), in_proj_covar=tensor([0.0351, 0.0349, 0.0348, 0.0375, 0.0449, 0.0381, 0.0330, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:00:46,052 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=118993.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:00:54,396 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-04-02 12:00:54,945 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3681, 1.4833, 1.9975, 1.7032, 2.7433, 2.4378, 3.1038, 1.5384], + device='cuda:3'), covar=tensor([0.2774, 0.4706, 0.2867, 0.2128, 0.2070, 0.2461, 0.1996, 0.4514], + device='cuda:3'), in_proj_covar=tensor([0.0520, 0.0625, 0.0682, 0.0469, 0.0618, 0.0524, 0.0659, 0.0533], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 12:00:56,064 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119002.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:00:58,541 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 12:01:05,223 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.034e+02 5.127e+02 5.908e+02 8.371e+02 2.467e+03, threshold=1.182e+03, percent-clipped=10.0 +2023-04-02 12:01:21,099 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119022.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:01:25,122 INFO [train.py:903] (3/4) Epoch 18, batch 2950, loss[loss=0.1668, simple_loss=0.2539, pruned_loss=0.03984, over 19849.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2907, pruned_loss=0.06682, over 3835217.46 frames. ], batch size: 52, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:01:26,625 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119027.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:01:50,472 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119047.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:02:08,484 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119062.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:02:24,526 INFO [train.py:903] (3/4) Epoch 18, batch 3000, loss[loss=0.243, simple_loss=0.3159, pruned_loss=0.08499, over 19092.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2908, pruned_loss=0.06718, over 3830281.27 frames. ], batch size: 75, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:02:24,526 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 12:02:37,004 INFO [train.py:937] (3/4) Epoch 18, validation: loss=0.1707, simple_loss=0.2711, pruned_loss=0.03521, over 944034.00 frames. +2023-04-02 12:02:37,005 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 12:02:40,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 12:02:50,811 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119087.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:02:57,672 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119093.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:03:17,132 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119108.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:03:17,948 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.980e+02 5.504e+02 6.538e+02 8.295e+02 4.074e+03, threshold=1.308e+03, percent-clipped=8.0 +2023-04-02 12:03:20,394 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119111.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:03:37,844 INFO [train.py:903] (3/4) Epoch 18, batch 3050, loss[loss=0.2291, simple_loss=0.3107, pruned_loss=0.07372, over 19648.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2911, pruned_loss=0.06743, over 3824836.61 frames. ], batch size: 58, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:04:37,433 INFO [train.py:903] (3/4) Epoch 18, batch 3100, loss[loss=0.1959, simple_loss=0.2869, pruned_loss=0.05247, over 17972.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2918, pruned_loss=0.06773, over 3823922.63 frames. ], batch size: 83, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:05:18,223 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.373e+02 4.828e+02 5.979e+02 7.400e+02 1.693e+03, threshold=1.196e+03, percent-clipped=2.0 +2023-04-02 12:05:39,402 INFO [train.py:903] (3/4) Epoch 18, batch 3150, loss[loss=0.2273, simple_loss=0.3057, pruned_loss=0.07441, over 19705.00 frames. ], tot_loss[loss=0.2132, simple_loss=0.2913, pruned_loss=0.06753, over 3827837.31 frames. ], batch size: 59, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:06:04,649 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 12:06:06,960 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 12:06:39,660 INFO [train.py:903] (3/4) Epoch 18, batch 3200, loss[loss=0.179, simple_loss=0.2687, pruned_loss=0.04466, over 19770.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.292, pruned_loss=0.0677, over 3827387.21 frames. ], batch size: 54, lr: 4.60e-03, grad_scale: 8.0 +2023-04-02 12:07:09,083 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1587, 2.0809, 1.8588, 1.6777, 1.6163, 1.7191, 0.5210, 0.9977], + device='cuda:3'), covar=tensor([0.0513, 0.0515, 0.0422, 0.0675, 0.1068, 0.0796, 0.1131, 0.0967], + device='cuda:3'), in_proj_covar=tensor([0.0349, 0.0345, 0.0345, 0.0371, 0.0446, 0.0378, 0.0326, 0.0333], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:07:18,202 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.241e+02 4.710e+02 5.964e+02 8.137e+02 2.705e+03, threshold=1.193e+03, percent-clipped=4.0 +2023-04-02 12:07:39,012 INFO [train.py:903] (3/4) Epoch 18, batch 3250, loss[loss=0.2637, simple_loss=0.324, pruned_loss=0.1017, over 13295.00 frames. ], tot_loss[loss=0.2148, simple_loss=0.2926, pruned_loss=0.06846, over 3817592.10 frames. ], batch size: 136, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:08:24,488 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119364.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:08:37,659 INFO [train.py:903] (3/4) Epoch 18, batch 3300, loss[loss=0.202, simple_loss=0.2793, pruned_loss=0.06234, over 19625.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2925, pruned_loss=0.06814, over 3831347.57 frames. ], batch size: 50, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:08:42,299 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 12:08:50,158 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3264, 3.7214, 4.0538, 4.1549, 1.6725, 3.8556, 3.2577, 3.5065], + device='cuda:3'), covar=tensor([0.2380, 0.1541, 0.1017, 0.1179, 0.7049, 0.1726, 0.1221, 0.2053], + device='cuda:3'), in_proj_covar=tensor([0.0758, 0.0701, 0.0907, 0.0790, 0.0808, 0.0662, 0.0549, 0.0841], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 12:08:53,576 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119389.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:09:17,496 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.159e+02 5.369e+02 6.623e+02 8.148e+02 1.799e+03, threshold=1.325e+03, percent-clipped=5.0 +2023-04-02 12:09:37,819 INFO [train.py:903] (3/4) Epoch 18, batch 3350, loss[loss=0.1852, simple_loss=0.2598, pruned_loss=0.05526, over 19734.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2926, pruned_loss=0.0681, over 3848244.57 frames. ], batch size: 45, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:09:39,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7642, 4.3048, 4.4905, 4.5139, 1.6026, 4.2076, 3.6760, 4.2003], + device='cuda:3'), covar=tensor([0.1542, 0.0686, 0.0582, 0.0619, 0.5882, 0.0735, 0.0610, 0.1160], + device='cuda:3'), in_proj_covar=tensor([0.0756, 0.0698, 0.0903, 0.0787, 0.0805, 0.0661, 0.0547, 0.0838], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 12:09:46,648 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3697, 1.5172, 1.7293, 1.5785, 2.7962, 1.3627, 2.4927, 3.2818], + device='cuda:3'), covar=tensor([0.0713, 0.3071, 0.2836, 0.2226, 0.1163, 0.2692, 0.1458, 0.0424], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0358, 0.0377, 0.0343, 0.0366, 0.0347, 0.0368, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:09:50,673 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=119437.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:10:11,914 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=119455.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:10:37,461 INFO [train.py:903] (3/4) Epoch 18, batch 3400, loss[loss=0.2048, simple_loss=0.2929, pruned_loss=0.0583, over 19588.00 frames. ], tot_loss[loss=0.2139, simple_loss=0.2922, pruned_loss=0.06775, over 3844805.40 frames. ], batch size: 61, lr: 4.59e-03, grad_scale: 4.0 +2023-04-02 12:11:18,514 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.251e+02 4.894e+02 5.911e+02 7.983e+02 1.559e+03, threshold=1.182e+03, percent-clipped=2.0 +2023-04-02 12:11:37,562 INFO [train.py:903] (3/4) Epoch 18, batch 3450, loss[loss=0.184, simple_loss=0.2705, pruned_loss=0.04875, over 19679.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.293, pruned_loss=0.06823, over 3834260.13 frames. ], batch size: 53, lr: 4.59e-03, grad_scale: 4.0 +2023-04-02 12:11:43,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 12:11:45,635 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119532.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:12:09,611 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119552.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:12:31,744 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119570.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:12:38,211 INFO [train.py:903] (3/4) Epoch 18, batch 3500, loss[loss=0.2686, simple_loss=0.344, pruned_loss=0.09656, over 19672.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2932, pruned_loss=0.06835, over 3828647.81 frames. ], batch size: 60, lr: 4.59e-03, grad_scale: 4.0 +2023-04-02 12:12:47,662 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.82 vs. limit=5.0 +2023-04-02 12:12:51,541 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4437, 1.4564, 1.7482, 1.6594, 2.6075, 2.2427, 2.6868, 1.1062], + device='cuda:3'), covar=tensor([0.2242, 0.3988, 0.2434, 0.1747, 0.1408, 0.2032, 0.1391, 0.4107], + device='cuda:3'), in_proj_covar=tensor([0.0520, 0.0620, 0.0681, 0.0469, 0.0616, 0.0522, 0.0657, 0.0531], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 12:12:55,123 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 12:13:20,027 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.151e+02 4.884e+02 6.516e+02 7.989e+02 1.670e+03, threshold=1.303e+03, percent-clipped=4.0 +2023-04-02 12:13:39,059 INFO [train.py:903] (3/4) Epoch 18, batch 3550, loss[loss=0.1992, simple_loss=0.2705, pruned_loss=0.06396, over 19746.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2932, pruned_loss=0.06834, over 3834746.19 frames. ], batch size: 47, lr: 4.59e-03, grad_scale: 4.0 +2023-04-02 12:14:39,530 INFO [train.py:903] (3/4) Epoch 18, batch 3600, loss[loss=0.2125, simple_loss=0.2794, pruned_loss=0.07275, over 19762.00 frames. ], tot_loss[loss=0.2146, simple_loss=0.2931, pruned_loss=0.06805, over 3827978.68 frames. ], batch size: 47, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:14:47,966 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119683.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:15:20,525 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.297e+02 4.936e+02 6.005e+02 7.350e+02 2.220e+03, threshold=1.201e+03, percent-clipped=3.0 +2023-04-02 12:15:39,380 INFO [train.py:903] (3/4) Epoch 18, batch 3650, loss[loss=0.247, simple_loss=0.3157, pruned_loss=0.08922, over 19769.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.293, pruned_loss=0.06801, over 3822307.01 frames. ], batch size: 54, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:16:40,014 INFO [train.py:903] (3/4) Epoch 18, batch 3700, loss[loss=0.2277, simple_loss=0.321, pruned_loss=0.0672, over 19677.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2935, pruned_loss=0.06814, over 3809751.76 frames. ], batch size: 55, lr: 4.59e-03, grad_scale: 8.0 +2023-04-02 12:16:58,975 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-02 12:17:19,681 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119808.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:17:21,521 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.455e+02 5.017e+02 5.995e+02 7.636e+02 1.424e+03, threshold=1.199e+03, percent-clipped=3.0 +2023-04-02 12:17:32,214 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3870, 1.3978, 1.8957, 1.5413, 2.7596, 3.6424, 3.4407, 3.8348], + device='cuda:3'), covar=tensor([0.1538, 0.3679, 0.3177, 0.2321, 0.0634, 0.0269, 0.0211, 0.0277], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0309, 0.0340, 0.0258, 0.0233, 0.0178, 0.0211, 0.0239], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 12:17:40,475 INFO [train.py:903] (3/4) Epoch 18, batch 3750, loss[loss=0.1825, simple_loss=0.2564, pruned_loss=0.05426, over 19756.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.294, pruned_loss=0.06886, over 3790101.49 frames. ], batch size: 45, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:17:40,915 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119826.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:17:48,730 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119833.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:17:58,523 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6916, 1.8527, 1.7289, 2.6477, 1.8163, 2.4746, 1.9018, 1.5068], + device='cuda:3'), covar=tensor([0.4876, 0.4117, 0.2670, 0.2688, 0.4340, 0.2223, 0.5892, 0.4871], + device='cuda:3'), in_proj_covar=tensor([0.0870, 0.0926, 0.0696, 0.0921, 0.0850, 0.0789, 0.0827, 0.0766], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 12:18:10,666 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119851.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:18:20,895 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1364, 5.1244, 5.9235, 5.9888, 1.8708, 5.6280, 4.7327, 5.5243], + device='cuda:3'), covar=tensor([0.1486, 0.0841, 0.0496, 0.0566, 0.6293, 0.0731, 0.0557, 0.1148], + device='cuda:3'), in_proj_covar=tensor([0.0757, 0.0696, 0.0905, 0.0790, 0.0806, 0.0656, 0.0546, 0.0839], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 12:18:33,223 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5009, 1.7142, 2.0998, 1.9134, 3.0847, 2.7018, 3.3714, 1.4992], + device='cuda:3'), covar=tensor([0.2404, 0.4103, 0.2497, 0.1776, 0.1665, 0.2011, 0.1659, 0.4093], + device='cuda:3'), in_proj_covar=tensor([0.0518, 0.0618, 0.0678, 0.0467, 0.0611, 0.0519, 0.0654, 0.0530], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:18:39,735 INFO [train.py:903] (3/4) Epoch 18, batch 3800, loss[loss=0.2319, simple_loss=0.3119, pruned_loss=0.076, over 19619.00 frames. ], tot_loss[loss=0.2154, simple_loss=0.2934, pruned_loss=0.06864, over 3806763.39 frames. ], batch size: 57, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:18:40,735 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=119876.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:19:12,066 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 12:19:22,041 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.281e+02 5.499e+02 6.857e+02 8.596e+02 2.059e+03, threshold=1.371e+03, percent-clipped=8.0 +2023-04-02 12:19:41,469 INFO [train.py:903] (3/4) Epoch 18, batch 3850, loss[loss=0.2162, simple_loss=0.2977, pruned_loss=0.06739, over 18823.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.292, pruned_loss=0.06758, over 3814664.72 frames. ], batch size: 74, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:20:40,743 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-02 12:20:43,326 INFO [train.py:903] (3/4) Epoch 18, batch 3900, loss[loss=0.2074, simple_loss=0.2914, pruned_loss=0.06173, over 19551.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2907, pruned_loss=0.06656, over 3829770.45 frames. ], batch size: 61, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:21:02,839 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119991.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:21:26,391 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.199e+02 4.757e+02 5.722e+02 7.251e+02 1.425e+03, threshold=1.144e+03, percent-clipped=2.0 +2023-04-02 12:21:45,628 INFO [train.py:903] (3/4) Epoch 18, batch 3950, loss[loss=0.2234, simple_loss=0.3061, pruned_loss=0.07036, over 19085.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.2903, pruned_loss=0.06617, over 3836336.99 frames. ], batch size: 69, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:21:47,751 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=120027.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:21:50,875 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 12:22:05,642 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0044, 3.6489, 2.4887, 3.2537, 1.0207, 3.5978, 3.4385, 3.5839], + device='cuda:3'), covar=tensor([0.0741, 0.1017, 0.1935, 0.0965, 0.3687, 0.0753, 0.0932, 0.1029], + device='cuda:3'), in_proj_covar=tensor([0.0480, 0.0392, 0.0479, 0.0341, 0.0396, 0.0415, 0.0408, 0.0442], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:22:47,956 INFO [train.py:903] (3/4) Epoch 18, batch 4000, loss[loss=0.2547, simple_loss=0.3299, pruned_loss=0.08973, over 17586.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2898, pruned_loss=0.06596, over 3838436.37 frames. ], batch size: 101, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:23:29,303 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.243e+02 4.630e+02 5.764e+02 7.444e+02 1.534e+03, threshold=1.153e+03, percent-clipped=2.0 +2023-04-02 12:23:36,025 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 12:23:49,506 INFO [train.py:903] (3/4) Epoch 18, batch 4050, loss[loss=0.2209, simple_loss=0.2968, pruned_loss=0.0725, over 19312.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.29, pruned_loss=0.06624, over 3829268.31 frames. ], batch size: 66, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:24:08,119 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=120142.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:24:47,334 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2575, 1.1606, 1.1558, 1.2918, 0.9974, 1.2456, 1.3322, 1.2103], + device='cuda:3'), covar=tensor([0.0940, 0.1069, 0.1141, 0.0722, 0.0886, 0.0872, 0.0814, 0.0838], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0224, 0.0244, 0.0227, 0.0208, 0.0188, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 12:24:49,140 INFO [train.py:903] (3/4) Epoch 18, batch 4100, loss[loss=0.2077, simple_loss=0.2912, pruned_loss=0.06211, over 17280.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2906, pruned_loss=0.06664, over 3834202.80 frames. ], batch size: 101, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:25:24,879 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 12:25:29,250 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.453e+02 5.087e+02 6.190e+02 7.903e+02 1.294e+03, threshold=1.238e+03, percent-clipped=4.0 +2023-04-02 12:25:48,080 INFO [train.py:903] (3/4) Epoch 18, batch 4150, loss[loss=0.2265, simple_loss=0.3087, pruned_loss=0.07215, over 19677.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2911, pruned_loss=0.06685, over 3830185.61 frames. ], batch size: 58, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:26:14,724 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=120247.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:26:46,375 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=120272.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:26:50,728 INFO [train.py:903] (3/4) Epoch 18, batch 4200, loss[loss=0.2077, simple_loss=0.2795, pruned_loss=0.06798, over 19381.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.292, pruned_loss=0.06754, over 3836067.17 frames. ], batch size: 48, lr: 4.58e-03, grad_scale: 8.0 +2023-04-02 12:26:57,176 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 12:27:06,726 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1472, 2.0733, 1.7612, 1.5719, 1.5241, 1.6192, 0.5580, 1.0398], + device='cuda:3'), covar=tensor([0.0508, 0.0595, 0.0455, 0.0772, 0.1210, 0.0966, 0.1222, 0.1008], + device='cuda:3'), in_proj_covar=tensor([0.0348, 0.0345, 0.0346, 0.0372, 0.0448, 0.0379, 0.0325, 0.0334], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:27:30,828 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.475e+02 4.968e+02 6.385e+02 7.954e+02 1.571e+03, threshold=1.277e+03, percent-clipped=3.0 +2023-04-02 12:27:51,846 INFO [train.py:903] (3/4) Epoch 18, batch 4250, loss[loss=0.2064, simple_loss=0.2769, pruned_loss=0.06798, over 19395.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2914, pruned_loss=0.06704, over 3839737.24 frames. ], batch size: 48, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:28:09,643 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 12:28:18,893 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 12:28:19,571 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.65 vs. limit=5.0 +2023-04-02 12:28:51,543 INFO [train.py:903] (3/4) Epoch 18, batch 4300, loss[loss=0.2594, simple_loss=0.3378, pruned_loss=0.09055, over 19342.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2917, pruned_loss=0.06703, over 3837918.95 frames. ], batch size: 66, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:29:19,875 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=120398.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:29:34,133 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.947e+02 5.084e+02 5.969e+02 7.631e+02 1.294e+03, threshold=1.194e+03, percent-clipped=1.0 +2023-04-02 12:29:46,157 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 12:29:49,849 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=120423.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:29:52,877 INFO [train.py:903] (3/4) Epoch 18, batch 4350, loss[loss=0.2199, simple_loss=0.3019, pruned_loss=0.06894, over 19671.00 frames. ], tot_loss[loss=0.214, simple_loss=0.293, pruned_loss=0.06755, over 3832847.54 frames. ], batch size: 60, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:29:57,442 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0105, 1.1454, 1.4691, 0.7386, 2.0183, 2.2174, 1.9921, 2.3191], + device='cuda:3'), covar=tensor([0.1491, 0.3564, 0.3034, 0.2508, 0.0733, 0.0390, 0.0367, 0.0420], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0311, 0.0339, 0.0258, 0.0234, 0.0179, 0.0211, 0.0238], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 12:30:20,710 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3018, 1.1303, 1.4642, 1.5015, 2.6516, 1.1020, 2.3408, 3.1499], + device='cuda:3'), covar=tensor([0.0688, 0.3422, 0.3070, 0.1993, 0.1240, 0.2791, 0.1240, 0.0464], + device='cuda:3'), in_proj_covar=tensor([0.0394, 0.0354, 0.0374, 0.0338, 0.0363, 0.0343, 0.0364, 0.0385], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:30:55,181 INFO [train.py:903] (3/4) Epoch 18, batch 4400, loss[loss=0.1966, simple_loss=0.2833, pruned_loss=0.05495, over 19317.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2924, pruned_loss=0.06746, over 3832018.03 frames. ], batch size: 66, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:31:18,744 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 12:31:28,328 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 12:31:35,104 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.222e+02 5.092e+02 6.306e+02 8.223e+02 1.474e+03, threshold=1.261e+03, percent-clipped=7.0 +2023-04-02 12:31:55,621 INFO [train.py:903] (3/4) Epoch 18, batch 4450, loss[loss=0.2308, simple_loss=0.3143, pruned_loss=0.07363, over 18140.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2929, pruned_loss=0.068, over 3829125.35 frames. ], batch size: 83, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:32:55,856 INFO [train.py:903] (3/4) Epoch 18, batch 4500, loss[loss=0.2477, simple_loss=0.3201, pruned_loss=0.08768, over 19506.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.292, pruned_loss=0.0674, over 3837202.19 frames. ], batch size: 64, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:33:37,594 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.517e+02 5.202e+02 6.117e+02 7.756e+02 2.072e+03, threshold=1.223e+03, percent-clipped=4.0 +2023-04-02 12:33:56,189 INFO [train.py:903] (3/4) Epoch 18, batch 4550, loss[loss=0.2193, simple_loss=0.2987, pruned_loss=0.06989, over 19776.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.292, pruned_loss=0.0677, over 3845545.77 frames. ], batch size: 56, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:34:05,863 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 12:34:29,497 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 12:34:56,266 INFO [train.py:903] (3/4) Epoch 18, batch 4600, loss[loss=0.2654, simple_loss=0.335, pruned_loss=0.09792, over 18778.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.294, pruned_loss=0.06927, over 3812008.88 frames. ], batch size: 74, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:35:17,310 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=120694.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:35:27,854 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0078, 1.9512, 1.9806, 2.7683, 1.9392, 2.5307, 2.4856, 2.4180], + device='cuda:3'), covar=tensor([0.0758, 0.0878, 0.0927, 0.0778, 0.0878, 0.0715, 0.0854, 0.0594], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0226, 0.0245, 0.0229, 0.0211, 0.0189, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 12:35:35,196 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.215e+02 4.865e+02 6.169e+02 7.995e+02 1.948e+03, threshold=1.234e+03, percent-clipped=9.0 +2023-04-02 12:35:55,082 INFO [train.py:903] (3/4) Epoch 18, batch 4650, loss[loss=0.2841, simple_loss=0.3408, pruned_loss=0.1137, over 12543.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.292, pruned_loss=0.06841, over 3825792.82 frames. ], batch size: 136, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:36:13,548 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 12:36:23,040 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 12:36:24,595 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 12:36:55,849 INFO [train.py:903] (3/4) Epoch 18, batch 4700, loss[loss=0.2083, simple_loss=0.2979, pruned_loss=0.05931, over 19351.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.292, pruned_loss=0.0676, over 3827302.96 frames. ], batch size: 70, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:37:18,766 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 12:37:37,347 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.471e+02 5.224e+02 6.544e+02 8.077e+02 2.112e+03, threshold=1.309e+03, percent-clipped=4.0 +2023-04-02 12:37:51,142 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4918, 1.3741, 1.4183, 1.7856, 1.5310, 1.6811, 1.5900, 1.5252], + device='cuda:3'), covar=tensor([0.0695, 0.0770, 0.0811, 0.0630, 0.0804, 0.0642, 0.0859, 0.0610], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0223, 0.0225, 0.0244, 0.0227, 0.0209, 0.0189, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 12:37:54,550 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-02 12:37:56,118 INFO [train.py:903] (3/4) Epoch 18, batch 4750, loss[loss=0.2958, simple_loss=0.3478, pruned_loss=0.1219, over 13495.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2931, pruned_loss=0.06867, over 3811310.29 frames. ], batch size: 137, lr: 4.57e-03, grad_scale: 8.0 +2023-04-02 12:38:57,283 INFO [train.py:903] (3/4) Epoch 18, batch 4800, loss[loss=0.2009, simple_loss=0.2763, pruned_loss=0.06272, over 19415.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2923, pruned_loss=0.06821, over 3822632.83 frames. ], batch size: 48, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:39:38,286 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.835e+02 5.063e+02 6.517e+02 8.294e+02 1.429e+03, threshold=1.303e+03, percent-clipped=1.0 +2023-04-02 12:39:57,799 INFO [train.py:903] (3/4) Epoch 18, batch 4850, loss[loss=0.228, simple_loss=0.3071, pruned_loss=0.07446, over 19307.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2915, pruned_loss=0.06792, over 3797516.35 frames. ], batch size: 66, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:40:19,482 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 12:40:38,945 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 12:40:44,430 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 12:40:45,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 12:40:55,301 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 12:40:57,465 INFO [train.py:903] (3/4) Epoch 18, batch 4900, loss[loss=0.2028, simple_loss=0.2906, pruned_loss=0.05745, over 19695.00 frames. ], tot_loss[loss=0.215, simple_loss=0.2925, pruned_loss=0.06873, over 3785908.63 frames. ], batch size: 59, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:41:15,405 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 12:41:38,440 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.463e+02 5.379e+02 6.693e+02 8.152e+02 1.326e+03, threshold=1.339e+03, percent-clipped=1.0 +2023-04-02 12:41:50,636 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121021.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:41:56,826 INFO [train.py:903] (3/4) Epoch 18, batch 4950, loss[loss=0.2109, simple_loss=0.2988, pruned_loss=0.0615, over 19423.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2933, pruned_loss=0.06934, over 3805197.41 frames. ], batch size: 70, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:42:11,752 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=121038.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:42:13,943 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 12:42:36,735 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 12:42:55,224 INFO [train.py:903] (3/4) Epoch 18, batch 5000, loss[loss=0.304, simple_loss=0.3529, pruned_loss=0.1275, over 13491.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2954, pruned_loss=0.0706, over 3801455.81 frames. ], batch size: 136, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:43:05,401 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 12:43:11,828 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6121, 1.4986, 1.5340, 2.1474, 1.5874, 1.9490, 1.9686, 1.7106], + device='cuda:3'), covar=tensor([0.0846, 0.0952, 0.1012, 0.0743, 0.0857, 0.0733, 0.0814, 0.0729], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0226, 0.0244, 0.0229, 0.0210, 0.0189, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 12:43:16,152 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 12:43:36,123 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.358e+02 5.049e+02 6.044e+02 8.198e+02 1.739e+03, threshold=1.209e+03, percent-clipped=8.0 +2023-04-02 12:43:55,696 INFO [train.py:903] (3/4) Epoch 18, batch 5050, loss[loss=0.2563, simple_loss=0.3194, pruned_loss=0.09659, over 13739.00 frames. ], tot_loss[loss=0.2177, simple_loss=0.2951, pruned_loss=0.07018, over 3818798.84 frames. ], batch size: 136, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:44:27,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121153.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:44:29,509 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 12:44:55,825 INFO [train.py:903] (3/4) Epoch 18, batch 5100, loss[loss=0.2065, simple_loss=0.2996, pruned_loss=0.05673, over 19533.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.296, pruned_loss=0.07059, over 3801083.59 frames. ], batch size: 64, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:45:03,048 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121182.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:45:05,057 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 12:45:07,244 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 12:45:12,795 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 12:45:37,241 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.930e+02 4.928e+02 6.063e+02 8.354e+02 2.244e+03, threshold=1.213e+03, percent-clipped=8.0 +2023-04-02 12:45:54,370 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4930, 2.2964, 1.7039, 1.4920, 2.1364, 1.3204, 1.3995, 1.9716], + device='cuda:3'), covar=tensor([0.0969, 0.0776, 0.0960, 0.0827, 0.0499, 0.1256, 0.0703, 0.0449], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0313, 0.0326, 0.0259, 0.0247, 0.0332, 0.0292, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:45:56,099 INFO [train.py:903] (3/4) Epoch 18, batch 5150, loss[loss=0.3051, simple_loss=0.3594, pruned_loss=0.1254, over 12989.00 frames. ], tot_loss[loss=0.2175, simple_loss=0.2951, pruned_loss=0.06989, over 3797448.40 frames. ], batch size: 136, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:46:05,287 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 12:46:40,455 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 12:46:55,813 INFO [train.py:903] (3/4) Epoch 18, batch 5200, loss[loss=0.2151, simple_loss=0.2966, pruned_loss=0.06679, over 19787.00 frames. ], tot_loss[loss=0.2166, simple_loss=0.2944, pruned_loss=0.06936, over 3808445.50 frames. ], batch size: 56, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:47:03,719 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5637, 1.9441, 2.0792, 2.0228, 3.3309, 1.7207, 2.8038, 3.4529], + device='cuda:3'), covar=tensor([0.0448, 0.2203, 0.2208, 0.1609, 0.0504, 0.2094, 0.1475, 0.0335], + device='cuda:3'), in_proj_covar=tensor([0.0394, 0.0355, 0.0376, 0.0341, 0.0365, 0.0347, 0.0366, 0.0387], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:47:08,967 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 12:47:36,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.274e+02 5.110e+02 6.332e+02 8.392e+02 3.036e+03, threshold=1.266e+03, percent-clipped=7.0 +2023-04-02 12:47:39,674 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1001, 1.3606, 1.6110, 1.2402, 2.5827, 3.5100, 3.2510, 3.7508], + device='cuda:3'), covar=tensor([0.1661, 0.3605, 0.3277, 0.2421, 0.0610, 0.0178, 0.0219, 0.0252], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0312, 0.0343, 0.0259, 0.0235, 0.0179, 0.0212, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 12:47:49,046 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 12:47:50,690 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4649, 1.5666, 1.8161, 1.7332, 2.6504, 2.4192, 2.7526, 1.2871], + device='cuda:3'), covar=tensor([0.2171, 0.3968, 0.2438, 0.1754, 0.1429, 0.1889, 0.1400, 0.3769], + device='cuda:3'), in_proj_covar=tensor([0.0517, 0.0621, 0.0680, 0.0466, 0.0612, 0.0519, 0.0653, 0.0531], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:47:54,555 INFO [train.py:903] (3/4) Epoch 18, batch 5250, loss[loss=0.2509, simple_loss=0.3336, pruned_loss=0.08405, over 19660.00 frames. ], tot_loss[loss=0.217, simple_loss=0.295, pruned_loss=0.06948, over 3809663.22 frames. ], batch size: 59, lr: 4.56e-03, grad_scale: 8.0 +2023-04-02 12:47:59,199 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121329.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 12:48:22,545 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5078, 1.6013, 1.8866, 1.8054, 2.8736, 2.4322, 2.9689, 1.3247], + device='cuda:3'), covar=tensor([0.2314, 0.4221, 0.2596, 0.1775, 0.1358, 0.1989, 0.1381, 0.4082], + device='cuda:3'), in_proj_covar=tensor([0.0517, 0.0620, 0.0680, 0.0466, 0.0611, 0.0518, 0.0653, 0.0531], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:48:40,812 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=121365.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:48:54,017 INFO [train.py:903] (3/4) Epoch 18, batch 5300, loss[loss=0.2367, simple_loss=0.319, pruned_loss=0.07714, over 19350.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2943, pruned_loss=0.06926, over 3820922.80 frames. ], batch size: 70, lr: 4.56e-03, grad_scale: 4.0 +2023-04-02 12:49:10,552 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 12:49:14,170 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3345, 2.3766, 2.5697, 3.2762, 2.3207, 3.0441, 2.7242, 2.5121], + device='cuda:3'), covar=tensor([0.4012, 0.3814, 0.1676, 0.2239, 0.4217, 0.1908, 0.4178, 0.2937], + device='cuda:3'), in_proj_covar=tensor([0.0870, 0.0928, 0.0695, 0.0923, 0.0851, 0.0788, 0.0826, 0.0764], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 12:49:33,524 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121409.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:49:35,381 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.570e+02 5.404e+02 6.878e+02 8.269e+02 2.518e+03, threshold=1.376e+03, percent-clipped=11.0 +2023-04-02 12:49:52,539 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121425.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:49:53,398 INFO [train.py:903] (3/4) Epoch 18, batch 5350, loss[loss=0.1937, simple_loss=0.2649, pruned_loss=0.06125, over 19773.00 frames. ], tot_loss[loss=0.2163, simple_loss=0.2943, pruned_loss=0.06917, over 3813763.12 frames. ], batch size: 47, lr: 4.55e-03, grad_scale: 4.0 +2023-04-02 12:50:03,610 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121434.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:50:26,223 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 12:50:53,958 INFO [train.py:903] (3/4) Epoch 18, batch 5400, loss[loss=0.1688, simple_loss=0.2468, pruned_loss=0.04537, over 18682.00 frames. ], tot_loss[loss=0.2167, simple_loss=0.2947, pruned_loss=0.06932, over 3824914.25 frames. ], batch size: 41, lr: 4.55e-03, grad_scale: 4.0 +2023-04-02 12:50:59,568 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121480.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:51:35,380 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.204e+02 5.292e+02 6.238e+02 7.901e+02 1.766e+03, threshold=1.248e+03, percent-clipped=4.0 +2023-04-02 12:51:54,666 INFO [train.py:903] (3/4) Epoch 18, batch 5450, loss[loss=0.2242, simple_loss=0.3106, pruned_loss=0.06888, over 18779.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2956, pruned_loss=0.06947, over 3826639.63 frames. ], batch size: 74, lr: 4.55e-03, grad_scale: 4.0 +2023-04-02 12:51:54,819 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=121526.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:51:56,524 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-04-02 12:52:54,549 INFO [train.py:903] (3/4) Epoch 18, batch 5500, loss[loss=0.2593, simple_loss=0.3318, pruned_loss=0.09335, over 19327.00 frames. ], tot_loss[loss=0.2161, simple_loss=0.2943, pruned_loss=0.06894, over 3826739.74 frames. ], batch size: 66, lr: 4.55e-03, grad_scale: 4.0 +2023-04-02 12:53:09,246 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9523, 1.8180, 1.6097, 2.0446, 1.7853, 1.7093, 1.5336, 1.8682], + device='cuda:3'), covar=tensor([0.1039, 0.1518, 0.1444, 0.1032, 0.1341, 0.0548, 0.1382, 0.0720], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0355, 0.0302, 0.0249, 0.0299, 0.0247, 0.0297, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:53:19,885 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 12:53:37,491 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.086e+02 4.754e+02 5.891e+02 7.661e+02 1.861e+03, threshold=1.178e+03, percent-clipped=6.0 +2023-04-02 12:53:55,066 INFO [train.py:903] (3/4) Epoch 18, batch 5550, loss[loss=0.1776, simple_loss=0.2619, pruned_loss=0.04665, over 19422.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2944, pruned_loss=0.06924, over 3824307.65 frames. ], batch size: 48, lr: 4.55e-03, grad_scale: 4.0 +2023-04-02 12:54:02,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 12:54:13,258 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121641.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:54:51,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 12:54:52,547 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=121673.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 12:54:55,719 INFO [train.py:903] (3/4) Epoch 18, batch 5600, loss[loss=0.2141, simple_loss=0.2942, pruned_loss=0.06701, over 19764.00 frames. ], tot_loss[loss=0.2158, simple_loss=0.2935, pruned_loss=0.06907, over 3822918.92 frames. ], batch size: 54, lr: 4.55e-03, grad_scale: 8.0 +2023-04-02 12:55:15,337 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8425, 4.3687, 2.6663, 3.8505, 1.3711, 4.3261, 4.1985, 4.3299], + device='cuda:3'), covar=tensor([0.0537, 0.0995, 0.2122, 0.0856, 0.3478, 0.0651, 0.0869, 0.1042], + device='cuda:3'), in_proj_covar=tensor([0.0484, 0.0391, 0.0479, 0.0340, 0.0392, 0.0418, 0.0408, 0.0439], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 12:55:37,561 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.280e+02 4.851e+02 5.729e+02 6.791e+02 1.695e+03, threshold=1.146e+03, percent-clipped=1.0 +2023-04-02 12:55:56,733 INFO [train.py:903] (3/4) Epoch 18, batch 5650, loss[loss=0.2078, simple_loss=0.3002, pruned_loss=0.05768, over 19614.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2934, pruned_loss=0.06881, over 3826800.64 frames. ], batch size: 57, lr: 4.55e-03, grad_scale: 8.0 +2023-04-02 12:56:08,932 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121736.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:56:37,728 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121761.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:56:44,587 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 12:56:48,206 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=121769.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:56:50,038 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.41 vs. limit=5.0 +2023-04-02 12:56:57,050 INFO [train.py:903] (3/4) Epoch 18, batch 5700, loss[loss=0.2456, simple_loss=0.3175, pruned_loss=0.08689, over 19647.00 frames. ], tot_loss[loss=0.2176, simple_loss=0.2951, pruned_loss=0.06998, over 3824799.50 frames. ], batch size: 60, lr: 4.55e-03, grad_scale: 8.0 +2023-04-02 12:57:10,998 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121788.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 12:57:39,339 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.986e+02 4.821e+02 5.859e+02 7.608e+02 1.521e+03, threshold=1.172e+03, percent-clipped=4.0 +2023-04-02 12:57:57,040 INFO [train.py:903] (3/4) Epoch 18, batch 5750, loss[loss=0.242, simple_loss=0.3168, pruned_loss=0.08357, over 19756.00 frames. ], tot_loss[loss=0.2173, simple_loss=0.2948, pruned_loss=0.06995, over 3817077.71 frames. ], batch size: 63, lr: 4.55e-03, grad_scale: 8.0 +2023-04-02 12:57:58,073 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 12:58:05,923 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 12:58:11,395 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 12:58:42,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9401, 1.8989, 1.7436, 1.5706, 1.4247, 1.5835, 0.4197, 0.8700], + device='cuda:3'), covar=tensor([0.0562, 0.0544, 0.0403, 0.0636, 0.1076, 0.0683, 0.1112, 0.0940], + device='cuda:3'), in_proj_covar=tensor([0.0351, 0.0347, 0.0349, 0.0375, 0.0451, 0.0379, 0.0329, 0.0336], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 12:58:43,873 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 12:58:57,512 INFO [train.py:903] (3/4) Epoch 18, batch 5800, loss[loss=0.2382, simple_loss=0.3143, pruned_loss=0.08105, over 19575.00 frames. ], tot_loss[loss=0.2168, simple_loss=0.2945, pruned_loss=0.06954, over 3827902.00 frames. ], batch size: 61, lr: 4.55e-03, grad_scale: 8.0 +2023-04-02 12:59:08,950 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121884.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:59:24,594 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121897.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:59:39,870 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.096e+02 5.174e+02 6.386e+02 8.157e+02 2.937e+03, threshold=1.277e+03, percent-clipped=7.0 +2023-04-02 12:59:54,634 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121922.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 12:59:59,627 INFO [train.py:903] (3/4) Epoch 18, batch 5850, loss[loss=0.2469, simple_loss=0.32, pruned_loss=0.08685, over 13677.00 frames. ], tot_loss[loss=0.2171, simple_loss=0.2945, pruned_loss=0.06986, over 3797736.81 frames. ], batch size: 136, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:00:06,630 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6397, 2.4234, 1.8354, 1.6636, 2.2527, 1.3862, 1.5095, 2.0398], + device='cuda:3'), covar=tensor([0.1065, 0.0788, 0.0970, 0.0817, 0.0509, 0.1277, 0.0732, 0.0480], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0313, 0.0327, 0.0257, 0.0245, 0.0331, 0.0291, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:00:51,267 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121968.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:01:00,889 INFO [train.py:903] (3/4) Epoch 18, batch 5900, loss[loss=0.2371, simple_loss=0.3249, pruned_loss=0.07468, over 19755.00 frames. ], tot_loss[loss=0.215, simple_loss=0.2933, pruned_loss=0.06837, over 3815167.97 frames. ], batch size: 63, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:01:02,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 13:01:23,289 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 13:01:39,600 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5955, 1.1976, 1.5008, 1.1771, 2.2516, 1.0485, 2.1732, 2.4297], + device='cuda:3'), covar=tensor([0.0675, 0.2697, 0.2603, 0.1704, 0.0847, 0.2045, 0.0964, 0.0496], + device='cuda:3'), in_proj_covar=tensor([0.0390, 0.0352, 0.0373, 0.0338, 0.0364, 0.0346, 0.0363, 0.0385], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:01:43,749 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.064e+02 4.950e+02 6.415e+02 8.144e+02 2.513e+03, threshold=1.283e+03, percent-clipped=4.0 +2023-04-02 13:02:01,670 INFO [train.py:903] (3/4) Epoch 18, batch 5950, loss[loss=0.2364, simple_loss=0.3101, pruned_loss=0.08135, over 19343.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.2947, pruned_loss=0.06913, over 3829109.23 frames. ], batch size: 66, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:02:24,950 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122044.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 13:02:42,896 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-04-02 13:02:53,782 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122069.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 13:03:01,811 INFO [train.py:903] (3/4) Epoch 18, batch 6000, loss[loss=0.2304, simple_loss=0.3038, pruned_loss=0.07846, over 19650.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.2951, pruned_loss=0.06938, over 3831991.24 frames. ], batch size: 55, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:03:01,812 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 13:03:14,300 INFO [train.py:937] (3/4) Epoch 18, validation: loss=0.1702, simple_loss=0.2706, pruned_loss=0.03489, over 944034.00 frames. +2023-04-02 13:03:14,302 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 13:03:57,709 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.792e+02 5.055e+02 6.106e+02 7.549e+02 1.634e+03, threshold=1.221e+03, percent-clipped=4.0 +2023-04-02 13:04:15,902 INFO [train.py:903] (3/4) Epoch 18, batch 6050, loss[loss=0.1981, simple_loss=0.2723, pruned_loss=0.0619, over 19723.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2939, pruned_loss=0.0686, over 3830205.87 frames. ], batch size: 51, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:04:28,039 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122135.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:04:33,704 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122140.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:05:04,404 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:05:18,133 INFO [train.py:903] (3/4) Epoch 18, batch 6100, loss[loss=0.2798, simple_loss=0.3364, pruned_loss=0.1116, over 17359.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2941, pruned_loss=0.06881, over 3831055.77 frames. ], batch size: 101, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:05:59,976 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.364e+02 5.309e+02 6.592e+02 8.010e+02 1.726e+03, threshold=1.318e+03, percent-clipped=1.0 +2023-04-02 13:06:03,477 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122213.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:06:17,058 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0095, 2.1125, 2.3654, 2.7418, 2.0498, 2.6734, 2.4641, 2.1669], + device='cuda:3'), covar=tensor([0.3939, 0.3654, 0.1706, 0.2148, 0.3858, 0.1826, 0.4065, 0.2981], + device='cuda:3'), in_proj_covar=tensor([0.0870, 0.0926, 0.0696, 0.0920, 0.0854, 0.0785, 0.0825, 0.0762], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 13:06:18,826 INFO [train.py:903] (3/4) Epoch 18, batch 6150, loss[loss=0.2015, simple_loss=0.2686, pruned_loss=0.06725, over 17813.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2933, pruned_loss=0.06881, over 3820330.76 frames. ], batch size: 39, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:06:38,374 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1642, 2.3334, 2.4296, 3.0738, 2.2881, 2.9593, 2.4454, 2.1098], + device='cuda:3'), covar=tensor([0.4375, 0.3974, 0.1892, 0.2523, 0.4376, 0.2045, 0.4823, 0.3515], + device='cuda:3'), in_proj_covar=tensor([0.0868, 0.0924, 0.0695, 0.0918, 0.0852, 0.0783, 0.0823, 0.0760], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 13:06:46,739 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 13:07:20,189 INFO [train.py:903] (3/4) Epoch 18, batch 6200, loss[loss=0.213, simple_loss=0.2963, pruned_loss=0.06486, over 19743.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2911, pruned_loss=0.0672, over 3831308.06 frames. ], batch size: 63, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:08:04,095 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.932e+02 4.549e+02 5.662e+02 6.707e+02 1.660e+03, threshold=1.132e+03, percent-clipped=3.0 +2023-04-02 13:08:05,435 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=122312.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:08:17,460 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 13:08:21,487 INFO [train.py:903] (3/4) Epoch 18, batch 6250, loss[loss=0.2154, simple_loss=0.2756, pruned_loss=0.0776, over 18674.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2908, pruned_loss=0.06713, over 3828698.66 frames. ], batch size: 41, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:08:35,738 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122337.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:08:53,453 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 13:09:23,683 INFO [train.py:903] (3/4) Epoch 18, batch 6300, loss[loss=0.2203, simple_loss=0.2926, pruned_loss=0.07407, over 19740.00 frames. ], tot_loss[loss=0.2118, simple_loss=0.2899, pruned_loss=0.06681, over 3825849.62 frames. ], batch size: 51, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:10:06,341 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.668e+02 4.826e+02 5.839e+02 7.420e+02 1.796e+03, threshold=1.168e+03, percent-clipped=3.0 +2023-04-02 13:10:25,342 INFO [train.py:903] (3/4) Epoch 18, batch 6350, loss[loss=0.2023, simple_loss=0.2899, pruned_loss=0.05736, over 19582.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2906, pruned_loss=0.06727, over 3810534.63 frames. ], batch size: 52, lr: 4.54e-03, grad_scale: 8.0 +2023-04-02 13:10:26,819 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122427.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:11:25,786 INFO [train.py:903] (3/4) Epoch 18, batch 6400, loss[loss=0.2586, simple_loss=0.3307, pruned_loss=0.09329, over 19652.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2909, pruned_loss=0.06733, over 3815087.57 frames. ], batch size: 58, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:11:29,350 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=122479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:11:30,830 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2147, 1.3026, 1.2516, 1.0521, 1.0699, 1.1206, 0.0894, 0.3468], + device='cuda:3'), covar=tensor([0.0680, 0.0639, 0.0436, 0.0560, 0.1394, 0.0631, 0.1285, 0.1068], + device='cuda:3'), in_proj_covar=tensor([0.0350, 0.0347, 0.0347, 0.0373, 0.0449, 0.0379, 0.0329, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 13:11:38,365 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122486.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:12:08,691 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.263e+02 4.572e+02 5.352e+02 7.001e+02 2.095e+03, threshold=1.070e+03, percent-clipped=5.0 +2023-04-02 13:12:27,168 INFO [train.py:903] (3/4) Epoch 18, batch 6450, loss[loss=0.2192, simple_loss=0.3095, pruned_loss=0.06445, over 19552.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2905, pruned_loss=0.06648, over 3821213.28 frames. ], batch size: 56, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:12:43,823 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-02 13:13:05,519 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=122557.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:13:09,884 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 13:13:27,310 INFO [train.py:903] (3/4) Epoch 18, batch 6500, loss[loss=0.2139, simple_loss=0.2969, pruned_loss=0.06546, over 19786.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2889, pruned_loss=0.06555, over 3820995.57 frames. ], batch size: 56, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:13:32,625 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 13:13:41,127 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.65 vs. limit=5.0 +2023-04-02 13:13:50,450 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122594.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:14:06,029 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2438, 2.1200, 2.0039, 1.7830, 1.5914, 1.8440, 0.5119, 1.2086], + device='cuda:3'), covar=tensor([0.0624, 0.0601, 0.0448, 0.0834, 0.1166, 0.0878, 0.1294, 0.0967], + device='cuda:3'), in_proj_covar=tensor([0.0352, 0.0347, 0.0348, 0.0375, 0.0451, 0.0380, 0.0330, 0.0335], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 13:14:10,392 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.928e+02 4.824e+02 5.508e+02 6.898e+02 1.222e+03, threshold=1.102e+03, percent-clipped=3.0 +2023-04-02 13:14:28,807 INFO [train.py:903] (3/4) Epoch 18, batch 6550, loss[loss=0.2213, simple_loss=0.31, pruned_loss=0.06626, over 19708.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2894, pruned_loss=0.06546, over 3825782.10 frames. ], batch size: 59, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:15:25,367 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122672.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:15:29,440 INFO [train.py:903] (3/4) Epoch 18, batch 6600, loss[loss=0.2119, simple_loss=0.2983, pruned_loss=0.0628, over 18801.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2895, pruned_loss=0.06568, over 3826622.45 frames. ], batch size: 74, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:15:35,566 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=122681.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:15:37,955 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122683.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:16:08,836 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122708.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:16:11,858 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.339e+02 4.976e+02 6.114e+02 8.137e+02 1.508e+03, threshold=1.223e+03, percent-clipped=10.0 +2023-04-02 13:16:29,835 INFO [train.py:903] (3/4) Epoch 18, batch 6650, loss[loss=0.1982, simple_loss=0.2886, pruned_loss=0.05388, over 19439.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2908, pruned_loss=0.06606, over 3829861.45 frames. ], batch size: 48, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:17:30,797 INFO [train.py:903] (3/4) Epoch 18, batch 6700, loss[loss=0.1852, simple_loss=0.2673, pruned_loss=0.05155, over 19847.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2905, pruned_loss=0.06596, over 3817115.35 frames. ], batch size: 52, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:17:55,704 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122796.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:18:12,038 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.354e+02 4.791e+02 5.844e+02 7.603e+02 1.856e+03, threshold=1.169e+03, percent-clipped=2.0 +2023-04-02 13:18:28,395 INFO [train.py:903] (3/4) Epoch 18, batch 6750, loss[loss=0.2445, simple_loss=0.3096, pruned_loss=0.08972, over 19496.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2914, pruned_loss=0.06679, over 3810198.41 frames. ], batch size: 49, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:18:33,091 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=122830.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:18:55,613 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122850.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:19:24,287 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122875.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:19:25,061 INFO [train.py:903] (3/4) Epoch 18, batch 6800, loss[loss=0.1989, simple_loss=0.2668, pruned_loss=0.06546, over 19309.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2909, pruned_loss=0.06666, over 3812809.79 frames. ], batch size: 44, lr: 4.53e-03, grad_scale: 8.0 +2023-04-02 13:19:40,837 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122890.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:20:09,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 13:20:09,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 13:20:12,597 INFO [train.py:903] (3/4) Epoch 19, batch 0, loss[loss=0.2186, simple_loss=0.2986, pruned_loss=0.06933, over 19782.00 frames. ], tot_loss[loss=0.2186, simple_loss=0.2986, pruned_loss=0.06933, over 19782.00 frames. ], batch size: 56, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:20:12,597 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 13:20:24,047 INFO [train.py:937] (3/4) Epoch 19, validation: loss=0.171, simple_loss=0.2713, pruned_loss=0.03533, over 944034.00 frames. +2023-04-02 13:20:24,048 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 13:20:32,696 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.525e+02 4.987e+02 6.075e+02 7.792e+02 1.350e+03, threshold=1.215e+03, percent-clipped=4.0 +2023-04-02 13:20:37,457 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 13:20:53,733 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122928.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:20:55,940 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0688, 5.1182, 5.9452, 5.9254, 1.9705, 5.5684, 4.7252, 5.5470], + device='cuda:3'), covar=tensor([0.1573, 0.0743, 0.0514, 0.0545, 0.6049, 0.0691, 0.0592, 0.1089], + device='cuda:3'), in_proj_covar=tensor([0.0751, 0.0696, 0.0898, 0.0796, 0.0803, 0.0652, 0.0542, 0.0826], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 13:21:14,005 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122945.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:21:17,738 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-04-02 13:21:24,691 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122953.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:21:25,465 INFO [train.py:903] (3/4) Epoch 19, batch 50, loss[loss=0.1689, simple_loss=0.2488, pruned_loss=0.04447, over 19743.00 frames. ], tot_loss[loss=0.2118, simple_loss=0.2895, pruned_loss=0.06709, over 864950.14 frames. ], batch size: 45, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:22:04,202 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 13:22:27,933 INFO [train.py:903] (3/4) Epoch 19, batch 100, loss[loss=0.2415, simple_loss=0.3024, pruned_loss=0.09034, over 19720.00 frames. ], tot_loss[loss=0.2132, simple_loss=0.2912, pruned_loss=0.06764, over 1526678.71 frames. ], batch size: 45, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:22:35,841 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.196e+02 4.975e+02 5.926e+02 8.151e+02 1.966e+03, threshold=1.185e+03, percent-clipped=7.0 +2023-04-02 13:22:40,154 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 13:22:51,561 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0130, 1.2631, 1.6288, 0.8993, 2.3684, 3.0621, 2.7946, 3.2708], + device='cuda:3'), covar=tensor([0.1648, 0.3592, 0.3240, 0.2583, 0.0573, 0.0205, 0.0231, 0.0274], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0311, 0.0341, 0.0259, 0.0234, 0.0179, 0.0211, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 13:23:25,708 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123052.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:23:27,454 INFO [train.py:903] (3/4) Epoch 19, batch 150, loss[loss=0.2305, simple_loss=0.3106, pruned_loss=0.0752, over 19319.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.294, pruned_loss=0.06935, over 2035179.40 frames. ], batch size: 70, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:23:55,479 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123077.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:24:27,285 INFO [train.py:903] (3/4) Epoch 19, batch 200, loss[loss=0.3056, simple_loss=0.3571, pruned_loss=0.1271, over 19736.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2936, pruned_loss=0.06877, over 2434424.89 frames. ], batch size: 63, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:24:27,696 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3448, 2.0713, 1.6079, 1.3759, 1.9304, 1.3198, 1.3145, 1.8001], + device='cuda:3'), covar=tensor([0.0885, 0.0719, 0.0974, 0.0779, 0.0443, 0.1180, 0.0576, 0.0396], + device='cuda:3'), in_proj_covar=tensor([0.0296, 0.0310, 0.0327, 0.0256, 0.0243, 0.0331, 0.0288, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:24:29,573 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 13:24:35,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.274e+02 5.223e+02 5.868e+02 7.012e+02 1.944e+03, threshold=1.174e+03, percent-clipped=7.0 +2023-04-02 13:25:27,092 INFO [train.py:903] (3/4) Epoch 19, batch 250, loss[loss=0.2036, simple_loss=0.2806, pruned_loss=0.06331, over 19733.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2928, pruned_loss=0.06799, over 2755605.89 frames. ], batch size: 51, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:26:25,153 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123201.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:26:28,233 INFO [train.py:903] (3/4) Epoch 19, batch 300, loss[loss=0.2298, simple_loss=0.3071, pruned_loss=0.07621, over 19272.00 frames. ], tot_loss[loss=0.214, simple_loss=0.2921, pruned_loss=0.06797, over 2994266.97 frames. ], batch size: 66, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:26:37,167 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.064e+02 4.833e+02 6.226e+02 7.852e+02 1.722e+03, threshold=1.245e+03, percent-clipped=4.0 +2023-04-02 13:26:55,621 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123226.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:27:04,614 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=123234.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:27:29,487 INFO [train.py:903] (3/4) Epoch 19, batch 350, loss[loss=0.2251, simple_loss=0.3033, pruned_loss=0.07352, over 19607.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2912, pruned_loss=0.06727, over 3182840.35 frames. ], batch size: 57, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:27:35,244 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 13:28:30,167 INFO [train.py:903] (3/4) Epoch 19, batch 400, loss[loss=0.2013, simple_loss=0.2845, pruned_loss=0.05902, over 19576.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2912, pruned_loss=0.06671, over 3320145.48 frames. ], batch size: 52, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:28:37,938 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.570e+02 5.151e+02 6.306e+02 7.944e+02 1.366e+03, threshold=1.261e+03, percent-clipped=2.0 +2023-04-02 13:28:55,716 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0685, 1.3105, 1.5889, 1.0022, 2.3060, 2.9903, 2.7438, 3.2404], + device='cuda:3'), covar=tensor([0.1605, 0.3589, 0.3334, 0.2486, 0.0613, 0.0232, 0.0271, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0310, 0.0341, 0.0258, 0.0235, 0.0179, 0.0211, 0.0242], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 13:29:04,100 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123332.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:29:24,968 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123349.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:29:30,465 INFO [train.py:903] (3/4) Epoch 19, batch 450, loss[loss=0.2062, simple_loss=0.2725, pruned_loss=0.07, over 19784.00 frames. ], tot_loss[loss=0.2132, simple_loss=0.2921, pruned_loss=0.06714, over 3436814.39 frames. ], batch size: 46, lr: 4.40e-03, grad_scale: 8.0 +2023-04-02 13:30:04,831 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 13:30:05,743 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 13:30:30,933 INFO [train.py:903] (3/4) Epoch 19, batch 500, loss[loss=0.1954, simple_loss=0.2702, pruned_loss=0.0603, over 19615.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2915, pruned_loss=0.06721, over 3534176.94 frames. ], batch size: 50, lr: 4.40e-03, grad_scale: 16.0 +2023-04-02 13:30:39,754 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.237e+02 5.352e+02 7.180e+02 8.361e+02 2.088e+03, threshold=1.436e+03, percent-clipped=3.0 +2023-04-02 13:31:21,112 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123446.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:31:30,114 INFO [train.py:903] (3/4) Epoch 19, batch 550, loss[loss=0.2332, simple_loss=0.3127, pruned_loss=0.07687, over 19502.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2925, pruned_loss=0.06746, over 3595468.89 frames. ], batch size: 64, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:32:30,274 INFO [train.py:903] (3/4) Epoch 19, batch 600, loss[loss=0.2167, simple_loss=0.2993, pruned_loss=0.06707, over 19605.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2912, pruned_loss=0.06725, over 3655905.50 frames. ], batch size: 57, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:32:39,906 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.253e+02 4.960e+02 5.982e+02 8.370e+02 1.865e+03, threshold=1.196e+03, percent-clipped=4.0 +2023-04-02 13:33:12,186 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1646, 1.2394, 1.6118, 1.0038, 2.3152, 3.0353, 2.6990, 3.2053], + device='cuda:3'), covar=tensor([0.1555, 0.3724, 0.3274, 0.2492, 0.0612, 0.0225, 0.0260, 0.0298], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0312, 0.0343, 0.0259, 0.0237, 0.0179, 0.0212, 0.0243], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 13:33:14,022 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 13:33:30,842 INFO [train.py:903] (3/4) Epoch 19, batch 650, loss[loss=0.2284, simple_loss=0.3132, pruned_loss=0.07185, over 19624.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2919, pruned_loss=0.06763, over 3691478.80 frames. ], batch size: 57, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:34:30,415 INFO [train.py:903] (3/4) Epoch 19, batch 700, loss[loss=0.2134, simple_loss=0.2997, pruned_loss=0.06351, over 19531.00 frames. ], tot_loss[loss=0.2139, simple_loss=0.2922, pruned_loss=0.06783, over 3713596.28 frames. ], batch size: 56, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:34:31,932 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123605.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:34:41,233 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.354e+02 5.460e+02 6.000e+02 7.674e+02 1.249e+03, threshold=1.200e+03, percent-clipped=2.0 +2023-04-02 13:35:02,792 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:35:31,845 INFO [train.py:903] (3/4) Epoch 19, batch 750, loss[loss=0.2188, simple_loss=0.299, pruned_loss=0.06931, over 19523.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2907, pruned_loss=0.06695, over 3745570.16 frames. ], batch size: 54, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:35:59,275 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=123676.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:36:34,034 INFO [train.py:903] (3/4) Epoch 19, batch 800, loss[loss=0.2077, simple_loss=0.2893, pruned_loss=0.06304, over 19651.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2916, pruned_loss=0.06694, over 3761797.91 frames. ], batch size: 55, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:36:43,873 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.634e+02 4.695e+02 6.255e+02 7.449e+02 1.390e+03, threshold=1.251e+03, percent-clipped=2.0 +2023-04-02 13:36:47,183 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 13:37:34,011 INFO [train.py:903] (3/4) Epoch 19, batch 850, loss[loss=0.1954, simple_loss=0.2821, pruned_loss=0.05434, over 17676.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.2925, pruned_loss=0.06727, over 3770592.29 frames. ], batch size: 101, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:38:17,263 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=123790.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:38:18,596 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123791.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:38:23,841 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 13:38:34,410 INFO [train.py:903] (3/4) Epoch 19, batch 900, loss[loss=0.2237, simple_loss=0.2945, pruned_loss=0.07643, over 17680.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2924, pruned_loss=0.06724, over 3781476.73 frames. ], batch size: 39, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:38:44,943 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.405e+02 4.993e+02 6.523e+02 8.112e+02 2.572e+03, threshold=1.305e+03, percent-clipped=9.0 +2023-04-02 13:39:35,390 INFO [train.py:903] (3/4) Epoch 19, batch 950, loss[loss=0.195, simple_loss=0.2713, pruned_loss=0.0593, over 19616.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2928, pruned_loss=0.06737, over 3791242.45 frames. ], batch size: 50, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:39:35,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 13:40:18,682 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123890.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:40:35,890 INFO [train.py:903] (3/4) Epoch 19, batch 1000, loss[loss=0.2228, simple_loss=0.3043, pruned_loss=0.07065, over 19673.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.293, pruned_loss=0.06792, over 3810730.71 frames. ], batch size: 60, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:40:37,306 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123905.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:40:44,459 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.613e+02 5.533e+02 6.903e+02 9.244e+02 2.435e+03, threshold=1.381e+03, percent-clipped=8.0 +2023-04-02 13:41:23,922 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 13:41:33,875 INFO [train.py:903] (3/4) Epoch 19, batch 1050, loss[loss=0.2045, simple_loss=0.2707, pruned_loss=0.06915, over 19776.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2923, pruned_loss=0.06758, over 3819177.89 frames. ], batch size: 47, lr: 4.39e-03, grad_scale: 8.0 +2023-04-02 13:41:37,444 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3173, 2.0193, 1.5976, 1.3478, 1.8654, 1.2596, 1.1938, 1.7766], + device='cuda:3'), covar=tensor([0.0908, 0.0818, 0.1126, 0.0830, 0.0508, 0.1304, 0.0735, 0.0447], + device='cuda:3'), in_proj_covar=tensor([0.0299, 0.0314, 0.0334, 0.0260, 0.0245, 0.0336, 0.0291, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:42:03,300 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 13:42:28,047 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0804, 1.7559, 2.0449, 1.6066, 4.6352, 1.1557, 2.5741, 4.9214], + device='cuda:3'), covar=tensor([0.0411, 0.2623, 0.2423, 0.2060, 0.0660, 0.2613, 0.1406, 0.0179], + device='cuda:3'), in_proj_covar=tensor([0.0400, 0.0357, 0.0377, 0.0342, 0.0368, 0.0350, 0.0369, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:42:36,021 INFO [train.py:903] (3/4) Epoch 19, batch 1100, loss[loss=0.2885, simple_loss=0.3523, pruned_loss=0.1123, over 19145.00 frames. ], tot_loss[loss=0.2133, simple_loss=0.2914, pruned_loss=0.06762, over 3814843.22 frames. ], batch size: 69, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:42:45,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.024e+02 4.737e+02 5.703e+02 7.685e+02 1.238e+03, threshold=1.141e+03, percent-clipped=0.0 +2023-04-02 13:43:28,388 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124047.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:43:36,792 INFO [train.py:903] (3/4) Epoch 19, batch 1150, loss[loss=0.3038, simple_loss=0.3533, pruned_loss=0.1271, over 13924.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2911, pruned_loss=0.06737, over 3814785.33 frames. ], batch size: 138, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:43:37,115 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124054.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:43:59,177 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124072.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:44:08,944 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124080.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:44:37,473 INFO [train.py:903] (3/4) Epoch 19, batch 1200, loss[loss=0.2325, simple_loss=0.3158, pruned_loss=0.07464, over 19075.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2912, pruned_loss=0.06722, over 3828506.84 frames. ], batch size: 69, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:44:48,187 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.573e+02 5.076e+02 6.121e+02 8.217e+02 1.455e+03, threshold=1.224e+03, percent-clipped=6.0 +2023-04-02 13:45:00,037 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-02 13:45:09,222 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 13:45:38,205 INFO [train.py:903] (3/4) Epoch 19, batch 1250, loss[loss=0.1727, simple_loss=0.2558, pruned_loss=0.04483, over 19606.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2914, pruned_loss=0.06742, over 3835640.38 frames. ], batch size: 50, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:45:47,124 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124161.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:46:16,831 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124186.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:46:38,208 INFO [train.py:903] (3/4) Epoch 19, batch 1300, loss[loss=0.2616, simple_loss=0.3422, pruned_loss=0.09049, over 18781.00 frames. ], tot_loss[loss=0.2141, simple_loss=0.2922, pruned_loss=0.06803, over 3830341.74 frames. ], batch size: 74, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:46:39,476 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3531, 3.9277, 2.5081, 3.4964, 0.7989, 3.8255, 3.7683, 3.8628], + device='cuda:3'), covar=tensor([0.0669, 0.1024, 0.2081, 0.0960, 0.4158, 0.0778, 0.0913, 0.1161], + device='cuda:3'), in_proj_covar=tensor([0.0484, 0.0394, 0.0479, 0.0338, 0.0394, 0.0419, 0.0407, 0.0442], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:46:48,366 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.504e+02 4.873e+02 5.866e+02 7.986e+02 1.872e+03, threshold=1.173e+03, percent-clipped=5.0 +2023-04-02 13:46:53,198 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.56 vs. limit=2.0 +2023-04-02 13:47:15,087 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124234.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:47:37,774 INFO [train.py:903] (3/4) Epoch 19, batch 1350, loss[loss=0.2306, simple_loss=0.3083, pruned_loss=0.07646, over 19139.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2928, pruned_loss=0.06829, over 3832052.87 frames. ], batch size: 69, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:48:22,197 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124290.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:48:30,542 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8464, 4.4202, 2.5627, 3.8850, 0.9310, 4.2066, 4.2483, 4.2888], + device='cuda:3'), covar=tensor([0.0527, 0.0906, 0.2074, 0.0804, 0.4072, 0.0747, 0.0831, 0.1091], + device='cuda:3'), in_proj_covar=tensor([0.0487, 0.0393, 0.0481, 0.0338, 0.0395, 0.0420, 0.0408, 0.0443], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 13:48:39,028 INFO [train.py:903] (3/4) Epoch 19, batch 1400, loss[loss=0.2188, simple_loss=0.3025, pruned_loss=0.06757, over 19616.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2927, pruned_loss=0.06812, over 3823272.75 frames. ], batch size: 50, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:48:48,825 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.237e+02 5.508e+02 6.829e+02 9.566e+02 2.163e+03, threshold=1.366e+03, percent-clipped=9.0 +2023-04-02 13:49:21,062 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-04-02 13:49:32,988 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124349.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:49:38,323 INFO [train.py:903] (3/4) Epoch 19, batch 1450, loss[loss=0.1863, simple_loss=0.2705, pruned_loss=0.05104, over 19770.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2912, pruned_loss=0.06739, over 3823948.37 frames. ], batch size: 54, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:49:40,272 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 13:49:53,733 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124366.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 13:50:04,342 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-04-02 13:50:32,827 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124398.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:50:40,519 INFO [train.py:903] (3/4) Epoch 19, batch 1500, loss[loss=0.2007, simple_loss=0.2922, pruned_loss=0.05461, over 19529.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2907, pruned_loss=0.06625, over 3838042.61 frames. ], batch size: 56, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:50:50,217 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.987e+02 4.822e+02 6.235e+02 8.378e+02 1.519e+03, threshold=1.247e+03, percent-clipped=2.0 +2023-04-02 13:51:04,439 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124424.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:51:28,513 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-02 13:51:35,999 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124450.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:51:39,966 INFO [train.py:903] (3/4) Epoch 19, batch 1550, loss[loss=0.2109, simple_loss=0.2839, pruned_loss=0.06893, over 19734.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.291, pruned_loss=0.06674, over 3844009.04 frames. ], batch size: 51, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:52:40,757 INFO [train.py:903] (3/4) Epoch 19, batch 1600, loss[loss=0.2212, simple_loss=0.2967, pruned_loss=0.07285, over 19653.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2907, pruned_loss=0.06662, over 3848049.62 frames. ], batch size: 53, lr: 4.38e-03, grad_scale: 8.0 +2023-04-02 13:52:51,837 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.803e+02 4.800e+02 6.281e+02 8.115e+02 1.566e+03, threshold=1.256e+03, percent-clipped=2.0 +2023-04-02 13:52:52,218 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124513.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:53:06,496 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 13:53:23,390 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124539.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:53:23,656 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 13:53:40,466 INFO [train.py:903] (3/4) Epoch 19, batch 1650, loss[loss=0.1982, simple_loss=0.2744, pruned_loss=0.06095, over 19848.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2914, pruned_loss=0.06713, over 3840208.22 frames. ], batch size: 52, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:54:43,095 INFO [train.py:903] (3/4) Epoch 19, batch 1700, loss[loss=0.1941, simple_loss=0.2866, pruned_loss=0.05078, over 19739.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.289, pruned_loss=0.06558, over 3849518.30 frames. ], batch size: 63, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:54:44,667 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124605.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:54:53,183 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.721e+02 4.942e+02 5.736e+02 7.226e+02 1.444e+03, threshold=1.147e+03, percent-clipped=3.0 +2023-04-02 13:55:14,615 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:55:18,820 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124634.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:55:21,980 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 13:55:42,975 INFO [train.py:903] (3/4) Epoch 19, batch 1750, loss[loss=0.2099, simple_loss=0.2794, pruned_loss=0.07024, over 19778.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2904, pruned_loss=0.06652, over 3845822.06 frames. ], batch size: 48, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:55:46,482 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124656.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:56:16,378 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124681.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:56:37,883 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-02 13:56:44,213 INFO [train.py:903] (3/4) Epoch 19, batch 1800, loss[loss=0.1965, simple_loss=0.2747, pruned_loss=0.05918, over 19621.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2903, pruned_loss=0.06645, over 3829433.34 frames. ], batch size: 50, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:56:51,811 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124710.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 13:56:54,991 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.267e+02 5.111e+02 6.286e+02 7.731e+02 1.656e+03, threshold=1.257e+03, percent-clipped=2.0 +2023-04-02 13:57:38,765 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 13:57:39,049 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124749.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:57:44,335 INFO [train.py:903] (3/4) Epoch 19, batch 1850, loss[loss=0.2, simple_loss=0.2775, pruned_loss=0.06124, over 19479.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2906, pruned_loss=0.06624, over 3821665.53 frames. ], batch size: 49, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:58:03,215 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124769.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:58:17,486 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 13:58:33,218 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=124794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:58:33,467 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:58:34,552 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124795.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:58:44,560 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0346, 2.1327, 2.3721, 2.6931, 2.0560, 2.6345, 2.4116, 2.2127], + device='cuda:3'), covar=tensor([0.4133, 0.3698, 0.1745, 0.2199, 0.3835, 0.1933, 0.4495, 0.3069], + device='cuda:3'), in_proj_covar=tensor([0.0870, 0.0928, 0.0696, 0.0914, 0.0852, 0.0787, 0.0824, 0.0762], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 13:58:46,482 INFO [train.py:903] (3/4) Epoch 19, batch 1900, loss[loss=0.2385, simple_loss=0.3077, pruned_loss=0.08467, over 19623.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2898, pruned_loss=0.06584, over 3816328.01 frames. ], batch size: 50, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 13:58:56,685 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.301e+02 4.862e+02 5.969e+02 7.822e+02 1.490e+03, threshold=1.194e+03, percent-clipped=1.0 +2023-04-02 13:59:01,922 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 13:59:05,770 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124820.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:59:07,770 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 13:59:11,242 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124825.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 13:59:26,795 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124838.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 13:59:31,995 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 13:59:46,593 INFO [train.py:903] (3/4) Epoch 19, batch 1950, loss[loss=0.1802, simple_loss=0.2589, pruned_loss=0.05081, over 19792.00 frames. ], tot_loss[loss=0.211, simple_loss=0.29, pruned_loss=0.06603, over 3818697.92 frames. ], batch size: 46, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:00:47,547 INFO [train.py:903] (3/4) Epoch 19, batch 2000, loss[loss=0.1638, simple_loss=0.2458, pruned_loss=0.04094, over 19391.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.2904, pruned_loss=0.06617, over 3807758.32 frames. ], batch size: 48, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:00:50,109 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1603, 2.0392, 1.8932, 1.7524, 1.5274, 1.6792, 0.6028, 1.0612], + device='cuda:3'), covar=tensor([0.0599, 0.0592, 0.0404, 0.0668, 0.1158, 0.0846, 0.1204, 0.0974], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0349, 0.0351, 0.0373, 0.0451, 0.0383, 0.0333, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:00:54,328 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124909.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:00:58,642 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.950e+02 4.950e+02 6.179e+02 7.428e+02 1.573e+03, threshold=1.236e+03, percent-clipped=5.0 +2023-04-02 14:01:45,076 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 14:01:48,486 INFO [train.py:903] (3/4) Epoch 19, batch 2050, loss[loss=0.234, simple_loss=0.3143, pruned_loss=0.07691, over 19623.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2895, pruned_loss=0.06566, over 3818582.61 frames. ], batch size: 57, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:02:06,429 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 14:02:07,323 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 14:02:25,175 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 14:02:44,759 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125000.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:02:50,764 INFO [train.py:903] (3/4) Epoch 19, batch 2100, loss[loss=0.2003, simple_loss=0.281, pruned_loss=0.05985, over 19540.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2893, pruned_loss=0.0659, over 3820974.87 frames. ], batch size: 56, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:02:52,331 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125005.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:03:00,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.657e+02 4.808e+02 5.766e+02 7.881e+02 2.968e+03, threshold=1.153e+03, percent-clipped=4.0 +2023-04-02 14:03:15,316 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125025.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:03:18,741 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 14:03:22,166 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125030.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:03:31,209 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4156, 1.5549, 1.7912, 1.7261, 2.7852, 2.2167, 2.9746, 1.3350], + device='cuda:3'), covar=tensor([0.2320, 0.4037, 0.2610, 0.1744, 0.1416, 0.2061, 0.1283, 0.4014], + device='cuda:3'), in_proj_covar=tensor([0.0515, 0.0623, 0.0684, 0.0467, 0.0614, 0.0518, 0.0653, 0.0532], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:03:39,536 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 14:03:49,807 INFO [train.py:903] (3/4) Epoch 19, batch 2150, loss[loss=0.167, simple_loss=0.2452, pruned_loss=0.04445, over 18653.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2898, pruned_loss=0.06621, over 3818606.92 frames. ], batch size: 41, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:04:22,807 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125081.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 14:04:29,459 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.03 vs. limit=5.0 +2023-04-02 14:04:50,319 INFO [train.py:903] (3/4) Epoch 19, batch 2200, loss[loss=0.2406, simple_loss=0.315, pruned_loss=0.08312, over 19367.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2906, pruned_loss=0.06655, over 3831407.48 frames. ], batch size: 66, lr: 4.37e-03, grad_scale: 8.0 +2023-04-02 14:04:53,049 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125106.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 14:05:01,343 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.808e+02 4.810e+02 5.592e+02 6.977e+02 1.826e+03, threshold=1.118e+03, percent-clipped=4.0 +2023-04-02 14:05:04,018 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125115.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:05:34,461 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125140.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:05:41,540 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125146.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:05:51,111 INFO [train.py:903] (3/4) Epoch 19, batch 2250, loss[loss=0.2478, simple_loss=0.3357, pruned_loss=0.07996, over 19332.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2913, pruned_loss=0.06671, over 3830555.97 frames. ], batch size: 66, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:06:05,775 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:06:25,691 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125182.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:06:35,060 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125190.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:06:52,197 INFO [train.py:903] (3/4) Epoch 19, batch 2300, loss[loss=0.2076, simple_loss=0.2936, pruned_loss=0.06078, over 19666.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2913, pruned_loss=0.06678, over 3824891.72 frames. ], batch size: 58, lr: 4.36e-03, grad_scale: 4.0 +2023-04-02 14:07:04,356 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.756e+02 5.218e+02 6.176e+02 8.185e+02 2.110e+03, threshold=1.235e+03, percent-clipped=6.0 +2023-04-02 14:07:06,715 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 14:07:52,839 INFO [train.py:903] (3/4) Epoch 19, batch 2350, loss[loss=0.1876, simple_loss=0.2721, pruned_loss=0.05148, over 19772.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2909, pruned_loss=0.06629, over 3833473.20 frames. ], batch size: 54, lr: 4.36e-03, grad_scale: 4.0 +2023-04-02 14:08:32,698 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 14:08:45,078 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125297.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:08:49,235 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 14:08:53,423 INFO [train.py:903] (3/4) Epoch 19, batch 2400, loss[loss=0.1894, simple_loss=0.2709, pruned_loss=0.0539, over 19582.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2909, pruned_loss=0.06655, over 3823348.58 frames. ], batch size: 52, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:09:05,367 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.876e+02 4.925e+02 6.150e+02 7.515e+02 1.529e+03, threshold=1.230e+03, percent-clipped=3.0 +2023-04-02 14:09:53,258 INFO [train.py:903] (3/4) Epoch 19, batch 2450, loss[loss=0.2205, simple_loss=0.2974, pruned_loss=0.07177, over 18711.00 frames. ], tot_loss[loss=0.2118, simple_loss=0.2905, pruned_loss=0.06652, over 3815726.25 frames. ], batch size: 74, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:10:14,740 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125371.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:10:17,719 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125374.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:10:41,885 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125394.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:10:44,292 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125396.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:10:44,326 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125396.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:10:53,808 INFO [train.py:903] (3/4) Epoch 19, batch 2500, loss[loss=0.2607, simple_loss=0.3317, pruned_loss=0.09488, over 18696.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2912, pruned_loss=0.06689, over 3830289.61 frames. ], batch size: 74, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:11:05,671 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.142e+02 4.898e+02 6.460e+02 8.264e+02 2.020e+03, threshold=1.292e+03, percent-clipped=4.0 +2023-04-02 14:11:13,988 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125421.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:11:35,784 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125439.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:11:39,886 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.1155, 3.8151, 2.9556, 3.2889, 1.6456, 3.6279, 3.5565, 3.6326], + device='cuda:3'), covar=tensor([0.0803, 0.1029, 0.1736, 0.0918, 0.2925, 0.0858, 0.1035, 0.1435], + device='cuda:3'), in_proj_covar=tensor([0.0484, 0.0393, 0.0478, 0.0335, 0.0391, 0.0417, 0.0408, 0.0443], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:11:54,214 INFO [train.py:903] (3/4) Epoch 19, batch 2550, loss[loss=0.1857, simple_loss=0.2706, pruned_loss=0.05044, over 19517.00 frames. ], tot_loss[loss=0.2138, simple_loss=0.2923, pruned_loss=0.06769, over 3801376.94 frames. ], batch size: 54, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:12:12,054 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0923, 5.0753, 5.8867, 5.9114, 1.8616, 5.5706, 4.7406, 5.5178], + device='cuda:3'), covar=tensor([0.1586, 0.0720, 0.0529, 0.0550, 0.6206, 0.0605, 0.0585, 0.1241], + device='cuda:3'), in_proj_covar=tensor([0.0760, 0.0702, 0.0910, 0.0796, 0.0806, 0.0659, 0.0547, 0.0839], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 14:12:30,646 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2973, 1.2087, 1.6773, 1.3134, 2.5545, 3.6747, 3.3923, 3.8038], + device='cuda:3'), covar=tensor([0.1611, 0.3902, 0.3416, 0.2380, 0.0660, 0.0175, 0.0204, 0.0260], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0313, 0.0344, 0.0261, 0.0238, 0.0180, 0.0212, 0.0244], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 14:12:38,053 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125490.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:12:47,117 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 14:12:53,798 INFO [train.py:903] (3/4) Epoch 19, batch 2600, loss[loss=0.2035, simple_loss=0.2905, pruned_loss=0.05821, over 19587.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2922, pruned_loss=0.0676, over 3801549.18 frames. ], batch size: 61, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:13:01,764 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7625, 1.8997, 2.1732, 2.3516, 1.7391, 2.2736, 2.2112, 2.0539], + device='cuda:3'), covar=tensor([0.4088, 0.3523, 0.1768, 0.2166, 0.3774, 0.1974, 0.4659, 0.3033], + device='cuda:3'), in_proj_covar=tensor([0.0871, 0.0929, 0.0699, 0.0916, 0.0853, 0.0789, 0.0827, 0.0764], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 14:13:05,878 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.683e+02 4.815e+02 5.841e+02 7.665e+02 1.339e+03, threshold=1.168e+03, percent-clipped=2.0 +2023-04-02 14:13:39,741 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0569, 1.4082, 1.7898, 1.3148, 2.7450, 3.7062, 3.3895, 3.8526], + device='cuda:3'), covar=tensor([0.1729, 0.3626, 0.3269, 0.2399, 0.0585, 0.0152, 0.0215, 0.0237], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0311, 0.0343, 0.0260, 0.0237, 0.0179, 0.0212, 0.0244], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 14:13:43,176 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7941, 1.6673, 1.6295, 2.2877, 1.7123, 2.1453, 2.0698, 1.8681], + device='cuda:3'), covar=tensor([0.0778, 0.0870, 0.0998, 0.0679, 0.0839, 0.0702, 0.0921, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0221, 0.0227, 0.0245, 0.0228, 0.0212, 0.0190, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 14:13:52,943 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125553.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:13:53,671 INFO [train.py:903] (3/4) Epoch 19, batch 2650, loss[loss=0.219, simple_loss=0.2897, pruned_loss=0.07415, over 19492.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2917, pruned_loss=0.06719, over 3815798.35 frames. ], batch size: 49, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:14:15,477 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 14:14:23,594 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125578.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:14:34,897 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 14:14:54,498 INFO [train.py:903] (3/4) Epoch 19, batch 2700, loss[loss=0.2031, simple_loss=0.28, pruned_loss=0.06314, over 19776.00 frames. ], tot_loss[loss=0.214, simple_loss=0.2924, pruned_loss=0.06777, over 3821459.62 frames. ], batch size: 54, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:14:56,033 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125605.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:14:59,365 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.59 vs. limit=5.0 +2023-04-02 14:15:07,182 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.594e+02 4.817e+02 5.964e+02 7.468e+02 1.608e+03, threshold=1.193e+03, percent-clipped=5.0 +2023-04-02 14:15:15,710 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125621.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:15:31,473 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8635, 1.4845, 1.4602, 1.8017, 1.5166, 1.5861, 1.4405, 1.7320], + device='cuda:3'), covar=tensor([0.1046, 0.1360, 0.1515, 0.0982, 0.1228, 0.0586, 0.1361, 0.0766], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0356, 0.0309, 0.0249, 0.0300, 0.0249, 0.0301, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:15:56,194 INFO [train.py:903] (3/4) Epoch 19, batch 2750, loss[loss=0.2172, simple_loss=0.2962, pruned_loss=0.06914, over 19786.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2919, pruned_loss=0.0675, over 3821672.06 frames. ], batch size: 56, lr: 4.36e-03, grad_scale: 8.0 +2023-04-02 14:16:49,132 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1023, 1.7459, 1.3596, 1.1356, 1.5836, 1.0401, 1.1013, 1.5542], + device='cuda:3'), covar=tensor([0.0832, 0.0793, 0.1042, 0.0850, 0.0533, 0.1268, 0.0642, 0.0438], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0312, 0.0333, 0.0260, 0.0245, 0.0334, 0.0290, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:16:55,540 INFO [train.py:903] (3/4) Epoch 19, batch 2800, loss[loss=0.1742, simple_loss=0.2453, pruned_loss=0.05151, over 19118.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2932, pruned_loss=0.06797, over 3829229.98 frames. ], batch size: 42, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:17:01,166 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6024, 2.1816, 2.2986, 2.6814, 2.3149, 2.3668, 2.0460, 2.4719], + device='cuda:3'), covar=tensor([0.0882, 0.1699, 0.1235, 0.0988, 0.1373, 0.0451, 0.1246, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0356, 0.0309, 0.0249, 0.0300, 0.0248, 0.0300, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:17:08,446 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.000e+02 5.188e+02 6.338e+02 7.813e+02 1.733e+03, threshold=1.268e+03, percent-clipped=8.0 +2023-04-02 14:17:12,865 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125718.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:17:37,023 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125738.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:17:42,797 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125743.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:17:55,248 INFO [train.py:903] (3/4) Epoch 19, batch 2850, loss[loss=0.212, simple_loss=0.2979, pruned_loss=0.06307, over 19658.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2933, pruned_loss=0.0678, over 3829412.40 frames. ], batch size: 58, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:18:04,516 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0634, 1.7744, 2.1494, 1.7561, 4.5476, 1.2659, 2.6275, 4.9614], + device='cuda:3'), covar=tensor([0.0397, 0.2617, 0.2491, 0.1978, 0.0722, 0.2581, 0.1348, 0.0161], + device='cuda:3'), in_proj_covar=tensor([0.0397, 0.0358, 0.0377, 0.0342, 0.0367, 0.0350, 0.0368, 0.0388], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:18:15,696 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1815, 1.8221, 1.4233, 1.2295, 1.6292, 1.1312, 1.1989, 1.6848], + device='cuda:3'), covar=tensor([0.0778, 0.0807, 0.1116, 0.0771, 0.0541, 0.1286, 0.0608, 0.0389], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0312, 0.0334, 0.0260, 0.0245, 0.0335, 0.0291, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:18:30,688 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125783.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:18:56,094 INFO [train.py:903] (3/4) Epoch 19, batch 2900, loss[loss=0.187, simple_loss=0.2634, pruned_loss=0.05534, over 19753.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2924, pruned_loss=0.06742, over 3833443.56 frames. ], batch size: 45, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:18:56,110 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 14:19:09,042 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.234e+02 4.625e+02 5.511e+02 7.350e+02 1.619e+03, threshold=1.102e+03, percent-clipped=3.0 +2023-04-02 14:19:31,569 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125833.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:19:47,779 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3348, 1.9856, 1.5412, 1.2575, 1.7456, 1.1039, 1.2716, 1.8838], + device='cuda:3'), covar=tensor([0.0796, 0.0738, 0.1086, 0.0775, 0.0554, 0.1248, 0.0638, 0.0399], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0314, 0.0336, 0.0261, 0.0247, 0.0337, 0.0293, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:19:55,782 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125853.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:19:56,472 INFO [train.py:903] (3/4) Epoch 19, batch 2950, loss[loss=0.2411, simple_loss=0.3104, pruned_loss=0.08596, over 19699.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2931, pruned_loss=0.06793, over 3811362.42 frames. ], batch size: 59, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:20:05,757 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125861.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:20:36,551 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125886.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:20:50,745 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125898.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:20:57,381 INFO [train.py:903] (3/4) Epoch 19, batch 3000, loss[loss=0.234, simple_loss=0.3059, pruned_loss=0.08104, over 19671.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2933, pruned_loss=0.06803, over 3817944.59 frames. ], batch size: 60, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:20:57,381 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 14:21:08,244 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5383, 1.6024, 1.5493, 1.3464, 1.2493, 1.3153, 0.3410, 0.6325], + device='cuda:3'), covar=tensor([0.0638, 0.0695, 0.0455, 0.0729, 0.1145, 0.0988, 0.1357, 0.1257], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0352, 0.0355, 0.0378, 0.0457, 0.0388, 0.0334, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:21:10,734 INFO [train.py:937] (3/4) Epoch 19, validation: loss=0.1696, simple_loss=0.2702, pruned_loss=0.03451, over 944034.00 frames. +2023-04-02 14:21:10,736 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 14:21:10,824 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 14:21:24,051 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.456e+02 4.999e+02 6.816e+02 8.693e+02 1.814e+03, threshold=1.363e+03, percent-clipped=12.0 +2023-04-02 14:22:11,565 INFO [train.py:903] (3/4) Epoch 19, batch 3050, loss[loss=0.1847, simple_loss=0.2655, pruned_loss=0.05193, over 19613.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2935, pruned_loss=0.06774, over 3823454.84 frames. ], batch size: 50, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:22:24,982 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=125965.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:23:10,423 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5635, 1.1748, 1.5646, 1.4788, 3.1247, 1.1146, 2.3068, 3.5210], + device='cuda:3'), covar=tensor([0.0524, 0.3072, 0.2770, 0.1944, 0.0741, 0.2639, 0.1348, 0.0288], + device='cuda:3'), in_proj_covar=tensor([0.0396, 0.0359, 0.0375, 0.0342, 0.0366, 0.0350, 0.0369, 0.0388], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:23:13,569 INFO [train.py:903] (3/4) Epoch 19, batch 3100, loss[loss=0.1946, simple_loss=0.2676, pruned_loss=0.06084, over 18219.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2931, pruned_loss=0.06767, over 3812215.04 frames. ], batch size: 40, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:23:21,474 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1047, 3.5115, 2.0254, 1.9096, 3.0793, 1.6987, 1.3628, 2.2755], + device='cuda:3'), covar=tensor([0.1329, 0.0545, 0.1023, 0.0941, 0.0527, 0.1159, 0.1000, 0.0683], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0313, 0.0334, 0.0261, 0.0246, 0.0335, 0.0292, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:23:26,818 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 4.651e+02 5.591e+02 6.916e+02 1.279e+03, threshold=1.118e+03, percent-clipped=0.0 +2023-04-02 14:23:28,124 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126016.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:24:13,738 INFO [train.py:903] (3/4) Epoch 19, batch 3150, loss[loss=0.1875, simple_loss=0.2741, pruned_loss=0.05042, over 19768.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2934, pruned_loss=0.06754, over 3821027.82 frames. ], batch size: 54, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:24:40,340 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 14:24:45,989 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126080.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:24:53,516 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=126087.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:24:55,931 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126089.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:25:04,721 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4189, 2.1667, 1.6278, 1.4886, 2.0038, 1.2336, 1.3826, 1.8753], + device='cuda:3'), covar=tensor([0.0981, 0.0735, 0.0963, 0.0744, 0.0505, 0.1163, 0.0621, 0.0437], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0311, 0.0332, 0.0259, 0.0243, 0.0333, 0.0289, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:25:14,245 INFO [train.py:903] (3/4) Epoch 19, batch 3200, loss[loss=0.218, simple_loss=0.3023, pruned_loss=0.0668, over 19662.00 frames. ], tot_loss[loss=0.2138, simple_loss=0.2931, pruned_loss=0.06727, over 3825939.61 frames. ], batch size: 58, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:25:21,281 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126109.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:25:26,837 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126114.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:25:27,595 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.283e+02 4.817e+02 6.226e+02 7.515e+02 1.545e+03, threshold=1.245e+03, percent-clipped=7.0 +2023-04-02 14:25:51,074 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126134.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:26:15,162 INFO [train.py:903] (3/4) Epoch 19, batch 3250, loss[loss=0.2176, simple_loss=0.2958, pruned_loss=0.06968, over 18207.00 frames. ], tot_loss[loss=0.2151, simple_loss=0.2937, pruned_loss=0.06824, over 3819069.78 frames. ], batch size: 83, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:26:15,582 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126154.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:26:46,075 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126179.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:26:58,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3509, 3.1117, 2.2389, 2.2979, 2.1544, 2.5258, 1.1506, 2.2419], + device='cuda:3'), covar=tensor([0.0625, 0.0559, 0.0766, 0.1158, 0.1075, 0.1252, 0.1375, 0.1028], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0349, 0.0352, 0.0377, 0.0454, 0.0386, 0.0333, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:27:14,784 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126202.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:27:16,666 INFO [train.py:903] (3/4) Epoch 19, batch 3300, loss[loss=0.2146, simple_loss=0.2856, pruned_loss=0.07177, over 19584.00 frames. ], tot_loss[loss=0.2132, simple_loss=0.2921, pruned_loss=0.06715, over 3826957.54 frames. ], batch size: 52, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:27:20,136 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 14:27:30,251 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.439e+02 5.162e+02 6.410e+02 7.971e+02 2.422e+03, threshold=1.282e+03, percent-clipped=4.0 +2023-04-02 14:28:17,454 INFO [train.py:903] (3/4) Epoch 19, batch 3350, loss[loss=0.2591, simple_loss=0.3217, pruned_loss=0.09821, over 18113.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2915, pruned_loss=0.06723, over 3812679.10 frames. ], batch size: 83, lr: 4.35e-03, grad_scale: 8.0 +2023-04-02 14:28:38,574 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2336, 1.2915, 1.2369, 1.0373, 1.0788, 1.0578, 0.0860, 0.3542], + device='cuda:3'), covar=tensor([0.0651, 0.0628, 0.0426, 0.0539, 0.1239, 0.0598, 0.1207, 0.1050], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0349, 0.0352, 0.0376, 0.0452, 0.0384, 0.0332, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:29:17,954 INFO [train.py:903] (3/4) Epoch 19, batch 3400, loss[loss=0.2223, simple_loss=0.3029, pruned_loss=0.0708, over 18372.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2913, pruned_loss=0.06693, over 3816937.21 frames. ], batch size: 84, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:29:31,302 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.222e+02 5.180e+02 6.014e+02 8.021e+02 1.733e+03, threshold=1.203e+03, percent-clipped=4.0 +2023-04-02 14:29:57,171 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126336.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:30:18,211 INFO [train.py:903] (3/4) Epoch 19, batch 3450, loss[loss=0.2186, simple_loss=0.2984, pruned_loss=0.06938, over 18866.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2904, pruned_loss=0.06632, over 3824454.23 frames. ], batch size: 74, lr: 4.34e-03, grad_scale: 4.0 +2023-04-02 14:30:22,545 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 14:30:27,175 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=126360.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:30:28,563 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126361.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:30:41,477 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 14:31:13,313 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126398.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:31:20,717 INFO [train.py:903] (3/4) Epoch 19, batch 3500, loss[loss=0.2813, simple_loss=0.3393, pruned_loss=0.1116, over 13847.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2907, pruned_loss=0.06684, over 3806525.83 frames. ], batch size: 136, lr: 4.34e-03, grad_scale: 4.0 +2023-04-02 14:31:23,782 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.48 vs. limit=2.0 +2023-04-02 14:31:34,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.238e+02 4.908e+02 6.053e+02 7.325e+02 1.346e+03, threshold=1.211e+03, percent-clipped=1.0 +2023-04-02 14:31:53,999 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9868, 3.6258, 2.4006, 3.2616, 0.7460, 3.4984, 3.4727, 3.5044], + device='cuda:3'), covar=tensor([0.0854, 0.1112, 0.2139, 0.0919, 0.4331, 0.0884, 0.1006, 0.1398], + device='cuda:3'), in_proj_covar=tensor([0.0488, 0.0394, 0.0480, 0.0337, 0.0396, 0.0419, 0.0409, 0.0446], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:32:21,707 INFO [train.py:903] (3/4) Epoch 19, batch 3550, loss[loss=0.2175, simple_loss=0.2946, pruned_loss=0.07017, over 19530.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2914, pruned_loss=0.06727, over 3808299.69 frames. ], batch size: 56, lr: 4.34e-03, grad_scale: 4.0 +2023-04-02 14:32:26,708 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126458.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:32:38,980 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126468.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:32:48,142 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126475.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:32:57,930 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126483.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:33:00,140 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0009, 1.9333, 1.7723, 1.5971, 1.4340, 1.5688, 0.4184, 0.9012], + device='cuda:3'), covar=tensor([0.0518, 0.0552, 0.0400, 0.0637, 0.1084, 0.0702, 0.1113, 0.0938], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0349, 0.0353, 0.0376, 0.0454, 0.0386, 0.0332, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:33:21,950 INFO [train.py:903] (3/4) Epoch 19, batch 3600, loss[loss=0.1974, simple_loss=0.2775, pruned_loss=0.05861, over 19853.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2908, pruned_loss=0.06683, over 3818416.05 frames. ], batch size: 52, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:33:37,204 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.060e+02 4.926e+02 5.826e+02 7.456e+02 2.258e+03, threshold=1.165e+03, percent-clipped=2.0 +2023-04-02 14:34:22,223 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8372, 4.4127, 2.7580, 3.8596, 1.0843, 4.2544, 4.2237, 4.2820], + device='cuda:3'), covar=tensor([0.0565, 0.0989, 0.2018, 0.0874, 0.3911, 0.0660, 0.0819, 0.1034], + device='cuda:3'), in_proj_covar=tensor([0.0488, 0.0395, 0.0483, 0.0338, 0.0395, 0.0419, 0.0409, 0.0447], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:34:23,084 INFO [train.py:903] (3/4) Epoch 19, batch 3650, loss[loss=0.2015, simple_loss=0.2747, pruned_loss=0.06414, over 19393.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.291, pruned_loss=0.06698, over 3822009.25 frames. ], batch size: 48, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:35:24,552 INFO [train.py:903] (3/4) Epoch 19, batch 3700, loss[loss=0.26, simple_loss=0.3296, pruned_loss=0.09524, over 19672.00 frames. ], tot_loss[loss=0.2133, simple_loss=0.2914, pruned_loss=0.06757, over 3824619.56 frames. ], batch size: 60, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:35:38,478 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.032e+02 5.326e+02 6.409e+02 8.349e+02 1.648e+03, threshold=1.282e+03, percent-clipped=7.0 +2023-04-02 14:36:06,933 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-02 14:36:23,985 INFO [train.py:903] (3/4) Epoch 19, batch 3750, loss[loss=0.2207, simple_loss=0.2981, pruned_loss=0.07169, over 19751.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2911, pruned_loss=0.06711, over 3821802.72 frames. ], batch size: 51, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:37:10,302 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9167, 1.3646, 1.0964, 1.0241, 1.2104, 1.0205, 1.0145, 1.2907], + device='cuda:3'), covar=tensor([0.0574, 0.0859, 0.1073, 0.0739, 0.0568, 0.1290, 0.0554, 0.0497], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0312, 0.0331, 0.0260, 0.0243, 0.0333, 0.0289, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:37:25,133 INFO [train.py:903] (3/4) Epoch 19, batch 3800, loss[loss=0.209, simple_loss=0.2905, pruned_loss=0.06374, over 19533.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2919, pruned_loss=0.06764, over 3801908.94 frames. ], batch size: 56, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:37:40,985 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.076e+02 4.757e+02 5.693e+02 7.302e+02 1.543e+03, threshold=1.139e+03, percent-clipped=1.0 +2023-04-02 14:37:57,333 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 14:37:58,875 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126731.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:38:11,174 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1778, 3.6376, 2.0604, 2.0767, 3.1741, 1.8164, 1.5076, 2.1579], + device='cuda:3'), covar=tensor([0.1357, 0.0471, 0.1143, 0.0917, 0.0524, 0.1291, 0.1033, 0.0767], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0312, 0.0332, 0.0261, 0.0244, 0.0334, 0.0290, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:38:12,100 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=126742.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:38:26,704 INFO [train.py:903] (3/4) Epoch 19, batch 3850, loss[loss=0.1632, simple_loss=0.238, pruned_loss=0.04417, over 18684.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2914, pruned_loss=0.06741, over 3808675.93 frames. ], batch size: 41, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:38:30,286 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126756.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:39:28,455 INFO [train.py:903] (3/4) Epoch 19, batch 3900, loss[loss=0.2235, simple_loss=0.3008, pruned_loss=0.07314, over 17546.00 frames. ], tot_loss[loss=0.2144, simple_loss=0.2925, pruned_loss=0.06811, over 3809949.67 frames. ], batch size: 101, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:39:30,208 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.27 vs. limit=5.0 +2023-04-02 14:39:37,802 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=126812.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:39:42,918 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.374e+02 5.082e+02 6.454e+02 7.734e+02 3.345e+03, threshold=1.291e+03, percent-clipped=6.0 +2023-04-02 14:40:04,533 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126834.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:40:29,140 INFO [train.py:903] (3/4) Epoch 19, batch 3950, loss[loss=0.1899, simple_loss=0.2802, pruned_loss=0.04976, over 19785.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2913, pruned_loss=0.06729, over 3815367.14 frames. ], batch size: 56, lr: 4.34e-03, grad_scale: 8.0 +2023-04-02 14:40:33,528 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126857.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:40:35,261 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 14:41:29,513 INFO [train.py:903] (3/4) Epoch 19, batch 4000, loss[loss=0.253, simple_loss=0.3232, pruned_loss=0.09138, over 17993.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.2914, pruned_loss=0.06775, over 3802537.96 frames. ], batch size: 83, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:41:43,557 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.579e+02 4.980e+02 6.258e+02 9.023e+02 1.716e+03, threshold=1.252e+03, percent-clipped=4.0 +2023-04-02 14:41:57,631 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:42:16,841 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 14:42:30,357 INFO [train.py:903] (3/4) Epoch 19, batch 4050, loss[loss=0.1641, simple_loss=0.2411, pruned_loss=0.04356, over 19767.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2911, pruned_loss=0.06724, over 3812273.30 frames. ], batch size: 47, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:43:17,812 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8416, 3.2682, 3.3259, 3.3572, 1.3425, 3.2221, 2.8407, 3.0956], + device='cuda:3'), covar=tensor([0.1673, 0.0970, 0.0827, 0.0902, 0.5380, 0.1000, 0.0815, 0.1326], + device='cuda:3'), in_proj_covar=tensor([0.0766, 0.0713, 0.0916, 0.0804, 0.0817, 0.0671, 0.0554, 0.0854], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 14:43:30,607 INFO [train.py:903] (3/4) Epoch 19, batch 4100, loss[loss=0.192, simple_loss=0.2767, pruned_loss=0.05363, over 19524.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2911, pruned_loss=0.06733, over 3814401.57 frames. ], batch size: 54, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:43:45,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.326e+02 4.699e+02 5.681e+02 7.096e+02 1.300e+03, threshold=1.136e+03, percent-clipped=1.0 +2023-04-02 14:44:06,969 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 14:44:31,640 INFO [train.py:903] (3/4) Epoch 19, batch 4150, loss[loss=0.2095, simple_loss=0.2952, pruned_loss=0.06194, over 19353.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2907, pruned_loss=0.06707, over 3805469.70 frames. ], batch size: 66, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:45:32,545 INFO [train.py:903] (3/4) Epoch 19, batch 4200, loss[loss=0.2201, simple_loss=0.3098, pruned_loss=0.06521, over 19637.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2907, pruned_loss=0.06689, over 3810116.76 frames. ], batch size: 58, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:45:35,885 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 14:45:43,861 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127113.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:45:46,907 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.729e+02 4.915e+02 5.762e+02 6.825e+02 1.362e+03, threshold=1.152e+03, percent-clipped=8.0 +2023-04-02 14:46:14,876 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127138.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:46:28,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2231, 1.3028, 1.2267, 1.0440, 1.0720, 1.0916, 0.1167, 0.3901], + device='cuda:3'), covar=tensor([0.0634, 0.0609, 0.0396, 0.0527, 0.1244, 0.0607, 0.1157, 0.0996], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0348, 0.0353, 0.0375, 0.0454, 0.0384, 0.0331, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 14:46:32,718 INFO [train.py:903] (3/4) Epoch 19, batch 4250, loss[loss=0.2231, simple_loss=0.2988, pruned_loss=0.07368, over 19767.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.29, pruned_loss=0.06667, over 3811707.17 frames. ], batch size: 54, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:46:46,857 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5671, 1.0915, 1.3225, 1.3155, 2.2260, 1.0679, 1.9940, 2.4792], + device='cuda:3'), covar=tensor([0.0653, 0.2836, 0.2898, 0.1585, 0.0808, 0.2043, 0.1082, 0.0470], + device='cuda:3'), in_proj_covar=tensor([0.0394, 0.0357, 0.0375, 0.0341, 0.0366, 0.0347, 0.0367, 0.0386], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:46:50,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 14:47:02,375 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 14:47:03,656 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=127178.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:47:09,530 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127183.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:47:34,690 INFO [train.py:903] (3/4) Epoch 19, batch 4300, loss[loss=0.2064, simple_loss=0.2848, pruned_loss=0.06405, over 19764.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2904, pruned_loss=0.06708, over 3810443.34 frames. ], batch size: 54, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:47:40,432 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127208.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:47:50,093 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.800e+02 4.893e+02 5.914e+02 7.996e+02 1.682e+03, threshold=1.183e+03, percent-clipped=7.0 +2023-04-02 14:48:28,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 14:48:35,535 INFO [train.py:903] (3/4) Epoch 19, batch 4350, loss[loss=0.1927, simple_loss=0.2853, pruned_loss=0.05007, over 19678.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2907, pruned_loss=0.06695, over 3818908.44 frames. ], batch size: 53, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:48:51,136 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127267.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:49:02,191 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 14:49:23,150 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127293.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:49:35,819 INFO [train.py:903] (3/4) Epoch 19, batch 4400, loss[loss=0.1834, simple_loss=0.254, pruned_loss=0.05639, over 19296.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2901, pruned_loss=0.06658, over 3822886.20 frames. ], batch size: 44, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:49:49,568 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.177e+02 5.239e+02 6.175e+02 6.853e+02 1.670e+03, threshold=1.235e+03, percent-clipped=2.0 +2023-04-02 14:50:02,112 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 14:50:12,010 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 14:50:12,999 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-02 14:50:36,323 INFO [train.py:903] (3/4) Epoch 19, batch 4450, loss[loss=0.2688, simple_loss=0.328, pruned_loss=0.1048, over 13420.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2902, pruned_loss=0.06642, over 3811280.83 frames. ], batch size: 136, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:51:38,004 INFO [train.py:903] (3/4) Epoch 19, batch 4500, loss[loss=0.2382, simple_loss=0.3227, pruned_loss=0.07682, over 19655.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2903, pruned_loss=0.06633, over 3815156.68 frames. ], batch size: 60, lr: 4.33e-03, grad_scale: 8.0 +2023-04-02 14:51:52,898 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.369e+02 5.116e+02 6.133e+02 7.767e+02 1.446e+03, threshold=1.227e+03, percent-clipped=3.0 +2023-04-02 14:51:53,551 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-02 14:52:28,102 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127445.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 14:52:39,295 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.93 vs. limit=5.0 +2023-04-02 14:52:39,717 INFO [train.py:903] (3/4) Epoch 19, batch 4550, loss[loss=0.2353, simple_loss=0.314, pruned_loss=0.07829, over 18192.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.291, pruned_loss=0.0664, over 3825374.39 frames. ], batch size: 83, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:52:48,339 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 14:53:11,963 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 14:53:29,921 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4926, 1.2916, 1.2660, 1.5473, 1.2875, 1.3211, 1.2801, 1.4068], + device='cuda:3'), covar=tensor([0.0834, 0.1165, 0.1113, 0.0668, 0.0978, 0.0467, 0.1071, 0.0604], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0350, 0.0302, 0.0246, 0.0294, 0.0243, 0.0295, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:53:40,565 INFO [train.py:903] (3/4) Epoch 19, batch 4600, loss[loss=0.2078, simple_loss=0.2966, pruned_loss=0.05949, over 18894.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2912, pruned_loss=0.06659, over 3818032.00 frames. ], batch size: 74, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:53:52,444 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5779, 2.3387, 2.1707, 2.6444, 2.2611, 2.1234, 1.9232, 2.3369], + device='cuda:3'), covar=tensor([0.0891, 0.1571, 0.1364, 0.1087, 0.1402, 0.0516, 0.1362, 0.0677], + device='cuda:3'), in_proj_covar=tensor([0.0262, 0.0351, 0.0303, 0.0247, 0.0295, 0.0244, 0.0296, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:53:54,253 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.388e+02 5.018e+02 6.286e+02 8.427e+02 2.189e+03, threshold=1.257e+03, percent-clipped=8.0 +2023-04-02 14:54:11,483 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3901, 1.4119, 1.9890, 1.6274, 3.0199, 4.5802, 4.5289, 5.0401], + device='cuda:3'), covar=tensor([0.1561, 0.3747, 0.3147, 0.2285, 0.0606, 0.0202, 0.0160, 0.0186], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0316, 0.0347, 0.0261, 0.0239, 0.0182, 0.0214, 0.0246], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 14:54:16,999 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7880, 4.2752, 4.4775, 4.4916, 1.6295, 4.2230, 3.6620, 4.1931], + device='cuda:3'), covar=tensor([0.1393, 0.0828, 0.0539, 0.0577, 0.5655, 0.0806, 0.0625, 0.1039], + device='cuda:3'), in_proj_covar=tensor([0.0758, 0.0711, 0.0910, 0.0795, 0.0812, 0.0666, 0.0551, 0.0848], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 14:54:32,459 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127547.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:54:34,766 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127549.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:54:37,339 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-04-02 14:54:39,919 INFO [train.py:903] (3/4) Epoch 19, batch 4650, loss[loss=0.1972, simple_loss=0.2849, pruned_loss=0.05476, over 19670.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.29, pruned_loss=0.06585, over 3818310.69 frames. ], batch size: 60, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:54:50,256 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3043, 3.8214, 3.9460, 3.9543, 1.6134, 3.7415, 3.2824, 3.7172], + device='cuda:3'), covar=tensor([0.1737, 0.0808, 0.0708, 0.0764, 0.5584, 0.0843, 0.0741, 0.1192], + device='cuda:3'), in_proj_covar=tensor([0.0761, 0.0715, 0.0914, 0.0798, 0.0814, 0.0668, 0.0553, 0.0850], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 14:54:55,882 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 14:55:05,459 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127574.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:55:07,391 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 14:55:16,480 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4871, 1.3213, 1.3322, 1.6972, 1.3723, 1.6240, 1.6936, 1.5351], + device='cuda:3'), covar=tensor([0.0926, 0.1057, 0.1130, 0.0823, 0.0912, 0.0839, 0.0888, 0.0745], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0227, 0.0246, 0.0229, 0.0212, 0.0190, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 14:55:40,716 INFO [train.py:903] (3/4) Epoch 19, batch 4700, loss[loss=0.2218, simple_loss=0.3015, pruned_loss=0.07101, over 19273.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2892, pruned_loss=0.0657, over 3823619.38 frames. ], batch size: 66, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:55:50,515 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=127611.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:55:55,864 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.385e+02 5.036e+02 6.203e+02 8.078e+02 1.735e+03, threshold=1.241e+03, percent-clipped=3.0 +2023-04-02 14:56:02,608 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 14:56:41,689 INFO [train.py:903] (3/4) Epoch 19, batch 4750, loss[loss=0.2261, simple_loss=0.3107, pruned_loss=0.07074, over 19371.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2885, pruned_loss=0.06554, over 3817230.45 frames. ], batch size: 70, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:57:13,621 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127680.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:57:41,813 INFO [train.py:903] (3/4) Epoch 19, batch 4800, loss[loss=0.2464, simple_loss=0.3162, pruned_loss=0.08834, over 19663.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2892, pruned_loss=0.06582, over 3813611.09 frames. ], batch size: 55, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:57:44,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6444, 1.3852, 1.5192, 1.5919, 3.2310, 1.1731, 2.2781, 3.6033], + device='cuda:3'), covar=tensor([0.0447, 0.2751, 0.2820, 0.1862, 0.0658, 0.2515, 0.1353, 0.0252], + device='cuda:3'), in_proj_covar=tensor([0.0394, 0.0358, 0.0377, 0.0341, 0.0367, 0.0347, 0.0369, 0.0386], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 14:57:55,392 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.353e+02 4.990e+02 6.329e+02 8.030e+02 1.437e+03, threshold=1.266e+03, percent-clipped=1.0 +2023-04-02 14:58:07,639 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127726.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:58:40,153 INFO [train.py:903] (3/4) Epoch 19, batch 4850, loss[loss=0.2089, simple_loss=0.3012, pruned_loss=0.05826, over 19603.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2913, pruned_loss=0.06712, over 3815081.50 frames. ], batch size: 57, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:59:04,712 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 14:59:22,982 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=127789.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 14:59:25,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 14:59:30,825 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 14:59:30,849 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 14:59:32,243 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127797.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 14:59:40,801 INFO [train.py:903] (3/4) Epoch 19, batch 4900, loss[loss=0.2126, simple_loss=0.2951, pruned_loss=0.06502, over 19554.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2911, pruned_loss=0.06686, over 3815653.92 frames. ], batch size: 56, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 14:59:40,836 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 14:59:55,908 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.219e+02 4.848e+02 5.865e+02 7.992e+02 2.664e+03, threshold=1.173e+03, percent-clipped=3.0 +2023-04-02 15:00:01,789 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 15:00:41,591 INFO [train.py:903] (3/4) Epoch 19, batch 4950, loss[loss=0.3139, simple_loss=0.368, pruned_loss=0.1298, over 12811.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2913, pruned_loss=0.06688, over 3827723.53 frames. ], batch size: 136, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 15:00:58,838 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127868.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:00:59,643 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 15:01:22,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 15:01:24,505 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127890.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:01:26,339 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=127891.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:01:41,515 INFO [train.py:903] (3/4) Epoch 19, batch 5000, loss[loss=0.1783, simple_loss=0.2643, pruned_loss=0.04614, over 19605.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2911, pruned_loss=0.0667, over 3830466.90 frames. ], batch size: 52, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 15:01:42,720 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127904.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:01:51,187 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 15:01:55,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.309e+02 4.675e+02 5.614e+02 6.818e+02 2.294e+03, threshold=1.123e+03, percent-clipped=3.0 +2023-04-02 15:02:03,226 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 15:02:23,457 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8804, 1.3068, 1.6356, 0.5814, 1.9835, 2.4293, 2.1228, 2.5915], + device='cuda:3'), covar=tensor([0.1681, 0.3575, 0.3138, 0.2680, 0.0597, 0.0257, 0.0339, 0.0351], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0314, 0.0345, 0.0260, 0.0237, 0.0180, 0.0213, 0.0245], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 15:02:41,854 INFO [train.py:903] (3/4) Epoch 19, batch 5050, loss[loss=0.2087, simple_loss=0.281, pruned_loss=0.06826, over 19417.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2912, pruned_loss=0.06668, over 3833453.58 frames. ], batch size: 48, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 15:02:55,372 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8424, 1.5190, 1.6579, 1.7713, 4.4001, 1.1636, 2.4605, 4.7742], + device='cuda:3'), covar=tensor([0.0445, 0.2857, 0.2971, 0.1876, 0.0760, 0.2727, 0.1539, 0.0163], + device='cuda:3'), in_proj_covar=tensor([0.0398, 0.0358, 0.0378, 0.0344, 0.0368, 0.0349, 0.0372, 0.0390], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:03:16,179 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127982.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:03:18,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 15:03:43,134 INFO [train.py:903] (3/4) Epoch 19, batch 5100, loss[loss=0.1828, simple_loss=0.2755, pruned_loss=0.04502, over 19297.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2913, pruned_loss=0.06672, over 3829405.99 frames. ], batch size: 66, lr: 4.32e-03, grad_scale: 8.0 +2023-04-02 15:03:45,734 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128006.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:03:47,029 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128007.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:03:56,497 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 15:03:58,282 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.139e+02 4.818e+02 5.706e+02 8.227e+02 1.561e+03, threshold=1.141e+03, percent-clipped=7.0 +2023-04-02 15:04:00,667 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 15:04:04,156 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 15:04:07,798 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128024.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:04:25,500 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8195, 1.3061, 1.4649, 1.4921, 3.2748, 1.0811, 2.3710, 3.7877], + device='cuda:3'), covar=tensor([0.0641, 0.3006, 0.2945, 0.2115, 0.0909, 0.2748, 0.1467, 0.0305], + device='cuda:3'), in_proj_covar=tensor([0.0398, 0.0360, 0.0379, 0.0345, 0.0370, 0.0350, 0.0373, 0.0391], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:04:43,095 INFO [train.py:903] (3/4) Epoch 19, batch 5150, loss[loss=0.2367, simple_loss=0.3057, pruned_loss=0.08384, over 19587.00 frames. ], tot_loss[loss=0.2118, simple_loss=0.2907, pruned_loss=0.06645, over 3829661.78 frames. ], batch size: 52, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:04:58,122 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 15:05:31,570 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 15:05:44,863 INFO [train.py:903] (3/4) Epoch 19, batch 5200, loss[loss=0.1575, simple_loss=0.2339, pruned_loss=0.04058, over 19332.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2906, pruned_loss=0.06627, over 3830207.20 frames. ], batch size: 44, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:05:59,005 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.969e+02 4.920e+02 6.208e+02 7.921e+02 1.726e+03, threshold=1.242e+03, percent-clipped=7.0 +2023-04-02 15:05:59,055 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 15:06:18,700 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1550, 5.0683, 5.9382, 5.9453, 1.7927, 5.6209, 4.7047, 5.5485], + device='cuda:3'), covar=tensor([0.1603, 0.0842, 0.0554, 0.0618, 0.6311, 0.0630, 0.0611, 0.1280], + device='cuda:3'), in_proj_covar=tensor([0.0753, 0.0710, 0.0910, 0.0798, 0.0806, 0.0666, 0.0548, 0.0842], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 15:06:28,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128139.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:06:30,660 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128141.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:06:41,119 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 15:06:45,759 INFO [train.py:903] (3/4) Epoch 19, batch 5250, loss[loss=0.2004, simple_loss=0.28, pruned_loss=0.06037, over 19275.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2903, pruned_loss=0.06634, over 3826587.29 frames. ], batch size: 66, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:06:50,379 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3752, 2.4153, 2.0413, 2.6779, 2.3764, 2.2684, 2.0038, 2.3980], + device='cuda:3'), covar=tensor([0.0999, 0.1489, 0.1431, 0.1007, 0.1295, 0.0479, 0.1306, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0355, 0.0307, 0.0250, 0.0297, 0.0247, 0.0298, 0.0250], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:06:53,717 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128160.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:07:23,111 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128185.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:07:45,449 INFO [train.py:903] (3/4) Epoch 19, batch 5300, loss[loss=0.2691, simple_loss=0.3324, pruned_loss=0.1028, over 18262.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2915, pruned_loss=0.06698, over 3817048.84 frames. ], batch size: 83, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:07:54,712 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128212.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:07:59,667 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.291e+02 5.001e+02 6.088e+02 7.600e+02 1.403e+03, threshold=1.218e+03, percent-clipped=2.0 +2023-04-02 15:08:00,872 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 15:08:21,813 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128234.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:08:46,733 INFO [train.py:903] (3/4) Epoch 19, batch 5350, loss[loss=0.1771, simple_loss=0.255, pruned_loss=0.04961, over 19775.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2913, pruned_loss=0.06682, over 3803866.32 frames. ], batch size: 47, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:08:50,322 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128256.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:08:57,122 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128262.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:09:04,899 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-02 15:09:20,021 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 15:09:26,926 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128287.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:09:47,391 INFO [train.py:903] (3/4) Epoch 19, batch 5400, loss[loss=0.2173, simple_loss=0.298, pruned_loss=0.06829, over 19657.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2911, pruned_loss=0.06646, over 3805111.97 frames. ], batch size: 55, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:10:01,763 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.288e+02 4.631e+02 5.571e+02 7.152e+02 1.493e+03, threshold=1.114e+03, percent-clipped=2.0 +2023-04-02 15:10:15,382 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128327.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:10:29,855 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128338.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:10:43,533 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128349.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:10:48,549 INFO [train.py:903] (3/4) Epoch 19, batch 5450, loss[loss=0.2457, simple_loss=0.3254, pruned_loss=0.08302, over 19675.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2908, pruned_loss=0.06627, over 3819976.15 frames. ], batch size: 59, lr: 4.31e-03, grad_scale: 16.0 +2023-04-02 15:11:39,585 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128395.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:11:50,176 INFO [train.py:903] (3/4) Epoch 19, batch 5500, loss[loss=0.2133, simple_loss=0.2908, pruned_loss=0.06792, over 19599.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2903, pruned_loss=0.06606, over 3808245.18 frames. ], batch size: 57, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:12:06,851 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.088e+02 5.166e+02 6.121e+02 7.872e+02 1.632e+03, threshold=1.224e+03, percent-clipped=5.0 +2023-04-02 15:12:10,665 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128420.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:12:12,701 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 15:12:23,815 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128431.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:12:50,272 INFO [train.py:903] (3/4) Epoch 19, batch 5550, loss[loss=0.2438, simple_loss=0.328, pruned_loss=0.07982, over 19668.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2901, pruned_loss=0.06619, over 3812451.83 frames. ], batch size: 60, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:12:56,488 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 15:13:25,601 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3220, 2.3629, 2.5898, 3.1152, 2.3251, 3.0654, 2.7576, 2.4045], + device='cuda:3'), covar=tensor([0.4200, 0.3985, 0.1739, 0.2360, 0.4184, 0.1942, 0.4170, 0.3043], + device='cuda:3'), in_proj_covar=tensor([0.0879, 0.0936, 0.0701, 0.0925, 0.0861, 0.0797, 0.0834, 0.0767], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 15:13:33,389 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.05 vs. limit=5.0 +2023-04-02 15:13:44,847 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 15:13:51,306 INFO [train.py:903] (3/4) Epoch 19, batch 5600, loss[loss=0.2248, simple_loss=0.2994, pruned_loss=0.07514, over 17545.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.29, pruned_loss=0.06609, over 3824055.59 frames. ], batch size: 101, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:14:01,896 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128512.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:14:07,029 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.064e+02 4.843e+02 5.877e+02 7.578e+02 1.194e+03, threshold=1.175e+03, percent-clipped=0.0 +2023-04-02 15:14:32,677 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128537.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:14:52,044 INFO [train.py:903] (3/4) Epoch 19, batch 5650, loss[loss=0.2935, simple_loss=0.3456, pruned_loss=0.1207, over 19787.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.29, pruned_loss=0.06627, over 3818980.84 frames. ], batch size: 54, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:15:27,799 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:15:37,259 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 15:15:53,141 INFO [train.py:903] (3/4) Epoch 19, batch 5700, loss[loss=0.2154, simple_loss=0.3029, pruned_loss=0.06397, over 19786.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2923, pruned_loss=0.06744, over 3807664.62 frames. ], batch size: 56, lr: 4.31e-03, grad_scale: 8.0 +2023-04-02 15:15:54,793 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128605.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:15:58,274 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128608.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:16:08,026 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.931e+02 5.156e+02 6.108e+02 7.232e+02 1.309e+03, threshold=1.222e+03, percent-clipped=4.0 +2023-04-02 15:16:25,772 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:16:53,490 INFO [train.py:903] (3/4) Epoch 19, batch 5750, loss[loss=0.2009, simple_loss=0.2822, pruned_loss=0.05984, over 19670.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2909, pruned_loss=0.06662, over 3815958.38 frames. ], batch size: 53, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:16:55,739 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 15:17:05,272 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 15:17:09,593 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 15:17:27,299 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128682.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:17:55,080 INFO [train.py:903] (3/4) Epoch 19, batch 5800, loss[loss=0.1723, simple_loss=0.2502, pruned_loss=0.04718, over 19411.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2905, pruned_loss=0.06646, over 3814943.39 frames. ], batch size: 48, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:18:10,461 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.713e+02 4.671e+02 6.414e+02 7.787e+02 1.302e+03, threshold=1.283e+03, percent-clipped=2.0 +2023-04-02 15:18:55,625 INFO [train.py:903] (3/4) Epoch 19, batch 5850, loss[loss=0.1946, simple_loss=0.2812, pruned_loss=0.05404, over 19660.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2903, pruned_loss=0.06636, over 3819696.49 frames. ], batch size: 55, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:19:20,645 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=128775.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:19:26,383 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6519, 4.2404, 2.6521, 3.7436, 1.2197, 4.1076, 4.0229, 4.0985], + device='cuda:3'), covar=tensor([0.0640, 0.0920, 0.2007, 0.0855, 0.3629, 0.0668, 0.0868, 0.1139], + device='cuda:3'), in_proj_covar=tensor([0.0494, 0.0398, 0.0487, 0.0342, 0.0399, 0.0422, 0.0413, 0.0450], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:19:48,478 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128797.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:19:50,618 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128799.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:19:55,918 INFO [train.py:903] (3/4) Epoch 19, batch 5900, loss[loss=0.201, simple_loss=0.2729, pruned_loss=0.06457, over 19720.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2895, pruned_loss=0.06588, over 3834301.94 frames. ], batch size: 47, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:20:02,596 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 15:20:11,628 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.258e+02 4.778e+02 5.849e+02 7.721e+02 1.320e+03, threshold=1.170e+03, percent-clipped=1.0 +2023-04-02 15:20:22,225 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 15:20:56,181 INFO [train.py:903] (3/4) Epoch 19, batch 5950, loss[loss=0.1875, simple_loss=0.2722, pruned_loss=0.05142, over 19698.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2899, pruned_loss=0.06596, over 3833384.73 frames. ], batch size: 59, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:21:13,042 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-02 15:21:41,370 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128890.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:21:57,518 INFO [train.py:903] (3/4) Epoch 19, batch 6000, loss[loss=0.1701, simple_loss=0.2489, pruned_loss=0.04567, over 19760.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2897, pruned_loss=0.06564, over 3824577.24 frames. ], batch size: 47, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:21:57,518 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 15:22:12,604 INFO [train.py:937] (3/4) Epoch 19, validation: loss=0.1702, simple_loss=0.2702, pruned_loss=0.03514, over 944034.00 frames. +2023-04-02 15:22:12,606 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 15:22:17,226 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128908.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:22:28,003 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.370e+02 5.136e+02 6.485e+02 9.043e+02 2.174e+03, threshold=1.297e+03, percent-clipped=7.0 +2023-04-02 15:23:11,990 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-02 15:23:13,563 INFO [train.py:903] (3/4) Epoch 19, batch 6050, loss[loss=0.221, simple_loss=0.3007, pruned_loss=0.07063, over 19601.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2891, pruned_loss=0.0655, over 3822493.56 frames. ], batch size: 61, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:24:15,361 INFO [train.py:903] (3/4) Epoch 19, batch 6100, loss[loss=0.1942, simple_loss=0.2735, pruned_loss=0.05745, over 19782.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.288, pruned_loss=0.06456, over 3840789.91 frames. ], batch size: 46, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:24:30,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.431e+02 4.994e+02 6.076e+02 7.380e+02 1.472e+03, threshold=1.215e+03, percent-clipped=4.0 +2023-04-02 15:25:15,141 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129053.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:25:15,828 INFO [train.py:903] (3/4) Epoch 19, batch 6150, loss[loss=0.1776, simple_loss=0.2553, pruned_loss=0.04999, over 19746.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2897, pruned_loss=0.0657, over 3835853.46 frames. ], batch size: 46, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:25:17,149 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9707, 2.0302, 2.1391, 1.9451, 3.6542, 1.6574, 2.9093, 3.6522], + device='cuda:3'), covar=tensor([0.0446, 0.2335, 0.2311, 0.1799, 0.0636, 0.2331, 0.1686, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0395, 0.0357, 0.0377, 0.0341, 0.0367, 0.0348, 0.0370, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:25:18,747 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.43 vs. limit=5.0 +2023-04-02 15:25:44,286 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 15:25:44,624 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129078.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:26:16,327 INFO [train.py:903] (3/4) Epoch 19, batch 6200, loss[loss=0.2242, simple_loss=0.3013, pruned_loss=0.07353, over 19617.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2917, pruned_loss=0.06663, over 3817269.79 frames. ], batch size: 57, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:26:32,071 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.839e+02 4.815e+02 6.250e+02 7.621e+02 1.523e+03, threshold=1.250e+03, percent-clipped=7.0 +2023-04-02 15:27:04,018 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=129143.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:27:07,735 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129146.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:27:17,161 INFO [train.py:903] (3/4) Epoch 19, batch 6250, loss[loss=0.2218, simple_loss=0.2985, pruned_loss=0.07257, over 12681.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2926, pruned_loss=0.06707, over 3810637.36 frames. ], batch size: 135, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:27:38,408 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129171.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:27:47,708 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 15:28:17,842 INFO [train.py:903] (3/4) Epoch 19, batch 6300, loss[loss=0.2019, simple_loss=0.2851, pruned_loss=0.05938, over 18768.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2922, pruned_loss=0.06689, over 3822937.30 frames. ], batch size: 74, lr: 4.30e-03, grad_scale: 8.0 +2023-04-02 15:28:33,726 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.168e+02 5.181e+02 6.373e+02 8.503e+02 1.874e+03, threshold=1.275e+03, percent-clipped=7.0 +2023-04-02 15:29:17,504 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=129252.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:29:19,488 INFO [train.py:903] (3/4) Epoch 19, batch 6350, loss[loss=0.2126, simple_loss=0.3002, pruned_loss=0.06249, over 18955.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2913, pruned_loss=0.0663, over 3823209.02 frames. ], batch size: 74, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:29:25,283 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129258.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:30:21,284 INFO [train.py:903] (3/4) Epoch 19, batch 6400, loss[loss=0.1996, simple_loss=0.2783, pruned_loss=0.06048, over 19725.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2902, pruned_loss=0.06601, over 3824345.32 frames. ], batch size: 51, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:30:36,125 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9544, 1.8760, 1.8116, 1.5066, 1.5018, 1.5365, 0.4683, 0.8266], + device='cuda:3'), covar=tensor([0.0562, 0.0565, 0.0380, 0.0632, 0.1037, 0.0787, 0.1145, 0.1004], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0347, 0.0350, 0.0374, 0.0449, 0.0380, 0.0329, 0.0336], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 15:30:36,856 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.331e+02 4.991e+02 6.008e+02 7.727e+02 1.608e+03, threshold=1.202e+03, percent-clipped=4.0 +2023-04-02 15:31:22,261 INFO [train.py:903] (3/4) Epoch 19, batch 6450, loss[loss=0.1693, simple_loss=0.2431, pruned_loss=0.04774, over 19732.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2903, pruned_loss=0.06558, over 3835706.20 frames. ], batch size: 45, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:31:39,037 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129367.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:32:06,817 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 15:32:22,354 INFO [train.py:903] (3/4) Epoch 19, batch 6500, loss[loss=0.164, simple_loss=0.2458, pruned_loss=0.04111, over 19712.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2901, pruned_loss=0.06562, over 3827378.15 frames. ], batch size: 45, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:32:29,749 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 15:32:38,766 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.146e+02 5.340e+02 6.897e+02 8.888e+02 1.987e+03, threshold=1.379e+03, percent-clipped=7.0 +2023-04-02 15:32:56,871 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129431.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:33:21,208 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1364, 1.9863, 1.9213, 1.6540, 1.5328, 1.6825, 0.5287, 1.0581], + device='cuda:3'), covar=tensor([0.0583, 0.0621, 0.0410, 0.0694, 0.1100, 0.0856, 0.1192, 0.0937], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0351, 0.0353, 0.0377, 0.0453, 0.0384, 0.0332, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 15:33:24,201 INFO [train.py:903] (3/4) Epoch 19, batch 6550, loss[loss=0.1936, simple_loss=0.2775, pruned_loss=0.05486, over 19053.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.292, pruned_loss=0.06696, over 3811522.55 frames. ], batch size: 69, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:34:01,180 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-02 15:34:03,300 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.04 vs. limit=5.0 +2023-04-02 15:34:24,464 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.56 vs. limit=2.0 +2023-04-02 15:34:24,939 INFO [train.py:903] (3/4) Epoch 19, batch 6600, loss[loss=0.2067, simple_loss=0.2907, pruned_loss=0.06136, over 19688.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2916, pruned_loss=0.06682, over 3819086.91 frames. ], batch size: 59, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:34:37,646 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129514.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:34:40,455 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.539e+02 5.847e+02 6.807e+02 8.552e+02 1.538e+03, threshold=1.361e+03, percent-clipped=4.0 +2023-04-02 15:35:07,250 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129539.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:35:25,978 INFO [train.py:903] (3/4) Epoch 19, batch 6650, loss[loss=0.1803, simple_loss=0.2605, pruned_loss=0.05003, over 19748.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.291, pruned_loss=0.06644, over 3806177.29 frames. ], batch size: 51, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:35:59,848 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3416, 1.1411, 1.2531, 1.3105, 2.1225, 1.1159, 1.8700, 2.3236], + device='cuda:3'), covar=tensor([0.0487, 0.2120, 0.2182, 0.1403, 0.0598, 0.1800, 0.1645, 0.0402], + device='cuda:3'), in_proj_covar=tensor([0.0395, 0.0357, 0.0376, 0.0339, 0.0366, 0.0347, 0.0368, 0.0389], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:36:26,349 INFO [train.py:903] (3/4) Epoch 19, batch 6700, loss[loss=0.1829, simple_loss=0.2635, pruned_loss=0.05111, over 19604.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2906, pruned_loss=0.0662, over 3812656.85 frames. ], batch size: 50, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:36:42,863 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.775e+02 4.923e+02 5.649e+02 7.598e+02 1.428e+03, threshold=1.130e+03, percent-clipped=1.0 +2023-04-02 15:36:51,107 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129623.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:36:58,789 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129630.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:37:17,658 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6138, 1.4726, 1.5060, 1.9672, 1.5479, 1.8822, 1.8745, 1.6560], + device='cuda:3'), covar=tensor([0.0799, 0.0895, 0.0994, 0.0670, 0.0819, 0.0679, 0.0826, 0.0714], + device='cuda:3'), in_proj_covar=tensor([0.0208, 0.0219, 0.0224, 0.0244, 0.0226, 0.0209, 0.0188, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 15:37:18,790 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129648.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:37:25,090 INFO [train.py:903] (3/4) Epoch 19, batch 6750, loss[loss=0.2047, simple_loss=0.2807, pruned_loss=0.06434, over 19752.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2905, pruned_loss=0.06622, over 3808823.61 frames. ], batch size: 54, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:38:20,255 INFO [train.py:903] (3/4) Epoch 19, batch 6800, loss[loss=0.2105, simple_loss=0.2894, pruned_loss=0.06582, over 19609.00 frames. ], tot_loss[loss=0.2125, simple_loss=0.2911, pruned_loss=0.06694, over 3812442.97 frames. ], batch size: 50, lr: 4.29e-03, grad_scale: 8.0 +2023-04-02 15:38:34,413 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.328e+02 4.884e+02 6.226e+02 8.201e+02 1.689e+03, threshold=1.245e+03, percent-clipped=11.0 +2023-04-02 15:39:05,129 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 15:39:06,149 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 15:39:08,367 INFO [train.py:903] (3/4) Epoch 20, batch 0, loss[loss=0.2291, simple_loss=0.2922, pruned_loss=0.08298, over 19306.00 frames. ], tot_loss[loss=0.2291, simple_loss=0.2922, pruned_loss=0.08298, over 19306.00 frames. ], batch size: 44, lr: 4.18e-03, grad_scale: 8.0 +2023-04-02 15:39:08,367 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 15:39:19,738 INFO [train.py:937] (3/4) Epoch 20, validation: loss=0.1695, simple_loss=0.2703, pruned_loss=0.03432, over 944034.00 frames. +2023-04-02 15:39:19,739 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 15:39:31,872 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 15:39:40,392 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8975, 1.6509, 1.6161, 1.9664, 1.7030, 1.6401, 1.5118, 1.8167], + device='cuda:3'), covar=tensor([0.1124, 0.1588, 0.1561, 0.1069, 0.1355, 0.0602, 0.1480, 0.0782], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0357, 0.0308, 0.0251, 0.0301, 0.0249, 0.0301, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:39:49,274 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7662, 1.2514, 1.5348, 1.4498, 3.3116, 1.1768, 2.3728, 3.7496], + device='cuda:3'), covar=tensor([0.0443, 0.2930, 0.2951, 0.1997, 0.0726, 0.2621, 0.1327, 0.0243], + device='cuda:3'), in_proj_covar=tensor([0.0396, 0.0358, 0.0377, 0.0340, 0.0366, 0.0348, 0.0371, 0.0390], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:39:50,422 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3204, 1.3589, 1.8078, 1.2774, 2.6132, 3.5432, 3.2863, 3.7787], + device='cuda:3'), covar=tensor([0.1507, 0.3668, 0.3149, 0.2383, 0.0567, 0.0183, 0.0211, 0.0246], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0317, 0.0348, 0.0263, 0.0239, 0.0182, 0.0214, 0.0249], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 15:40:12,661 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=129775.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:40:20,201 INFO [train.py:903] (3/4) Epoch 20, batch 50, loss[loss=0.1888, simple_loss=0.267, pruned_loss=0.05528, over 19773.00 frames. ], tot_loss[loss=0.213, simple_loss=0.2928, pruned_loss=0.06661, over 874181.36 frames. ], batch size: 49, lr: 4.18e-03, grad_scale: 8.0 +2023-04-02 15:40:51,343 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129809.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:40:54,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 15:41:03,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.371e+02 5.543e+02 6.891e+02 8.835e+02 1.770e+03, threshold=1.378e+03, percent-clipped=8.0 +2023-04-02 15:41:20,220 INFO [train.py:903] (3/4) Epoch 20, batch 100, loss[loss=0.2476, simple_loss=0.3177, pruned_loss=0.08879, over 13148.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2921, pruned_loss=0.06631, over 1522410.08 frames. ], batch size: 135, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:41:31,342 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 15:41:56,935 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3001, 1.8910, 1.8969, 2.6592, 1.7960, 2.4243, 2.3570, 2.2886], + device='cuda:3'), covar=tensor([0.0749, 0.0900, 0.0960, 0.0830, 0.0956, 0.0689, 0.0906, 0.0630], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0219, 0.0225, 0.0243, 0.0226, 0.0210, 0.0188, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 15:42:21,223 INFO [train.py:903] (3/4) Epoch 20, batch 150, loss[loss=0.2187, simple_loss=0.2988, pruned_loss=0.06925, over 19602.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2924, pruned_loss=0.066, over 2038359.15 frames. ], batch size: 50, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:42:30,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129890.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 15:43:03,460 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.452e+02 4.799e+02 5.935e+02 7.467e+02 3.197e+03, threshold=1.187e+03, percent-clipped=3.0 +2023-04-02 15:43:21,685 INFO [train.py:903] (3/4) Epoch 20, batch 200, loss[loss=0.1971, simple_loss=0.2649, pruned_loss=0.06466, over 19072.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2916, pruned_loss=0.06578, over 2437084.97 frames. ], batch size: 42, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:43:22,845 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 15:44:13,780 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=129974.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:44:23,550 INFO [train.py:903] (3/4) Epoch 20, batch 250, loss[loss=0.2243, simple_loss=0.3065, pruned_loss=0.07102, over 19659.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2922, pruned_loss=0.06582, over 2741986.01 frames. ], batch size: 60, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:45:06,933 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.070e+02 5.298e+02 6.354e+02 8.098e+02 1.543e+03, threshold=1.271e+03, percent-clipped=7.0 +2023-04-02 15:45:08,342 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130018.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:45:10,561 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130020.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:45:25,485 INFO [train.py:903] (3/4) Epoch 20, batch 300, loss[loss=0.1992, simple_loss=0.2915, pruned_loss=0.05349, over 19682.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2917, pruned_loss=0.06618, over 2997798.22 frames. ], batch size: 59, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:46:25,933 INFO [train.py:903] (3/4) Epoch 20, batch 350, loss[loss=0.2621, simple_loss=0.335, pruned_loss=0.09458, over 19537.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2902, pruned_loss=0.06577, over 3179680.85 frames. ], batch size: 56, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:46:35,018 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 15:46:35,365 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130089.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:46:50,775 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.65 vs. limit=5.0 +2023-04-02 15:47:08,654 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.196e+02 5.230e+02 6.397e+02 7.792e+02 1.393e+03, threshold=1.279e+03, percent-clipped=3.0 +2023-04-02 15:47:26,645 INFO [train.py:903] (3/4) Epoch 20, batch 400, loss[loss=0.1881, simple_loss=0.2753, pruned_loss=0.05048, over 19577.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2906, pruned_loss=0.06629, over 3316202.86 frames. ], batch size: 52, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:47:32,417 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7532, 4.2420, 4.4513, 4.4404, 1.7833, 4.1918, 3.6469, 4.1515], + device='cuda:3'), covar=tensor([0.1600, 0.0945, 0.0545, 0.0633, 0.5769, 0.0926, 0.0644, 0.0971], + device='cuda:3'), in_proj_covar=tensor([0.0759, 0.0712, 0.0914, 0.0800, 0.0811, 0.0672, 0.0550, 0.0849], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 15:47:42,950 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130146.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:47:53,134 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130153.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:48:14,998 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130171.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 15:48:27,089 INFO [train.py:903] (3/4) Epoch 20, batch 450, loss[loss=0.2104, simple_loss=0.2947, pruned_loss=0.063, over 17317.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2893, pruned_loss=0.06565, over 3440573.23 frames. ], batch size: 101, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:49:03,625 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 15:49:04,563 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 15:49:09,155 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.114e+02 5.003e+02 6.443e+02 8.043e+02 1.786e+03, threshold=1.289e+03, percent-clipped=5.0 +2023-04-02 15:49:27,174 INFO [train.py:903] (3/4) Epoch 20, batch 500, loss[loss=0.2029, simple_loss=0.2937, pruned_loss=0.05606, over 19544.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.29, pruned_loss=0.06585, over 3520180.76 frames. ], batch size: 56, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:49:37,967 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.45 vs. limit=2.0 +2023-04-02 15:50:10,067 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130268.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:50:15,785 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-02 15:50:27,988 INFO [train.py:903] (3/4) Epoch 20, batch 550, loss[loss=0.2365, simple_loss=0.3165, pruned_loss=0.07823, over 19700.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2894, pruned_loss=0.06544, over 3598288.58 frames. ], batch size: 59, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:51:11,220 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.373e+02 5.062e+02 6.327e+02 8.479e+02 2.088e+03, threshold=1.265e+03, percent-clipped=6.0 +2023-04-02 15:51:28,460 INFO [train.py:903] (3/4) Epoch 20, batch 600, loss[loss=0.2125, simple_loss=0.2933, pruned_loss=0.06583, over 19766.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2894, pruned_loss=0.06562, over 3657674.87 frames. ], batch size: 56, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:51:44,890 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130345.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:52:06,331 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130362.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:52:08,732 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130364.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:52:15,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 15:52:16,083 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130370.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:52:30,045 INFO [train.py:903] (3/4) Epoch 20, batch 650, loss[loss=0.2305, simple_loss=0.3068, pruned_loss=0.07705, over 18198.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.2903, pruned_loss=0.06621, over 3697268.43 frames. ], batch size: 83, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:52:56,085 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130402.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:53:13,835 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.680e+02 5.222e+02 6.307e+02 8.250e+02 2.391e+03, threshold=1.261e+03, percent-clipped=5.0 +2023-04-02 15:53:31,367 INFO [train.py:903] (3/4) Epoch 20, batch 700, loss[loss=0.2399, simple_loss=0.3316, pruned_loss=0.07412, over 19783.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2904, pruned_loss=0.06605, over 3715720.77 frames. ], batch size: 56, lr: 4.17e-03, grad_scale: 8.0 +2023-04-02 15:53:55,131 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130450.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:54:02,155 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130456.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:54:28,625 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130477.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:54:32,015 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:54:35,653 INFO [train.py:903] (3/4) Epoch 20, batch 750, loss[loss=0.2146, simple_loss=0.3006, pruned_loss=0.06433, over 19744.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2918, pruned_loss=0.06693, over 3728497.70 frames. ], batch size: 63, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:54:40,887 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0803, 2.1219, 2.3276, 2.1473, 3.1708, 2.7545, 3.2660, 2.2084], + device='cuda:3'), covar=tensor([0.1837, 0.3218, 0.2106, 0.1541, 0.1248, 0.1678, 0.1313, 0.3352], + device='cuda:3'), in_proj_covar=tensor([0.0522, 0.0628, 0.0692, 0.0472, 0.0612, 0.0522, 0.0654, 0.0537], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 15:55:19,213 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.433e+02 4.790e+02 6.042e+02 7.311e+02 1.890e+03, threshold=1.208e+03, percent-clipped=4.0 +2023-04-02 15:55:28,398 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130524.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:55:37,193 INFO [train.py:903] (3/4) Epoch 20, batch 800, loss[loss=0.2561, simple_loss=0.3237, pruned_loss=0.0943, over 13794.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2918, pruned_loss=0.06675, over 3729991.12 frames. ], batch size: 135, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:55:53,540 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 15:55:58,670 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130549.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:56:21,738 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130566.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:56:40,733 INFO [train.py:903] (3/4) Epoch 20, batch 850, loss[loss=0.2257, simple_loss=0.2958, pruned_loss=0.07778, over 19831.00 frames. ], tot_loss[loss=0.2142, simple_loss=0.2931, pruned_loss=0.06762, over 3727205.92 frames. ], batch size: 52, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:56:41,920 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:56:44,152 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1112, 1.1595, 1.4975, 1.3897, 2.7169, 1.1061, 2.2426, 2.9957], + device='cuda:3'), covar=tensor([0.0523, 0.2855, 0.2668, 0.1763, 0.0738, 0.2305, 0.1081, 0.0356], + device='cuda:3'), in_proj_covar=tensor([0.0397, 0.0355, 0.0375, 0.0337, 0.0366, 0.0344, 0.0370, 0.0390], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 15:56:57,585 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4044, 1.4950, 1.8231, 1.4614, 2.3676, 2.6948, 2.6067, 2.8542], + device='cuda:3'), covar=tensor([0.1344, 0.3107, 0.2735, 0.2382, 0.1096, 0.0386, 0.0252, 0.0349], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0315, 0.0346, 0.0261, 0.0237, 0.0182, 0.0213, 0.0247], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 15:57:25,285 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.130e+02 4.920e+02 5.760e+02 7.852e+02 1.760e+03, threshold=1.152e+03, percent-clipped=6.0 +2023-04-02 15:57:33,245 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 15:57:40,812 INFO [train.py:903] (3/4) Epoch 20, batch 900, loss[loss=0.2375, simple_loss=0.3113, pruned_loss=0.08186, over 19700.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2942, pruned_loss=0.06852, over 3748024.59 frames. ], batch size: 59, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:58:44,441 INFO [train.py:903] (3/4) Epoch 20, batch 950, loss[loss=0.2069, simple_loss=0.2955, pruned_loss=0.05911, over 18215.00 frames. ], tot_loss[loss=0.2153, simple_loss=0.294, pruned_loss=0.0683, over 3761921.20 frames. ], batch size: 83, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:58:47,660 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 15:59:28,722 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.021e+02 4.928e+02 5.917e+02 7.294e+02 1.421e+03, threshold=1.183e+03, percent-clipped=3.0 +2023-04-02 15:59:46,665 INFO [train.py:903] (3/4) Epoch 20, batch 1000, loss[loss=0.2229, simple_loss=0.3047, pruned_loss=0.07059, over 19432.00 frames. ], tot_loss[loss=0.2147, simple_loss=0.2931, pruned_loss=0.06816, over 3762279.83 frames. ], batch size: 70, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 15:59:48,220 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130733.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 15:59:50,480 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130735.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:00:03,671 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130746.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:00:18,654 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130758.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:00:20,910 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130760.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:00:38,585 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 16:00:48,485 INFO [train.py:903] (3/4) Epoch 20, batch 1050, loss[loss=0.2227, simple_loss=0.2989, pruned_loss=0.07324, over 18282.00 frames. ], tot_loss[loss=0.2145, simple_loss=0.2927, pruned_loss=0.06815, over 3754114.41 frames. ], batch size: 84, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 16:01:00,660 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8043, 4.3123, 4.5119, 4.5419, 1.7087, 4.2727, 3.7394, 4.2664], + device='cuda:3'), covar=tensor([0.1480, 0.0709, 0.0557, 0.0610, 0.5592, 0.0754, 0.0591, 0.1025], + device='cuda:3'), in_proj_covar=tensor([0.0759, 0.0717, 0.0918, 0.0801, 0.0817, 0.0677, 0.0555, 0.0857], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 16:01:02,849 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130794.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:01:10,718 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130800.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:01:20,589 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 16:01:33,079 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.731e+02 5.562e+02 6.742e+02 8.268e+02 2.102e+03, threshold=1.348e+03, percent-clipped=2.0 +2023-04-02 16:01:49,835 INFO [train.py:903] (3/4) Epoch 20, batch 1100, loss[loss=0.1989, simple_loss=0.2738, pruned_loss=0.06196, over 19727.00 frames. ], tot_loss[loss=0.2141, simple_loss=0.2926, pruned_loss=0.0678, over 3772318.53 frames. ], batch size: 51, lr: 4.16e-03, grad_scale: 4.0 +2023-04-02 16:02:28,187 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130861.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:02:42,048 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3888, 2.4609, 2.0084, 2.5781, 2.4075, 1.8681, 2.0302, 2.2314], + device='cuda:3'), covar=tensor([0.1152, 0.1750, 0.1669, 0.1108, 0.1490, 0.0766, 0.1552, 0.0860], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0360, 0.0312, 0.0252, 0.0302, 0.0251, 0.0305, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:02:52,361 INFO [train.py:903] (3/4) Epoch 20, batch 1150, loss[loss=0.1765, simple_loss=0.256, pruned_loss=0.04845, over 17727.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2927, pruned_loss=0.06793, over 3780138.58 frames. ], batch size: 39, lr: 4.16e-03, grad_scale: 4.0 +2023-04-02 16:02:53,748 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8584, 1.6027, 1.4629, 1.8296, 1.5131, 1.5646, 1.4696, 1.6764], + device='cuda:3'), covar=tensor([0.1013, 0.1284, 0.1489, 0.0960, 0.1190, 0.0573, 0.1318, 0.0748], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0359, 0.0311, 0.0252, 0.0301, 0.0251, 0.0304, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:03:26,741 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130909.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:03:27,655 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130910.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:03:33,842 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130915.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:03:39,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.314e+02 5.063e+02 6.056e+02 7.993e+02 1.743e+03, threshold=1.211e+03, percent-clipped=5.0 +2023-04-02 16:03:50,408 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=130927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:03:55,686 INFO [train.py:903] (3/4) Epoch 20, batch 1200, loss[loss=0.2464, simple_loss=0.3208, pruned_loss=0.08596, over 19744.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2911, pruned_loss=0.06722, over 3790491.43 frames. ], batch size: 63, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 16:04:23,990 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 16:04:56,114 INFO [train.py:903] (3/4) Epoch 20, batch 1250, loss[loss=0.2188, simple_loss=0.3079, pruned_loss=0.06481, over 19788.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2905, pruned_loss=0.06686, over 3786619.57 frames. ], batch size: 56, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 16:05:07,180 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 16:05:42,762 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.304e+02 5.121e+02 6.297e+02 7.673e+02 2.016e+03, threshold=1.259e+03, percent-clipped=4.0 +2023-04-02 16:05:51,140 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131025.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:05:58,785 INFO [train.py:903] (3/4) Epoch 20, batch 1300, loss[loss=0.1965, simple_loss=0.2818, pruned_loss=0.05563, over 19658.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2895, pruned_loss=0.06613, over 3805056.74 frames. ], batch size: 55, lr: 4.16e-03, grad_scale: 8.0 +2023-04-02 16:06:12,061 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131042.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:06:31,638 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1737, 5.6524, 3.1971, 4.9814, 1.3844, 5.7152, 5.5678, 5.6952], + device='cuda:3'), covar=tensor([0.0325, 0.0750, 0.1647, 0.0631, 0.3632, 0.0489, 0.0636, 0.0952], + device='cuda:3'), in_proj_covar=tensor([0.0491, 0.0395, 0.0484, 0.0343, 0.0400, 0.0421, 0.0415, 0.0447], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:06:51,900 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9995, 4.4355, 4.7618, 4.7506, 1.8294, 4.4389, 3.8715, 4.4294], + device='cuda:3'), covar=tensor([0.1630, 0.0902, 0.0528, 0.0617, 0.5626, 0.0811, 0.0616, 0.1055], + device='cuda:3'), in_proj_covar=tensor([0.0758, 0.0716, 0.0916, 0.0798, 0.0814, 0.0675, 0.0551, 0.0855], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 16:06:59,468 INFO [train.py:903] (3/4) Epoch 20, batch 1350, loss[loss=0.2003, simple_loss=0.2799, pruned_loss=0.06035, over 19832.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2896, pruned_loss=0.0662, over 3812077.11 frames. ], batch size: 52, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:07:43,044 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131117.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:07:44,787 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.358e+02 5.090e+02 6.517e+02 8.267e+02 2.193e+03, threshold=1.303e+03, percent-clipped=8.0 +2023-04-02 16:08:02,292 INFO [train.py:903] (3/4) Epoch 20, batch 1400, loss[loss=0.228, simple_loss=0.3152, pruned_loss=0.07037, over 19677.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2892, pruned_loss=0.06567, over 3819893.26 frames. ], batch size: 58, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:08:15,155 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131142.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:08:42,766 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:08:50,703 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131171.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:08:51,746 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6597, 1.4966, 1.5625, 2.2804, 1.7414, 2.0944, 2.0902, 1.7241], + device='cuda:3'), covar=tensor([0.0843, 0.0979, 0.1068, 0.0745, 0.0837, 0.0738, 0.0909, 0.0733], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0219, 0.0225, 0.0242, 0.0226, 0.0209, 0.0188, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 16:09:03,070 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3864, 1.4382, 1.6748, 1.6266, 2.5623, 2.1574, 2.5743, 1.1264], + device='cuda:3'), covar=tensor([0.2490, 0.4263, 0.2729, 0.1936, 0.1454, 0.2158, 0.1408, 0.4365], + device='cuda:3'), in_proj_covar=tensor([0.0526, 0.0630, 0.0695, 0.0475, 0.0614, 0.0526, 0.0658, 0.0541], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 16:09:03,834 INFO [train.py:903] (3/4) Epoch 20, batch 1450, loss[loss=0.2167, simple_loss=0.3004, pruned_loss=0.06649, over 19675.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2891, pruned_loss=0.06581, over 3828789.29 frames. ], batch size: 60, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:09:06,070 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 16:09:14,530 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131190.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:09:16,774 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131192.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:09:21,594 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131196.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:09:50,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.524e+02 5.028e+02 6.181e+02 7.641e+02 1.699e+03, threshold=1.236e+03, percent-clipped=6.0 +2023-04-02 16:10:06,714 INFO [train.py:903] (3/4) Epoch 20, batch 1500, loss[loss=0.1566, simple_loss=0.233, pruned_loss=0.0401, over 19004.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2897, pruned_loss=0.06597, over 3820423.14 frames. ], batch size: 42, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:10:20,896 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0013, 1.9087, 1.7530, 1.5228, 1.4257, 1.5559, 0.3754, 0.8863], + device='cuda:3'), covar=tensor([0.0535, 0.0573, 0.0402, 0.0685, 0.1143, 0.0794, 0.1182, 0.0971], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0349, 0.0352, 0.0377, 0.0452, 0.0383, 0.0332, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 16:11:06,980 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131281.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:11:07,712 INFO [train.py:903] (3/4) Epoch 20, batch 1550, loss[loss=0.236, simple_loss=0.3149, pruned_loss=0.07856, over 19663.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2908, pruned_loss=0.0665, over 3817252.92 frames. ], batch size: 55, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:11:29,267 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131298.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:11:38,467 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131306.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:11:53,760 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.770e+02 5.087e+02 6.243e+02 7.473e+02 1.350e+03, threshold=1.249e+03, percent-clipped=1.0 +2023-04-02 16:11:58,753 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131323.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:12:10,212 INFO [train.py:903] (3/4) Epoch 20, batch 1600, loss[loss=0.2141, simple_loss=0.3114, pruned_loss=0.05846, over 19700.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2917, pruned_loss=0.0668, over 3806864.68 frames. ], batch size: 59, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:12:36,128 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 16:13:12,820 INFO [train.py:903] (3/4) Epoch 20, batch 1650, loss[loss=0.2259, simple_loss=0.3089, pruned_loss=0.0714, over 19276.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2917, pruned_loss=0.06674, over 3820729.21 frames. ], batch size: 66, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:13:59,216 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.697e+02 5.218e+02 6.304e+02 8.075e+02 1.501e+03, threshold=1.261e+03, percent-clipped=5.0 +2023-04-02 16:14:08,616 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131427.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:14:15,064 INFO [train.py:903] (3/4) Epoch 20, batch 1700, loss[loss=0.2418, simple_loss=0.3172, pruned_loss=0.08314, over 19545.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2911, pruned_loss=0.06652, over 3823643.76 frames. ], batch size: 54, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:14:17,629 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131434.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:14:51,571 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5771, 1.4618, 1.7574, 1.6923, 3.8756, 1.1194, 2.7600, 4.2240], + device='cuda:3'), covar=tensor([0.0525, 0.3406, 0.3148, 0.2206, 0.1029, 0.3211, 0.1498, 0.0330], + device='cuda:3'), in_proj_covar=tensor([0.0396, 0.0356, 0.0376, 0.0338, 0.0365, 0.0346, 0.0370, 0.0391], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:14:55,854 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 16:15:13,169 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0514, 3.5749, 2.0182, 2.1744, 3.0867, 1.9492, 1.5797, 2.2819], + device='cuda:3'), covar=tensor([0.1508, 0.0607, 0.1096, 0.0888, 0.0555, 0.1172, 0.0973, 0.0698], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0313, 0.0331, 0.0261, 0.0245, 0.0335, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:15:16,179 INFO [train.py:903] (3/4) Epoch 20, batch 1750, loss[loss=0.1777, simple_loss=0.2578, pruned_loss=0.04883, over 19390.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.291, pruned_loss=0.06677, over 3804196.26 frames. ], batch size: 48, lr: 4.15e-03, grad_scale: 8.0 +2023-04-02 16:15:52,678 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131510.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:16:01,414 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-02 16:16:02,700 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.468e+02 5.359e+02 6.412e+02 7.691e+02 1.507e+03, threshold=1.282e+03, percent-clipped=3.0 +2023-04-02 16:16:18,695 INFO [train.py:903] (3/4) Epoch 20, batch 1800, loss[loss=0.2218, simple_loss=0.2881, pruned_loss=0.07775, over 19389.00 frames. ], tot_loss[loss=0.2118, simple_loss=0.2906, pruned_loss=0.06654, over 3809071.04 frames. ], batch size: 48, lr: 4.15e-03, grad_scale: 4.0 +2023-04-02 16:16:24,467 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=131536.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:17:16,153 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 16:17:22,072 INFO [train.py:903] (3/4) Epoch 20, batch 1850, loss[loss=0.2018, simple_loss=0.2831, pruned_loss=0.06024, over 19844.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2917, pruned_loss=0.06724, over 3798155.92 frames. ], batch size: 52, lr: 4.15e-03, grad_scale: 4.0 +2023-04-02 16:17:54,263 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 16:18:09,983 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.335e+02 4.826e+02 6.634e+02 9.045e+02 2.049e+03, threshold=1.327e+03, percent-clipped=6.0 +2023-04-02 16:18:25,168 INFO [train.py:903] (3/4) Epoch 20, batch 1900, loss[loss=0.2261, simple_loss=0.2988, pruned_loss=0.07668, over 19784.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2907, pruned_loss=0.06622, over 3814782.39 frames. ], batch size: 56, lr: 4.15e-03, grad_scale: 4.0 +2023-04-02 16:18:40,091 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 16:18:45,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 16:18:47,991 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131651.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:19:10,735 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 16:19:25,447 INFO [train.py:903] (3/4) Epoch 20, batch 1950, loss[loss=0.2236, simple_loss=0.3101, pruned_loss=0.06858, over 19614.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2915, pruned_loss=0.06668, over 3807671.74 frames. ], batch size: 57, lr: 4.15e-03, grad_scale: 4.0 +2023-04-02 16:20:11,291 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7681, 3.2724, 3.3089, 3.3237, 1.4106, 3.1944, 2.7738, 3.0687], + device='cuda:3'), covar=tensor([0.1874, 0.1053, 0.0852, 0.0939, 0.5452, 0.1030, 0.0875, 0.1335], + device='cuda:3'), in_proj_covar=tensor([0.0766, 0.0718, 0.0922, 0.0810, 0.0818, 0.0683, 0.0555, 0.0860], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 16:20:13,300 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.458e+02 5.129e+02 6.435e+02 8.344e+02 2.370e+03, threshold=1.287e+03, percent-clipped=2.0 +2023-04-02 16:20:28,635 INFO [train.py:903] (3/4) Epoch 20, batch 2000, loss[loss=0.2266, simple_loss=0.3044, pruned_loss=0.07436, over 19530.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2924, pruned_loss=0.06718, over 3812584.69 frames. ], batch size: 56, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:21:18,598 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=131771.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:21:25,592 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 16:21:28,789 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=131778.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:21:33,331 INFO [train.py:903] (3/4) Epoch 20, batch 2050, loss[loss=0.1914, simple_loss=0.2799, pruned_loss=0.0515, over 18841.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.2924, pruned_loss=0.06732, over 3794237.52 frames. ], batch size: 74, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:21:45,621 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 16:21:46,760 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 16:22:07,045 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0318, 1.9507, 1.7682, 1.5360, 1.5057, 1.5732, 0.3508, 0.8968], + device='cuda:3'), covar=tensor([0.0599, 0.0645, 0.0428, 0.0710, 0.1122, 0.0816, 0.1224, 0.1018], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0351, 0.0354, 0.0380, 0.0454, 0.0384, 0.0334, 0.0337], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 16:22:07,705 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 16:22:22,559 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.239e+02 4.827e+02 6.005e+02 7.777e+02 1.829e+03, threshold=1.201e+03, percent-clipped=4.0 +2023-04-02 16:22:35,844 INFO [train.py:903] (3/4) Epoch 20, batch 2100, loss[loss=0.2035, simple_loss=0.2766, pruned_loss=0.06519, over 19747.00 frames. ], tot_loss[loss=0.2143, simple_loss=0.2928, pruned_loss=0.06784, over 3802538.92 frames. ], batch size: 51, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:23:02,420 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 16:23:03,499 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=131854.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:23:26,836 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 16:23:29,495 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131875.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:23:37,624 INFO [train.py:903] (3/4) Epoch 20, batch 2150, loss[loss=0.2327, simple_loss=0.3073, pruned_loss=0.07909, over 19333.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2914, pruned_loss=0.06711, over 3813343.78 frames. ], batch size: 66, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:23:42,694 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131886.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:23:51,980 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131893.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:24:09,611 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131907.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:24:26,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.293e+02 5.118e+02 6.039e+02 8.265e+02 1.505e+03, threshold=1.208e+03, percent-clipped=4.0 +2023-04-02 16:24:39,702 INFO [train.py:903] (3/4) Epoch 20, batch 2200, loss[loss=0.2171, simple_loss=0.2962, pruned_loss=0.06898, over 19540.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2915, pruned_loss=0.06657, over 3823284.43 frames. ], batch size: 54, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:24:40,123 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131932.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:24:42,801 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-04-02 16:25:26,165 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131969.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:25:42,873 INFO [train.py:903] (3/4) Epoch 20, batch 2250, loss[loss=0.199, simple_loss=0.2652, pruned_loss=0.06638, over 19684.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.2917, pruned_loss=0.06727, over 3805227.41 frames. ], batch size: 45, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:25:52,172 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1701, 1.7989, 1.4197, 1.2345, 1.6253, 1.1597, 1.1365, 1.6430], + device='cuda:3'), covar=tensor([0.0823, 0.0837, 0.1112, 0.0807, 0.0510, 0.1328, 0.0658, 0.0465], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0315, 0.0333, 0.0261, 0.0245, 0.0336, 0.0291, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:26:31,963 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.368e+02 4.978e+02 6.272e+02 7.965e+02 1.499e+03, threshold=1.254e+03, percent-clipped=2.0 +2023-04-02 16:26:44,540 INFO [train.py:903] (3/4) Epoch 20, batch 2300, loss[loss=0.191, simple_loss=0.2715, pruned_loss=0.05522, over 19416.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2915, pruned_loss=0.06713, over 3822515.96 frames. ], batch size: 48, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:26:58,062 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 16:27:06,368 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132049.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:27:35,729 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132072.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:27:47,312 INFO [train.py:903] (3/4) Epoch 20, batch 2350, loss[loss=0.1965, simple_loss=0.2707, pruned_loss=0.06111, over 19675.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2909, pruned_loss=0.06669, over 3838805.94 frames. ], batch size: 53, lr: 4.14e-03, grad_scale: 4.0 +2023-04-02 16:28:26,825 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 16:28:35,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.260e+02 4.795e+02 5.776e+02 7.778e+02 1.972e+03, threshold=1.155e+03, percent-clipped=8.0 +2023-04-02 16:28:42,740 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 16:28:49,774 INFO [train.py:903] (3/4) Epoch 20, batch 2400, loss[loss=0.2222, simple_loss=0.309, pruned_loss=0.06769, over 19732.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2912, pruned_loss=0.06705, over 3816799.59 frames. ], batch size: 63, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:29:03,490 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132142.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:29:11,530 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132149.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:29:32,902 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132167.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:29:41,722 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132174.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:29:51,982 INFO [train.py:903] (3/4) Epoch 20, batch 2450, loss[loss=0.1982, simple_loss=0.2884, pruned_loss=0.054, over 19519.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2915, pruned_loss=0.067, over 3825477.37 frames. ], batch size: 54, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:30:38,913 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=132219.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:30:41,019 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 4.995e+02 6.263e+02 8.063e+02 1.363e+03, threshold=1.253e+03, percent-clipped=5.0 +2023-04-02 16:30:47,015 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132225.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:30:54,271 INFO [train.py:903] (3/4) Epoch 20, batch 2500, loss[loss=0.2201, simple_loss=0.2988, pruned_loss=0.07066, over 19752.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2914, pruned_loss=0.06669, over 3842083.33 frames. ], batch size: 63, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:31:06,001 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1238, 1.8045, 1.7805, 2.0402, 1.7557, 1.8856, 1.7382, 1.9604], + device='cuda:3'), covar=tensor([0.0961, 0.1538, 0.1370, 0.1033, 0.1359, 0.0491, 0.1322, 0.0728], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0354, 0.0308, 0.0249, 0.0299, 0.0248, 0.0304, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:31:15,979 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132250.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:31:43,685 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-02 16:31:56,371 INFO [train.py:903] (3/4) Epoch 20, batch 2550, loss[loss=0.2135, simple_loss=0.2942, pruned_loss=0.06635, over 19598.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2907, pruned_loss=0.06618, over 3847072.10 frames. ], batch size: 61, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:32:30,461 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9773, 1.3619, 1.0507, 1.0549, 1.1898, 1.0378, 0.9806, 1.2921], + device='cuda:3'), covar=tensor([0.0580, 0.0804, 0.1140, 0.0700, 0.0568, 0.1329, 0.0551, 0.0480], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0316, 0.0333, 0.0260, 0.0247, 0.0336, 0.0292, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:32:45,605 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.513e+02 4.985e+02 5.787e+02 8.066e+02 1.995e+03, threshold=1.157e+03, percent-clipped=4.0 +2023-04-02 16:32:52,689 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 16:32:58,517 INFO [train.py:903] (3/4) Epoch 20, batch 2600, loss[loss=0.2406, simple_loss=0.3221, pruned_loss=0.07954, over 18710.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2905, pruned_loss=0.06578, over 3843178.73 frames. ], batch size: 74, lr: 4.14e-03, grad_scale: 8.0 +2023-04-02 16:33:02,441 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132334.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:34:01,398 INFO [train.py:903] (3/4) Epoch 20, batch 2650, loss[loss=0.2196, simple_loss=0.2929, pruned_loss=0.07319, over 19362.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2916, pruned_loss=0.06641, over 3844124.69 frames. ], batch size: 47, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:34:14,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8181, 1.4521, 1.5384, 1.5332, 3.4024, 1.0826, 2.4465, 3.8283], + device='cuda:3'), covar=tensor([0.0437, 0.2662, 0.2762, 0.1875, 0.0696, 0.2631, 0.1247, 0.0228], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0357, 0.0377, 0.0340, 0.0368, 0.0348, 0.0373, 0.0393], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:34:15,341 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=132393.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:34:23,149 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 16:34:44,237 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=132416.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:34:50,450 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.050e+02 4.865e+02 6.149e+02 7.368e+02 1.585e+03, threshold=1.230e+03, percent-clipped=4.0 +2023-04-02 16:35:04,063 INFO [train.py:903] (3/4) Epoch 20, batch 2700, loss[loss=0.2546, simple_loss=0.3226, pruned_loss=0.0933, over 19657.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2907, pruned_loss=0.06617, over 3838354.59 frames. ], batch size: 60, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:36:00,514 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5589, 1.6769, 1.9528, 1.8381, 2.6172, 2.4299, 2.7423, 1.1204], + device='cuda:3'), covar=tensor([0.2302, 0.4008, 0.2510, 0.1739, 0.1490, 0.1882, 0.1476, 0.4350], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0636, 0.0700, 0.0479, 0.0618, 0.0528, 0.0662, 0.0543], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 16:36:06,700 INFO [train.py:903] (3/4) Epoch 20, batch 2750, loss[loss=0.1884, simple_loss=0.2612, pruned_loss=0.05775, over 19751.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2894, pruned_loss=0.06575, over 3831584.70 frames. ], batch size: 46, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:36:39,228 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132508.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:36:55,660 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.965e+02 5.186e+02 6.180e+02 7.968e+02 1.505e+03, threshold=1.236e+03, percent-clipped=2.0 +2023-04-02 16:36:59,734 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0772, 3.4142, 2.0249, 1.6824, 3.2200, 1.6517, 1.5132, 2.4891], + device='cuda:3'), covar=tensor([0.1254, 0.0600, 0.1065, 0.1140, 0.0474, 0.1326, 0.0990, 0.0573], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0316, 0.0335, 0.0261, 0.0247, 0.0336, 0.0292, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:37:07,829 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132531.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:37:08,668 INFO [train.py:903] (3/4) Epoch 20, batch 2800, loss[loss=0.2442, simple_loss=0.3119, pruned_loss=0.08819, over 13658.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2893, pruned_loss=0.06559, over 3826396.27 frames. ], batch size: 136, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:37:34,501 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-02 16:38:13,010 INFO [train.py:903] (3/4) Epoch 20, batch 2850, loss[loss=0.2093, simple_loss=0.288, pruned_loss=0.06527, over 19479.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2891, pruned_loss=0.06539, over 3826890.39 frames. ], batch size: 49, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:38:22,432 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132590.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:38:45,315 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-04-02 16:38:52,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132615.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:39:01,717 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.639e+02 4.861e+02 5.815e+02 7.642e+02 3.357e+03, threshold=1.163e+03, percent-clipped=7.0 +2023-04-02 16:39:14,636 INFO [train.py:903] (3/4) Epoch 20, batch 2900, loss[loss=0.2195, simple_loss=0.2792, pruned_loss=0.07992, over 19759.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2894, pruned_loss=0.06539, over 3840401.64 frames. ], batch size: 46, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:39:14,680 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 16:40:13,481 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132678.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:40:18,639 INFO [train.py:903] (3/4) Epoch 20, batch 2950, loss[loss=0.222, simple_loss=0.3011, pruned_loss=0.07146, over 19517.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2898, pruned_loss=0.06546, over 3843720.29 frames. ], batch size: 54, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:41:09,139 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.010e+02 4.627e+02 5.679e+02 7.371e+02 2.153e+03, threshold=1.136e+03, percent-clipped=3.0 +2023-04-02 16:41:20,811 INFO [train.py:903] (3/4) Epoch 20, batch 3000, loss[loss=0.2078, simple_loss=0.2947, pruned_loss=0.06044, over 19640.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2891, pruned_loss=0.06542, over 3842014.63 frames. ], batch size: 58, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:41:20,811 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 16:41:34,260 INFO [train.py:937] (3/4) Epoch 20, validation: loss=0.1695, simple_loss=0.2697, pruned_loss=0.03462, over 944034.00 frames. +2023-04-02 16:41:34,261 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 16:41:40,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 16:42:12,989 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132764.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:42:13,128 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132764.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:42:23,090 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3518, 3.5964, 2.2337, 2.2927, 3.3543, 1.9624, 1.5595, 2.5332], + device='cuda:3'), covar=tensor([0.1264, 0.0545, 0.0905, 0.0854, 0.0490, 0.1147, 0.0976, 0.0588], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0311, 0.0329, 0.0257, 0.0243, 0.0333, 0.0288, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:42:35,226 INFO [train.py:903] (3/4) Epoch 20, batch 3050, loss[loss=0.1913, simple_loss=0.2709, pruned_loss=0.0558, over 19659.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2888, pruned_loss=0.0656, over 3848034.61 frames. ], batch size: 53, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:42:41,474 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132787.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:42:43,583 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132789.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:43:13,139 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132812.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:43:24,237 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.615e+02 4.785e+02 6.187e+02 7.720e+02 1.879e+03, threshold=1.237e+03, percent-clipped=7.0 +2023-04-02 16:43:37,018 INFO [train.py:903] (3/4) Epoch 20, batch 3100, loss[loss=0.2262, simple_loss=0.3165, pruned_loss=0.06796, over 17946.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2893, pruned_loss=0.06556, over 3835161.75 frames. ], batch size: 83, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:44:40,224 INFO [train.py:903] (3/4) Epoch 20, batch 3150, loss[loss=0.2139, simple_loss=0.2874, pruned_loss=0.07022, over 19777.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2882, pruned_loss=0.06502, over 3837076.30 frames. ], batch size: 47, lr: 4.13e-03, grad_scale: 4.0 +2023-04-02 16:45:07,812 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 16:45:29,824 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 5.025e+02 5.951e+02 7.011e+02 1.371e+03, threshold=1.190e+03, percent-clipped=2.0 +2023-04-02 16:45:42,517 INFO [train.py:903] (3/4) Epoch 20, batch 3200, loss[loss=0.2036, simple_loss=0.2702, pruned_loss=0.0685, over 19789.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2885, pruned_loss=0.06544, over 3836207.93 frames. ], batch size: 48, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:46:46,032 INFO [train.py:903] (3/4) Epoch 20, batch 3250, loss[loss=0.2908, simple_loss=0.3613, pruned_loss=0.1101, over 19335.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2892, pruned_loss=0.06583, over 3838988.81 frames. ], batch size: 66, lr: 4.13e-03, grad_scale: 8.0 +2023-04-02 16:47:37,647 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.383e+02 4.867e+02 6.333e+02 8.818e+02 1.782e+03, threshold=1.267e+03, percent-clipped=7.0 +2023-04-02 16:47:37,839 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=133022.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:47:49,130 INFO [train.py:903] (3/4) Epoch 20, batch 3300, loss[loss=0.2711, simple_loss=0.3265, pruned_loss=0.1078, over 13334.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2892, pruned_loss=0.06613, over 3814420.14 frames. ], batch size: 137, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:47:50,544 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5049, 2.0606, 1.5825, 1.4319, 1.9656, 1.2992, 1.3809, 1.8504], + device='cuda:3'), covar=tensor([0.0978, 0.0870, 0.1012, 0.0833, 0.0478, 0.1256, 0.0709, 0.0505], + device='cuda:3'), in_proj_covar=tensor([0.0299, 0.0312, 0.0332, 0.0259, 0.0244, 0.0334, 0.0288, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:47:57,157 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 16:48:26,007 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5283, 1.5685, 1.9132, 1.8485, 2.8317, 2.5389, 2.9460, 1.3484], + device='cuda:3'), covar=tensor([0.2352, 0.4197, 0.2702, 0.1828, 0.1426, 0.1889, 0.1411, 0.4156], + device='cuda:3'), in_proj_covar=tensor([0.0526, 0.0633, 0.0697, 0.0478, 0.0614, 0.0525, 0.0658, 0.0541], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 16:48:38,078 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-02 16:48:54,786 INFO [train.py:903] (3/4) Epoch 20, batch 3350, loss[loss=0.2048, simple_loss=0.2705, pruned_loss=0.06958, over 19762.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.288, pruned_loss=0.06514, over 3823501.71 frames. ], batch size: 47, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:49:27,433 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=133108.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:49:45,520 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.359e+02 4.956e+02 6.093e+02 7.175e+02 1.819e+03, threshold=1.219e+03, percent-clipped=1.0 +2023-04-02 16:49:57,529 INFO [train.py:903] (3/4) Epoch 20, batch 3400, loss[loss=0.2336, simple_loss=0.3075, pruned_loss=0.0799, over 19758.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2897, pruned_loss=0.06621, over 3818241.17 frames. ], batch size: 56, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:50:06,221 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133137.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:51:02,180 INFO [train.py:903] (3/4) Epoch 20, batch 3450, loss[loss=0.1883, simple_loss=0.2698, pruned_loss=0.05341, over 19742.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2903, pruned_loss=0.06673, over 3820046.83 frames. ], batch size: 51, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:51:08,031 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 16:51:52,698 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.981e+02 4.746e+02 5.634e+02 7.504e+02 1.582e+03, threshold=1.127e+03, percent-clipped=2.0 +2023-04-02 16:51:54,150 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133223.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:52:04,226 INFO [train.py:903] (3/4) Epoch 20, batch 3500, loss[loss=0.2181, simple_loss=0.2916, pruned_loss=0.07231, over 19665.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2902, pruned_loss=0.06701, over 3830241.22 frames. ], batch size: 53, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:52:56,236 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8972, 5.0233, 5.7395, 5.7264, 2.0455, 5.4216, 4.6358, 5.4278], + device='cuda:3'), covar=tensor([0.1787, 0.0777, 0.0572, 0.0653, 0.6091, 0.0802, 0.0561, 0.1198], + device='cuda:3'), in_proj_covar=tensor([0.0758, 0.0717, 0.0915, 0.0803, 0.0815, 0.0676, 0.0549, 0.0849], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 16:53:08,050 INFO [train.py:903] (3/4) Epoch 20, batch 3550, loss[loss=0.1885, simple_loss=0.2689, pruned_loss=0.05404, over 19741.00 frames. ], tot_loss[loss=0.213, simple_loss=0.291, pruned_loss=0.06751, over 3830621.21 frames. ], batch size: 51, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:53:34,557 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-02 16:53:58,718 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.076e+02 5.017e+02 5.933e+02 7.980e+02 2.795e+03, threshold=1.187e+03, percent-clipped=11.0 +2023-04-02 16:54:10,372 INFO [train.py:903] (3/4) Epoch 20, batch 3600, loss[loss=0.2265, simple_loss=0.3082, pruned_loss=0.0724, over 19768.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2911, pruned_loss=0.0673, over 3834334.67 frames. ], batch size: 54, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:55:15,915 INFO [train.py:903] (3/4) Epoch 20, batch 3650, loss[loss=0.215, simple_loss=0.2861, pruned_loss=0.07195, over 19404.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2903, pruned_loss=0.06676, over 3819519.80 frames. ], batch size: 48, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:55:28,146 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133392.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 16:55:29,490 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133393.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:55:30,503 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133394.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:56:02,492 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133418.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:56:06,669 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.934e+02 5.113e+02 6.456e+02 7.889e+02 1.610e+03, threshold=1.291e+03, percent-clipped=2.0 +2023-04-02 16:56:18,436 INFO [train.py:903] (3/4) Epoch 20, batch 3700, loss[loss=0.188, simple_loss=0.2652, pruned_loss=0.05539, over 19034.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2905, pruned_loss=0.06702, over 3821723.12 frames. ], batch size: 42, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:56:21,232 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.3370, 5.7464, 3.4747, 5.1137, 1.1104, 5.8090, 5.7374, 5.8883], + device='cuda:3'), covar=tensor([0.0369, 0.0955, 0.1735, 0.0671, 0.4382, 0.0548, 0.0714, 0.0757], + device='cuda:3'), in_proj_covar=tensor([0.0489, 0.0396, 0.0480, 0.0341, 0.0399, 0.0420, 0.0412, 0.0446], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 16:57:19,839 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:57:23,017 INFO [train.py:903] (3/4) Epoch 20, batch 3750, loss[loss=0.2362, simple_loss=0.3181, pruned_loss=0.07712, over 19370.00 frames. ], tot_loss[loss=0.2115, simple_loss=0.2898, pruned_loss=0.06661, over 3816243.63 frames. ], batch size: 70, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:57:34,894 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133492.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:57:50,980 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133504.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 16:58:13,502 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.259e+02 5.077e+02 5.935e+02 8.206e+02 1.595e+03, threshold=1.187e+03, percent-clipped=3.0 +2023-04-02 16:58:24,815 INFO [train.py:903] (3/4) Epoch 20, batch 3800, loss[loss=0.2869, simple_loss=0.3442, pruned_loss=0.1149, over 19735.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.2894, pruned_loss=0.06663, over 3827396.60 frames. ], batch size: 63, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 16:58:58,163 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 16:59:30,207 INFO [train.py:903] (3/4) Epoch 20, batch 3850, loss[loss=0.1759, simple_loss=0.2589, pruned_loss=0.04645, over 19480.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2892, pruned_loss=0.06642, over 3809587.65 frames. ], batch size: 49, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 17:00:00,653 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5536, 2.1799, 1.7738, 1.4573, 2.0545, 1.4270, 1.5266, 1.9570], + device='cuda:3'), covar=tensor([0.0923, 0.0758, 0.0982, 0.0835, 0.0509, 0.1237, 0.0646, 0.0427], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0311, 0.0332, 0.0258, 0.0245, 0.0334, 0.0289, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:00:20,971 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.481e+02 5.079e+02 6.219e+02 7.261e+02 1.808e+03, threshold=1.244e+03, percent-clipped=5.0 +2023-04-02 17:00:25,931 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0192, 2.0788, 1.9286, 1.7618, 1.6298, 1.8085, 1.1179, 1.4124], + device='cuda:3'), covar=tensor([0.0553, 0.0576, 0.0393, 0.0615, 0.0919, 0.0826, 0.1131, 0.0865], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0350, 0.0356, 0.0380, 0.0454, 0.0383, 0.0331, 0.0336], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:00:32,689 INFO [train.py:903] (3/4) Epoch 20, batch 3900, loss[loss=0.1902, simple_loss=0.268, pruned_loss=0.05617, over 19691.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2891, pruned_loss=0.06605, over 3814827.79 frames. ], batch size: 53, lr: 4.12e-03, grad_scale: 8.0 +2023-04-02 17:01:21,152 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133670.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:01:37,398 INFO [train.py:903] (3/4) Epoch 20, batch 3950, loss[loss=0.1965, simple_loss=0.2691, pruned_loss=0.06193, over 17238.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2891, pruned_loss=0.06586, over 3799374.22 frames. ], batch size: 38, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:01:42,243 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 17:02:02,383 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7850, 4.1789, 4.5068, 4.5085, 1.9312, 4.2445, 3.7258, 4.2501], + device='cuda:3'), covar=tensor([0.1530, 0.1471, 0.0566, 0.0611, 0.5709, 0.0978, 0.0622, 0.0981], + device='cuda:3'), in_proj_covar=tensor([0.0769, 0.0725, 0.0923, 0.0812, 0.0822, 0.0687, 0.0557, 0.0862], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 17:02:26,845 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.611e+02 5.167e+02 6.244e+02 7.613e+02 1.189e+03, threshold=1.249e+03, percent-clipped=0.0 +2023-04-02 17:02:38,778 INFO [train.py:903] (3/4) Epoch 20, batch 4000, loss[loss=0.2029, simple_loss=0.2906, pruned_loss=0.0576, over 19615.00 frames. ], tot_loss[loss=0.21, simple_loss=0.289, pruned_loss=0.06554, over 3802081.16 frames. ], batch size: 57, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:02:43,769 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=133736.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 17:02:47,222 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=133738.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:03:26,433 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 17:03:41,996 INFO [train.py:903] (3/4) Epoch 20, batch 4050, loss[loss=0.2279, simple_loss=0.3013, pruned_loss=0.07723, over 17441.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2887, pruned_loss=0.06556, over 3799342.36 frames. ], batch size: 101, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:04:30,941 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.559e+02 4.971e+02 6.413e+02 8.150e+02 1.897e+03, threshold=1.283e+03, percent-clipped=7.0 +2023-04-02 17:04:42,295 INFO [train.py:903] (3/4) Epoch 20, batch 4100, loss[loss=0.2156, simple_loss=0.3007, pruned_loss=0.06528, over 19693.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2888, pruned_loss=0.06575, over 3804254.39 frames. ], batch size: 59, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:04:47,153 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=133836.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:04:59,743 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6006, 1.4004, 1.4287, 1.9313, 1.5603, 1.7830, 1.9938, 1.5926], + device='cuda:3'), covar=tensor([0.0831, 0.0936, 0.0985, 0.0684, 0.0755, 0.0734, 0.0695, 0.0728], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0219, 0.0223, 0.0240, 0.0225, 0.0208, 0.0186, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 17:05:06,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133851.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 17:05:09,070 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133853.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:05:18,885 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 17:05:28,136 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3492, 3.9933, 2.5325, 3.5545, 1.2454, 3.8526, 3.8076, 3.8216], + device='cuda:3'), covar=tensor([0.0680, 0.1029, 0.2047, 0.0849, 0.3615, 0.0765, 0.0917, 0.1306], + device='cuda:3'), in_proj_covar=tensor([0.0491, 0.0400, 0.0483, 0.0341, 0.0400, 0.0424, 0.0416, 0.0449], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:05:45,979 INFO [train.py:903] (3/4) Epoch 20, batch 4150, loss[loss=0.2062, simple_loss=0.2907, pruned_loss=0.06088, over 19549.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2887, pruned_loss=0.06531, over 3820882.58 frames. ], batch size: 56, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:05:56,151 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.86 vs. limit=5.0 +2023-04-02 17:06:02,512 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0281, 1.1176, 1.3704, 1.3176, 2.4217, 1.0463, 2.2165, 2.9190], + device='cuda:3'), covar=tensor([0.0809, 0.3632, 0.3478, 0.2122, 0.1287, 0.2859, 0.1309, 0.0470], + device='cuda:3'), in_proj_covar=tensor([0.0399, 0.0357, 0.0377, 0.0339, 0.0368, 0.0347, 0.0370, 0.0394], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:06:35,676 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.109e+02 4.757e+02 5.877e+02 6.653e+02 1.329e+03, threshold=1.175e+03, percent-clipped=1.0 +2023-04-02 17:06:47,886 INFO [train.py:903] (3/4) Epoch 20, batch 4200, loss[loss=0.1875, simple_loss=0.2778, pruned_loss=0.04864, over 19611.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2894, pruned_loss=0.06576, over 3821853.95 frames. ], batch size: 57, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:06:51,418 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 17:07:11,273 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8710, 1.9409, 2.1818, 2.2728, 1.7349, 2.1844, 2.2418, 2.0399], + device='cuda:3'), covar=tensor([0.3789, 0.3280, 0.1662, 0.2147, 0.3470, 0.1921, 0.4304, 0.2972], + device='cuda:3'), in_proj_covar=tensor([0.0889, 0.0950, 0.0709, 0.0933, 0.0869, 0.0802, 0.0837, 0.0775], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 17:07:12,219 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133951.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:07:50,937 INFO [train.py:903] (3/4) Epoch 20, batch 4250, loss[loss=0.2242, simple_loss=0.3082, pruned_loss=0.07008, over 19605.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2893, pruned_loss=0.06525, over 3827523.75 frames. ], batch size: 57, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:07:58,102 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5806, 1.6943, 1.9575, 1.9703, 1.4629, 1.8578, 1.9997, 1.8040], + device='cuda:3'), covar=tensor([0.4080, 0.3533, 0.1793, 0.2228, 0.3662, 0.2056, 0.4649, 0.3243], + device='cuda:3'), in_proj_covar=tensor([0.0887, 0.0948, 0.0707, 0.0930, 0.0867, 0.0801, 0.0836, 0.0772], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 17:08:08,179 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 17:08:21,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 17:08:32,207 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=134014.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:08:33,434 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134015.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:08:41,177 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.039e+02 5.008e+02 5.790e+02 6.980e+02 1.679e+03, threshold=1.158e+03, percent-clipped=4.0 +2023-04-02 17:08:54,656 INFO [train.py:903] (3/4) Epoch 20, batch 4300, loss[loss=0.2645, simple_loss=0.3379, pruned_loss=0.09552, over 19558.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2896, pruned_loss=0.06538, over 3831441.25 frames. ], batch size: 61, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:09:48,888 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8956, 1.9749, 2.2329, 2.5203, 1.8889, 2.4515, 2.2818, 1.9895], + device='cuda:3'), covar=tensor([0.3889, 0.3491, 0.1738, 0.2141, 0.3575, 0.1852, 0.4438, 0.3183], + device='cuda:3'), in_proj_covar=tensor([0.0888, 0.0950, 0.0710, 0.0932, 0.0868, 0.0803, 0.0835, 0.0774], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 17:09:50,637 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 17:09:56,295 INFO [train.py:903] (3/4) Epoch 20, batch 4350, loss[loss=0.2259, simple_loss=0.3009, pruned_loss=0.07549, over 19773.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2895, pruned_loss=0.06546, over 3819044.11 frames. ], batch size: 54, lr: 4.11e-03, grad_scale: 4.0 +2023-04-02 17:10:00,763 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-04-02 17:10:30,290 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134107.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 17:10:32,501 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134109.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:10:48,203 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.336e+02 5.216e+02 6.315e+02 8.361e+02 2.012e+03, threshold=1.263e+03, percent-clipped=10.0 +2023-04-02 17:10:57,742 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134129.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:11:00,859 INFO [train.py:903] (3/4) Epoch 20, batch 4400, loss[loss=0.2266, simple_loss=0.3052, pruned_loss=0.074, over 19685.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2896, pruned_loss=0.06544, over 3818179.45 frames. ], batch size: 55, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:11:01,288 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134132.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 17:11:03,623 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134134.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:11:29,062 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 17:11:37,427 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 17:12:05,047 INFO [train.py:903] (3/4) Epoch 20, batch 4450, loss[loss=0.1937, simple_loss=0.2678, pruned_loss=0.05983, over 16053.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2877, pruned_loss=0.06461, over 3827980.11 frames. ], batch size: 35, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:12:36,963 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134207.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:12:56,173 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.329e+02 4.942e+02 5.989e+02 7.537e+02 1.405e+03, threshold=1.198e+03, percent-clipped=2.0 +2023-04-02 17:13:08,117 INFO [train.py:903] (3/4) Epoch 20, batch 4500, loss[loss=0.2126, simple_loss=0.2918, pruned_loss=0.06673, over 19595.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2871, pruned_loss=0.06441, over 3821304.48 frames. ], batch size: 57, lr: 4.11e-03, grad_scale: 8.0 +2023-04-02 17:13:08,554 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134232.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:14:10,218 INFO [train.py:903] (3/4) Epoch 20, batch 4550, loss[loss=0.1913, simple_loss=0.2765, pruned_loss=0.05303, over 19683.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2891, pruned_loss=0.06558, over 3824764.36 frames. ], batch size: 53, lr: 4.11e-03, grad_scale: 4.0 +2023-04-02 17:14:15,066 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2226, 2.0707, 2.0201, 1.8206, 1.6776, 1.7962, 0.5845, 1.1432], + device='cuda:3'), covar=tensor([0.0601, 0.0599, 0.0401, 0.0692, 0.1117, 0.0823, 0.1186, 0.0946], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0351, 0.0355, 0.0379, 0.0455, 0.0383, 0.0332, 0.0338], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:14:19,194 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 17:14:42,931 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 17:15:02,249 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134322.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:15:04,308 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.805e+02 4.924e+02 5.886e+02 8.898e+02 2.816e+03, threshold=1.177e+03, percent-clipped=9.0 +2023-04-02 17:15:09,063 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134327.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:15:14,419 INFO [train.py:903] (3/4) Epoch 20, batch 4600, loss[loss=0.2644, simple_loss=0.3368, pruned_loss=0.09602, over 13765.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2888, pruned_loss=0.06546, over 3816918.57 frames. ], batch size: 136, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:15:30,248 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-02 17:15:47,469 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=134359.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:15:58,905 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4908, 1.1056, 1.3561, 1.1424, 2.1067, 0.9206, 2.1018, 2.3567], + device='cuda:3'), covar=tensor([0.0936, 0.3097, 0.3004, 0.1884, 0.1188, 0.2307, 0.1143, 0.0590], + device='cuda:3'), in_proj_covar=tensor([0.0403, 0.0360, 0.0381, 0.0342, 0.0372, 0.0347, 0.0373, 0.0396], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:16:13,235 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-04-02 17:16:15,900 INFO [train.py:903] (3/4) Epoch 20, batch 4650, loss[loss=0.1753, simple_loss=0.2489, pruned_loss=0.05091, over 19032.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2894, pruned_loss=0.06593, over 3817022.00 frames. ], batch size: 42, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:16:20,699 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134385.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:16:32,930 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 17:16:44,641 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 17:16:52,769 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134410.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:17:06,585 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-04-02 17:17:09,007 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.247e+02 4.575e+02 5.939e+02 8.134e+02 1.295e+03, threshold=1.188e+03, percent-clipped=3.0 +2023-04-02 17:17:19,255 INFO [train.py:903] (3/4) Epoch 20, batch 4700, loss[loss=0.2461, simple_loss=0.3173, pruned_loss=0.08743, over 19614.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2887, pruned_loss=0.06519, over 3825114.38 frames. ], batch size: 61, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:17:40,728 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1937, 1.8171, 1.4585, 1.1792, 1.5886, 1.1820, 1.2000, 1.6460], + device='cuda:3'), covar=tensor([0.0707, 0.0770, 0.1007, 0.0810, 0.0504, 0.1243, 0.0605, 0.0387], + device='cuda:3'), in_proj_covar=tensor([0.0294, 0.0310, 0.0332, 0.0256, 0.0244, 0.0334, 0.0286, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:17:41,482 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 17:18:11,968 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134474.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:18:20,721 INFO [train.py:903] (3/4) Epoch 20, batch 4750, loss[loss=0.2909, simple_loss=0.3429, pruned_loss=0.1194, over 19589.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2888, pruned_loss=0.06544, over 3827990.16 frames. ], batch size: 61, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:19:08,617 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0472, 2.0163, 1.7272, 2.1723, 2.0275, 1.8301, 1.7776, 2.0060], + device='cuda:3'), covar=tensor([0.1016, 0.1323, 0.1343, 0.0943, 0.1151, 0.0520, 0.1261, 0.0667], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0354, 0.0306, 0.0249, 0.0297, 0.0248, 0.0303, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:19:08,976 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.83 vs. limit=5.0 +2023-04-02 17:19:14,951 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.087e+02 5.001e+02 5.955e+02 7.090e+02 1.974e+03, threshold=1.191e+03, percent-clipped=7.0 +2023-04-02 17:19:25,460 INFO [train.py:903] (3/4) Epoch 20, batch 4800, loss[loss=0.1772, simple_loss=0.2531, pruned_loss=0.05063, over 17369.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.287, pruned_loss=0.06441, over 3830237.94 frames. ], batch size: 38, lr: 4.10e-03, grad_scale: 8.0 +2023-04-02 17:20:12,437 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3047, 2.3425, 2.5400, 3.0155, 2.3098, 2.8644, 2.5846, 2.2984], + device='cuda:3'), covar=tensor([0.4010, 0.3670, 0.1798, 0.2350, 0.4124, 0.2110, 0.4443, 0.3164], + device='cuda:3'), in_proj_covar=tensor([0.0881, 0.0946, 0.0707, 0.0925, 0.0866, 0.0800, 0.0832, 0.0772], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 17:20:26,814 INFO [train.py:903] (3/4) Epoch 20, batch 4850, loss[loss=0.2732, simple_loss=0.3383, pruned_loss=0.1041, over 19294.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2885, pruned_loss=0.06501, over 3826755.59 frames. ], batch size: 66, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:20:49,974 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 17:21:13,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 17:21:18,735 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 17:21:18,763 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 17:21:21,956 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.496e+02 4.787e+02 5.710e+02 7.760e+02 1.554e+03, threshold=1.142e+03, percent-clipped=3.0 +2023-04-02 17:21:31,390 INFO [train.py:903] (3/4) Epoch 20, batch 4900, loss[loss=0.2071, simple_loss=0.2909, pruned_loss=0.0616, over 19432.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2868, pruned_loss=0.06414, over 3829212.74 frames. ], batch size: 70, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:21:31,391 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 17:21:51,446 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 17:22:03,290 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1534, 3.5314, 2.1866, 2.1531, 3.3190, 2.0041, 1.3335, 2.2107], + device='cuda:3'), covar=tensor([0.1416, 0.0688, 0.1033, 0.0887, 0.0455, 0.1132, 0.1083, 0.0718], + device='cuda:3'), in_proj_covar=tensor([0.0297, 0.0312, 0.0333, 0.0257, 0.0245, 0.0335, 0.0287, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:22:14,400 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=134666.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:22:21,072 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=134671.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:22:33,369 INFO [train.py:903] (3/4) Epoch 20, batch 4950, loss[loss=0.2316, simple_loss=0.3069, pruned_loss=0.07811, over 19659.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2871, pruned_loss=0.06432, over 3836896.95 frames. ], batch size: 58, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:22:49,211 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 17:23:15,498 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 17:23:28,332 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.953e+02 4.876e+02 5.783e+02 7.285e+02 1.244e+03, threshold=1.157e+03, percent-clipped=2.0 +2023-04-02 17:23:34,575 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134730.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:23:37,531 INFO [train.py:903] (3/4) Epoch 20, batch 5000, loss[loss=0.2618, simple_loss=0.3323, pruned_loss=0.09562, over 17514.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2873, pruned_loss=0.06436, over 3831715.81 frames. ], batch size: 101, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:23:45,565 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 17:23:53,692 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3669, 1.4766, 1.7887, 1.6408, 2.8086, 2.2308, 2.8918, 1.4495], + device='cuda:3'), covar=tensor([0.2451, 0.4278, 0.2628, 0.1940, 0.1424, 0.2189, 0.1524, 0.4020], + device='cuda:3'), in_proj_covar=tensor([0.0523, 0.0632, 0.0694, 0.0476, 0.0612, 0.0521, 0.0655, 0.0540], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:23:56,668 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 17:24:06,115 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134755.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:24:37,413 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134781.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:24:38,201 INFO [train.py:903] (3/4) Epoch 20, batch 5050, loss[loss=0.2228, simple_loss=0.3029, pruned_loss=0.07135, over 19325.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2886, pruned_loss=0.0652, over 3829246.86 frames. ], batch size: 70, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:24:44,242 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134786.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:25:13,074 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 17:25:31,563 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 4.915e+02 6.341e+02 8.226e+02 2.739e+03, threshold=1.268e+03, percent-clipped=9.0 +2023-04-02 17:25:41,324 INFO [train.py:903] (3/4) Epoch 20, batch 5100, loss[loss=0.236, simple_loss=0.3165, pruned_loss=0.07775, over 17531.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2886, pruned_loss=0.06516, over 3835837.43 frames. ], batch size: 101, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:25:50,468 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 17:25:53,843 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 17:25:58,156 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 17:26:02,200 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0374, 5.0685, 5.9264, 5.9572, 2.2212, 5.5731, 4.7521, 5.5429], + device='cuda:3'), covar=tensor([0.1707, 0.0906, 0.0535, 0.0545, 0.5506, 0.0697, 0.0587, 0.1160], + device='cuda:3'), in_proj_covar=tensor([0.0766, 0.0721, 0.0922, 0.0808, 0.0816, 0.0686, 0.0553, 0.0857], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 17:26:41,287 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134880.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:26:43,298 INFO [train.py:903] (3/4) Epoch 20, batch 5150, loss[loss=0.2092, simple_loss=0.2903, pruned_loss=0.06406, over 19793.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2896, pruned_loss=0.06541, over 3829305.12 frames. ], batch size: 56, lr: 4.10e-03, grad_scale: 4.0 +2023-04-02 17:26:47,036 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8758, 1.4336, 1.8689, 1.6883, 4.2719, 1.1901, 2.6857, 4.7288], + device='cuda:3'), covar=tensor([0.0504, 0.3188, 0.2909, 0.2096, 0.0778, 0.2859, 0.1427, 0.0204], + device='cuda:3'), in_proj_covar=tensor([0.0401, 0.0361, 0.0380, 0.0342, 0.0371, 0.0346, 0.0372, 0.0397], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:26:55,950 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 17:27:03,555 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 17:27:10,858 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 17:27:32,563 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 17:27:37,090 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.618e+02 4.964e+02 6.321e+02 8.056e+02 1.479e+03, threshold=1.264e+03, percent-clipped=3.0 +2023-04-02 17:27:46,171 INFO [train.py:903] (3/4) Epoch 20, batch 5200, loss[loss=0.2053, simple_loss=0.29, pruned_loss=0.06031, over 19582.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2889, pruned_loss=0.06503, over 3817613.19 frames. ], batch size: 52, lr: 4.10e-03, grad_scale: 8.0 +2023-04-02 17:28:00,824 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 17:28:46,951 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 17:28:49,189 INFO [train.py:903] (3/4) Epoch 20, batch 5250, loss[loss=0.2267, simple_loss=0.3075, pruned_loss=0.07298, over 19533.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2889, pruned_loss=0.06495, over 3829164.17 frames. ], batch size: 56, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:29:42,806 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.326e+02 4.845e+02 5.899e+02 8.450e+02 1.811e+03, threshold=1.180e+03, percent-clipped=4.0 +2023-04-02 17:29:51,989 INFO [train.py:903] (3/4) Epoch 20, batch 5300, loss[loss=0.2048, simple_loss=0.2901, pruned_loss=0.05978, over 19525.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2885, pruned_loss=0.06543, over 3818519.77 frames. ], batch size: 54, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:29:59,112 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135037.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:30:04,905 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135042.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:30:11,575 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 17:30:30,714 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135062.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:30:36,544 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135067.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:30:54,699 INFO [train.py:903] (3/4) Epoch 20, batch 5350, loss[loss=0.2337, simple_loss=0.3114, pruned_loss=0.07802, over 19647.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2891, pruned_loss=0.06541, over 3808143.99 frames. ], batch size: 55, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:31:29,776 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 17:31:48,516 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.364e+02 5.248e+02 6.375e+02 8.534e+02 1.946e+03, threshold=1.275e+03, percent-clipped=9.0 +2023-04-02 17:31:57,756 INFO [train.py:903] (3/4) Epoch 20, batch 5400, loss[loss=0.2125, simple_loss=0.2966, pruned_loss=0.06419, over 19681.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2895, pruned_loss=0.06544, over 3809798.74 frames. ], batch size: 59, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:31:59,403 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3692, 1.3890, 1.6561, 1.5813, 2.5559, 2.2543, 2.7687, 1.2165], + device='cuda:3'), covar=tensor([0.2555, 0.4380, 0.2789, 0.2037, 0.1691, 0.2096, 0.1478, 0.4397], + device='cuda:3'), in_proj_covar=tensor([0.0522, 0.0628, 0.0694, 0.0474, 0.0610, 0.0519, 0.0650, 0.0538], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:33:00,881 INFO [train.py:903] (3/4) Epoch 20, batch 5450, loss[loss=0.1781, simple_loss=0.2684, pruned_loss=0.04387, over 19626.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2901, pruned_loss=0.06533, over 3803339.48 frames. ], batch size: 57, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:33:09,042 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=135189.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:33:15,483 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6619, 2.5599, 2.2170, 2.6346, 2.3984, 2.2141, 2.2171, 2.5680], + device='cuda:3'), covar=tensor([0.0950, 0.1492, 0.1349, 0.1066, 0.1403, 0.0516, 0.1213, 0.0637], + device='cuda:3'), in_proj_covar=tensor([0.0263, 0.0351, 0.0305, 0.0248, 0.0297, 0.0247, 0.0303, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:33:52,715 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=135224.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:33:54,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.649e+02 5.255e+02 6.010e+02 7.558e+02 1.824e+03, threshold=1.202e+03, percent-clipped=3.0 +2023-04-02 17:34:03,096 INFO [train.py:903] (3/4) Epoch 20, batch 5500, loss[loss=0.1846, simple_loss=0.2637, pruned_loss=0.05277, over 19625.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2891, pruned_loss=0.06529, over 3799839.90 frames. ], batch size: 50, lr: 4.09e-03, grad_scale: 4.0 +2023-04-02 17:34:29,170 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 17:34:30,819 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2098, 1.2857, 1.2640, 1.0107, 1.1339, 1.0908, 0.0845, 0.3155], + device='cuda:3'), covar=tensor([0.0702, 0.0673, 0.0423, 0.0578, 0.1269, 0.0632, 0.1298, 0.1110], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0351, 0.0354, 0.0379, 0.0455, 0.0384, 0.0333, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:35:05,527 INFO [train.py:903] (3/4) Epoch 20, batch 5550, loss[loss=0.2196, simple_loss=0.3015, pruned_loss=0.06884, over 19683.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2895, pruned_loss=0.06558, over 3804833.92 frames. ], batch size: 55, lr: 4.09e-03, grad_scale: 4.0 +2023-04-02 17:35:13,932 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 17:35:20,846 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5584, 3.9770, 4.3606, 4.4055, 1.6077, 4.0771, 3.5784, 3.7904], + device='cuda:3'), covar=tensor([0.2169, 0.1319, 0.0906, 0.1083, 0.7561, 0.1624, 0.1035, 0.2033], + device='cuda:3'), in_proj_covar=tensor([0.0769, 0.0724, 0.0928, 0.0813, 0.0821, 0.0691, 0.0554, 0.0862], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 17:35:38,280 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2073, 1.2987, 1.2809, 1.0654, 1.1374, 1.1179, 0.0951, 0.3578], + device='cuda:3'), covar=tensor([0.0725, 0.0685, 0.0438, 0.0597, 0.1345, 0.0667, 0.1316, 0.1130], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0353, 0.0355, 0.0380, 0.0457, 0.0386, 0.0335, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:36:00,656 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.142e+02 4.884e+02 6.418e+02 8.039e+02 2.322e+03, threshold=1.284e+03, percent-clipped=8.0 +2023-04-02 17:36:04,009 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 17:36:07,361 INFO [train.py:903] (3/4) Epoch 20, batch 5600, loss[loss=0.1842, simple_loss=0.2651, pruned_loss=0.05163, over 19748.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2885, pruned_loss=0.06484, over 3809867.01 frames. ], batch size: 51, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:36:18,152 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=135339.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:36:31,084 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1633, 2.8002, 2.2806, 2.2167, 1.9865, 2.5471, 0.8282, 2.0096], + device='cuda:3'), covar=tensor([0.0657, 0.0617, 0.0660, 0.1146, 0.1163, 0.1113, 0.1486, 0.1088], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0354, 0.0355, 0.0381, 0.0458, 0.0388, 0.0335, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:37:11,260 INFO [train.py:903] (3/4) Epoch 20, batch 5650, loss[loss=0.1941, simple_loss=0.2822, pruned_loss=0.05298, over 19467.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2889, pruned_loss=0.06497, over 3806359.12 frames. ], batch size: 49, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:38:01,350 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 17:38:05,718 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.932e+02 4.892e+02 5.590e+02 6.975e+02 1.698e+03, threshold=1.118e+03, percent-clipped=3.0 +2023-04-02 17:38:12,541 INFO [train.py:903] (3/4) Epoch 20, batch 5700, loss[loss=0.2229, simple_loss=0.3034, pruned_loss=0.07117, over 18291.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2895, pruned_loss=0.06575, over 3810653.32 frames. ], batch size: 83, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:39:14,105 INFO [train.py:903] (3/4) Epoch 20, batch 5750, loss[loss=0.1794, simple_loss=0.2582, pruned_loss=0.05032, over 19736.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2899, pruned_loss=0.06597, over 3810744.92 frames. ], batch size: 46, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:39:17,223 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 17:39:25,448 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 17:39:31,256 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 17:40:10,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.984e+02 5.062e+02 5.931e+02 7.729e+02 1.708e+03, threshold=1.186e+03, percent-clipped=6.0 +2023-04-02 17:40:18,218 INFO [train.py:903] (3/4) Epoch 20, batch 5800, loss[loss=0.2104, simple_loss=0.2982, pruned_loss=0.06128, over 19673.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.289, pruned_loss=0.06554, over 3813300.55 frames. ], batch size: 58, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:40:19,532 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=135533.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:41:20,583 INFO [train.py:903] (3/4) Epoch 20, batch 5850, loss[loss=0.2412, simple_loss=0.3169, pruned_loss=0.0828, over 19712.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2899, pruned_loss=0.06586, over 3819179.11 frames. ], batch size: 63, lr: 4.09e-03, grad_scale: 8.0 +2023-04-02 17:41:36,370 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135595.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:42:08,943 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135620.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:42:11,758 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.68 vs. limit=2.0 +2023-04-02 17:42:15,863 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.765e+02 5.164e+02 6.346e+02 8.557e+02 2.066e+03, threshold=1.269e+03, percent-clipped=9.0 +2023-04-02 17:42:23,022 INFO [train.py:903] (3/4) Epoch 20, batch 5900, loss[loss=0.2127, simple_loss=0.2973, pruned_loss=0.06401, over 19335.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.2903, pruned_loss=0.0662, over 3812699.23 frames. ], batch size: 66, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:42:25,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 17:42:43,086 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=135648.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:42:43,155 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=135648.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:42:46,320 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 17:43:24,712 INFO [train.py:903] (3/4) Epoch 20, batch 5950, loss[loss=0.1921, simple_loss=0.278, pruned_loss=0.05311, over 19723.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2898, pruned_loss=0.06598, over 3831199.73 frames. ], batch size: 63, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:44:19,051 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.081e+02 5.126e+02 6.708e+02 1.004e+03 2.382e+03, threshold=1.342e+03, percent-clipped=11.0 +2023-04-02 17:44:27,252 INFO [train.py:903] (3/4) Epoch 20, batch 6000, loss[loss=0.2102, simple_loss=0.2899, pruned_loss=0.06519, over 19752.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.29, pruned_loss=0.0661, over 3828052.81 frames. ], batch size: 54, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:44:27,252 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 17:44:39,926 INFO [train.py:937] (3/4) Epoch 20, validation: loss=0.1697, simple_loss=0.2697, pruned_loss=0.0349, over 944034.00 frames. +2023-04-02 17:44:39,927 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 17:45:41,594 INFO [train.py:903] (3/4) Epoch 20, batch 6050, loss[loss=0.2629, simple_loss=0.3329, pruned_loss=0.09647, over 13689.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.2901, pruned_loss=0.06634, over 3826277.13 frames. ], batch size: 136, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:46:36,913 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.594e+02 4.873e+02 5.806e+02 7.519e+02 1.887e+03, threshold=1.161e+03, percent-clipped=2.0 +2023-04-02 17:46:43,930 INFO [train.py:903] (3/4) Epoch 20, batch 6100, loss[loss=0.2004, simple_loss=0.2851, pruned_loss=0.05791, over 18710.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2901, pruned_loss=0.06658, over 3818052.81 frames. ], batch size: 74, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:47:05,979 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4988, 1.5659, 1.8599, 1.7452, 2.6348, 2.3447, 2.7321, 1.1542], + device='cuda:3'), covar=tensor([0.2420, 0.4273, 0.2739, 0.1941, 0.1519, 0.2018, 0.1478, 0.4479], + device='cuda:3'), in_proj_covar=tensor([0.0525, 0.0634, 0.0697, 0.0476, 0.0615, 0.0522, 0.0655, 0.0541], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:47:46,908 INFO [train.py:903] (3/4) Epoch 20, batch 6150, loss[loss=0.2036, simple_loss=0.2776, pruned_loss=0.06482, over 19613.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2891, pruned_loss=0.06593, over 3824366.08 frames. ], batch size: 50, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:48:15,330 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135904.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:48:16,084 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 17:48:21,068 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0578, 1.2790, 1.7265, 1.2637, 2.8724, 3.7736, 3.5078, 3.9999], + device='cuda:3'), covar=tensor([0.1725, 0.3762, 0.3381, 0.2450, 0.0546, 0.0188, 0.0202, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0318, 0.0347, 0.0264, 0.0239, 0.0183, 0.0215, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 17:48:38,647 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7885, 4.2715, 4.4970, 4.4937, 1.6883, 4.2623, 3.6623, 4.2223], + device='cuda:3'), covar=tensor([0.1616, 0.0773, 0.0617, 0.0649, 0.5953, 0.0775, 0.0647, 0.1193], + device='cuda:3'), in_proj_covar=tensor([0.0768, 0.0723, 0.0927, 0.0809, 0.0820, 0.0685, 0.0557, 0.0859], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 17:48:41,891 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 4.701e+02 6.211e+02 7.118e+02 1.417e+03, threshold=1.242e+03, percent-clipped=3.0 +2023-04-02 17:48:43,239 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1236, 1.2551, 1.7876, 1.1290, 2.4978, 3.3149, 3.0341, 3.5697], + device='cuda:3'), covar=tensor([0.1635, 0.3839, 0.3222, 0.2592, 0.0582, 0.0220, 0.0232, 0.0267], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0319, 0.0348, 0.0264, 0.0239, 0.0183, 0.0215, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 17:48:45,696 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135929.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:48:49,676 INFO [train.py:903] (3/4) Epoch 20, batch 6200, loss[loss=0.2407, simple_loss=0.3249, pruned_loss=0.07827, over 18014.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2889, pruned_loss=0.06571, over 3825426.16 frames. ], batch size: 83, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:49:52,113 INFO [train.py:903] (3/4) Epoch 20, batch 6250, loss[loss=0.1807, simple_loss=0.2654, pruned_loss=0.04799, over 19853.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2889, pruned_loss=0.06516, over 3827883.63 frames. ], batch size: 52, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:50:04,695 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=135992.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:50:15,890 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136000.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:50:19,573 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4565, 1.5601, 1.7617, 1.6915, 2.6289, 2.3054, 2.7625, 1.1534], + device='cuda:3'), covar=tensor([0.2422, 0.4264, 0.2727, 0.1861, 0.1453, 0.2119, 0.1474, 0.4333], + device='cuda:3'), in_proj_covar=tensor([0.0527, 0.0636, 0.0699, 0.0477, 0.0617, 0.0525, 0.0658, 0.0543], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 17:50:24,785 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 17:50:30,750 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136012.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:50:48,357 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.357e+02 5.185e+02 6.758e+02 8.208e+02 2.074e+03, threshold=1.352e+03, percent-clipped=8.0 +2023-04-02 17:50:55,344 INFO [train.py:903] (3/4) Epoch 20, batch 6300, loss[loss=0.2377, simple_loss=0.3141, pruned_loss=0.08067, over 19538.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.289, pruned_loss=0.06572, over 3832467.84 frames. ], batch size: 56, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:50:56,869 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3184, 1.7916, 1.9723, 1.9059, 2.8778, 1.6082, 2.8308, 3.2415], + device='cuda:3'), covar=tensor([0.0765, 0.2763, 0.2714, 0.2038, 0.0998, 0.2502, 0.1789, 0.0476], + device='cuda:3'), in_proj_covar=tensor([0.0404, 0.0360, 0.0380, 0.0342, 0.0372, 0.0347, 0.0372, 0.0397], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:51:58,440 INFO [train.py:903] (3/4) Epoch 20, batch 6350, loss[loss=0.2199, simple_loss=0.2914, pruned_loss=0.07424, over 19858.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2888, pruned_loss=0.06517, over 3835313.45 frames. ], batch size: 52, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:52:30,875 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136107.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:52:53,613 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.408e+02 5.045e+02 5.932e+02 7.156e+02 1.987e+03, threshold=1.186e+03, percent-clipped=3.0 +2023-04-02 17:53:01,266 INFO [train.py:903] (3/4) Epoch 20, batch 6400, loss[loss=0.2488, simple_loss=0.3107, pruned_loss=0.09345, over 13687.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2892, pruned_loss=0.06524, over 3831825.78 frames. ], batch size: 135, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:53:36,363 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136160.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:54:02,563 INFO [train.py:903] (3/4) Epoch 20, batch 6450, loss[loss=0.233, simple_loss=0.3139, pruned_loss=0.07604, over 19671.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2905, pruned_loss=0.06592, over 3830918.28 frames. ], batch size: 55, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:54:51,320 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 17:54:58,994 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.358e+02 5.012e+02 6.099e+02 8.089e+02 3.011e+03, threshold=1.220e+03, percent-clipped=7.0 +2023-04-02 17:55:01,042 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-04-02 17:55:07,146 INFO [train.py:903] (3/4) Epoch 20, batch 6500, loss[loss=0.1912, simple_loss=0.267, pruned_loss=0.05768, over 19610.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2915, pruned_loss=0.06638, over 3812335.11 frames. ], batch size: 50, lr: 4.08e-03, grad_scale: 8.0 +2023-04-02 17:55:14,051 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 17:55:20,946 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0019, 2.0638, 2.2997, 2.5962, 1.9046, 2.4662, 2.3567, 2.1549], + device='cuda:3'), covar=tensor([0.4194, 0.3893, 0.1880, 0.2401, 0.4146, 0.2180, 0.4592, 0.3292], + device='cuda:3'), in_proj_covar=tensor([0.0878, 0.0942, 0.0700, 0.0919, 0.0862, 0.0793, 0.0826, 0.0768], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 17:55:41,824 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7334, 1.7252, 1.3016, 1.7495, 1.6667, 1.4074, 1.3459, 1.5766], + device='cuda:3'), covar=tensor([0.1219, 0.1461, 0.1902, 0.1117, 0.1364, 0.0982, 0.1866, 0.1002], + device='cuda:3'), in_proj_covar=tensor([0.0264, 0.0353, 0.0305, 0.0248, 0.0297, 0.0246, 0.0304, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 17:56:09,976 INFO [train.py:903] (3/4) Epoch 20, batch 6550, loss[loss=0.2433, simple_loss=0.3198, pruned_loss=0.08342, over 19364.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2913, pruned_loss=0.06646, over 3825776.32 frames. ], batch size: 70, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 17:56:48,277 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 17:57:06,361 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.205e+02 5.319e+02 6.594e+02 8.030e+02 1.579e+03, threshold=1.319e+03, percent-clipped=2.0 +2023-04-02 17:57:14,430 INFO [train.py:903] (3/4) Epoch 20, batch 6600, loss[loss=0.253, simple_loss=0.3269, pruned_loss=0.08952, over 19525.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2903, pruned_loss=0.0657, over 3816905.58 frames. ], batch size: 56, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 17:57:28,429 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=136344.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:57:41,111 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136353.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:57:45,650 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=136356.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:57:53,895 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136363.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:58:16,828 INFO [train.py:903] (3/4) Epoch 20, batch 6650, loss[loss=0.2108, simple_loss=0.2948, pruned_loss=0.06338, over 19260.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.2906, pruned_loss=0.06601, over 3818426.47 frames. ], batch size: 66, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 17:58:26,169 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136388.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 17:58:43,792 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-02 17:59:14,402 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.147e+02 4.763e+02 5.879e+02 7.861e+02 2.647e+03, threshold=1.176e+03, percent-clipped=6.0 +2023-04-02 17:59:22,071 INFO [train.py:903] (3/4) Epoch 20, batch 6700, loss[loss=0.239, simple_loss=0.3209, pruned_loss=0.0786, over 19666.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2914, pruned_loss=0.06626, over 3829714.96 frames. ], batch size: 55, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 17:59:55,295 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136459.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:00:08,534 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136471.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:00:20,885 INFO [train.py:903] (3/4) Epoch 20, batch 6750, loss[loss=0.1877, simple_loss=0.2715, pruned_loss=0.05198, over 19762.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2913, pruned_loss=0.06633, over 3827579.88 frames. ], batch size: 54, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 18:00:21,335 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6033, 1.6840, 1.9222, 1.8932, 1.4551, 1.8376, 1.9327, 1.7699], + device='cuda:3'), covar=tensor([0.4103, 0.3708, 0.1974, 0.2415, 0.3793, 0.2272, 0.5098, 0.3507], + device='cuda:3'), in_proj_covar=tensor([0.0882, 0.0947, 0.0704, 0.0924, 0.0865, 0.0797, 0.0830, 0.0773], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 18:00:45,723 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=136504.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:01:11,793 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.251e+02 5.547e+02 6.932e+02 9.039e+02 1.788e+03, threshold=1.386e+03, percent-clipped=9.0 +2023-04-02 18:01:19,066 INFO [train.py:903] (3/4) Epoch 20, batch 6800, loss[loss=0.2365, simple_loss=0.3179, pruned_loss=0.07756, over 19619.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2903, pruned_loss=0.06582, over 3820594.48 frames. ], batch size: 57, lr: 4.07e-03, grad_scale: 8.0 +2023-04-02 18:02:04,754 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 18:02:05,227 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 18:02:07,962 INFO [train.py:903] (3/4) Epoch 21, batch 0, loss[loss=0.2207, simple_loss=0.2916, pruned_loss=0.07493, over 19373.00 frames. ], tot_loss[loss=0.2207, simple_loss=0.2916, pruned_loss=0.07493, over 19373.00 frames. ], batch size: 47, lr: 3.97e-03, grad_scale: 8.0 +2023-04-02 18:02:07,962 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 18:02:18,720 INFO [train.py:937] (3/4) Epoch 21, validation: loss=0.1691, simple_loss=0.2696, pruned_loss=0.03427, over 944034.00 frames. +2023-04-02 18:02:18,721 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 18:02:30,979 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 18:03:20,704 INFO [train.py:903] (3/4) Epoch 21, batch 50, loss[loss=0.2446, simple_loss=0.3186, pruned_loss=0.08528, over 19514.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2851, pruned_loss=0.06417, over 860719.80 frames. ], batch size: 64, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:03:24,696 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-02 18:03:33,346 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136619.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:03:43,287 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.449e+02 4.947e+02 6.173e+02 6.953e+02 1.295e+03, threshold=1.235e+03, percent-clipped=0.0 +2023-04-02 18:03:54,607 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 18:04:22,847 INFO [train.py:903] (3/4) Epoch 21, batch 100, loss[loss=0.2164, simple_loss=0.2768, pruned_loss=0.07799, over 17310.00 frames. ], tot_loss[loss=0.206, simple_loss=0.284, pruned_loss=0.06402, over 1513240.45 frames. ], batch size: 38, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:04:25,191 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136661.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:04:34,145 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 18:04:42,067 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2790, 1.9995, 1.9298, 2.7683, 2.0990, 2.5494, 2.5389, 2.2296], + device='cuda:3'), covar=tensor([0.0749, 0.0863, 0.0962, 0.0799, 0.0828, 0.0690, 0.0861, 0.0646], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0226, 0.0242, 0.0226, 0.0212, 0.0187, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 18:05:00,465 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.65 vs. limit=2.0 +2023-04-02 18:05:09,051 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=136697.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:05:25,010 INFO [train.py:903] (3/4) Epoch 21, batch 150, loss[loss=0.2307, simple_loss=0.3099, pruned_loss=0.07573, over 18284.00 frames. ], tot_loss[loss=0.207, simple_loss=0.2856, pruned_loss=0.06425, over 2028293.10 frames. ], batch size: 83, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:05:31,976 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136715.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:05:45,589 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.079e+02 5.031e+02 5.954e+02 7.665e+02 1.668e+03, threshold=1.191e+03, percent-clipped=3.0 +2023-04-02 18:05:45,999 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136727.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:06:02,669 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136740.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:06:18,654 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136752.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:06:25,310 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 18:06:27,565 INFO [train.py:903] (3/4) Epoch 21, batch 200, loss[loss=0.207, simple_loss=0.2789, pruned_loss=0.0675, over 19586.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2871, pruned_loss=0.0654, over 2418401.97 frames. ], batch size: 52, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:07:29,898 INFO [train.py:903] (3/4) Epoch 21, batch 250, loss[loss=0.1994, simple_loss=0.2907, pruned_loss=0.05399, over 19769.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.288, pruned_loss=0.06575, over 2725638.67 frames. ], batch size: 54, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:07:32,551 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8808, 1.4433, 1.5536, 1.4856, 3.4190, 1.2314, 2.4481, 3.9381], + device='cuda:3'), covar=tensor([0.0446, 0.2760, 0.2719, 0.1968, 0.0703, 0.2487, 0.1308, 0.0206], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0361, 0.0380, 0.0344, 0.0371, 0.0348, 0.0372, 0.0397], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:07:32,600 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136812.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:07:47,509 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136824.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:07:52,057 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.077e+02 4.815e+02 6.209e+02 8.068e+02 1.278e+03, threshold=1.242e+03, percent-clipped=1.0 +2023-04-02 18:08:32,585 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 18:08:33,040 INFO [train.py:903] (3/4) Epoch 21, batch 300, loss[loss=0.2018, simple_loss=0.2703, pruned_loss=0.06665, over 19747.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2857, pruned_loss=0.06441, over 2975413.33 frames. ], batch size: 47, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:08:53,757 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136875.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:09:07,904 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.02 vs. limit=5.0 +2023-04-02 18:09:23,358 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136900.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:09:36,407 INFO [train.py:903] (3/4) Epoch 21, batch 350, loss[loss=0.1658, simple_loss=0.2457, pruned_loss=0.04295, over 15558.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2887, pruned_loss=0.06601, over 3162124.96 frames. ], batch size: 34, lr: 3.97e-03, grad_scale: 4.0 +2023-04-02 18:09:38,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 18:09:56,983 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.395e+02 4.835e+02 5.943e+02 7.281e+02 1.741e+03, threshold=1.189e+03, percent-clipped=3.0 +2023-04-02 18:10:38,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4533, 1.4139, 1.9850, 1.6286, 2.8870, 4.5934, 4.4549, 5.1051], + device='cuda:3'), covar=tensor([0.1553, 0.3993, 0.3451, 0.2338, 0.0682, 0.0219, 0.0176, 0.0135], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0319, 0.0347, 0.0264, 0.0240, 0.0183, 0.0215, 0.0251], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 18:10:39,349 INFO [train.py:903] (3/4) Epoch 21, batch 400, loss[loss=0.1892, simple_loss=0.2834, pruned_loss=0.04744, over 19793.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2888, pruned_loss=0.06568, over 3316068.77 frames. ], batch size: 56, lr: 3.97e-03, grad_scale: 8.0 +2023-04-02 18:11:35,383 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=137005.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:11:41,216 INFO [train.py:903] (3/4) Epoch 21, batch 450, loss[loss=0.1846, simple_loss=0.258, pruned_loss=0.05555, over 19773.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2889, pruned_loss=0.0658, over 3434266.72 frames. ], batch size: 47, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:12:03,757 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.868e+02 4.613e+02 5.973e+02 7.527e+02 1.521e+03, threshold=1.195e+03, percent-clipped=6.0 +2023-04-02 18:12:12,939 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 18:12:14,060 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 18:12:43,472 INFO [train.py:903] (3/4) Epoch 21, batch 500, loss[loss=0.2649, simple_loss=0.3291, pruned_loss=0.1003, over 19493.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2903, pruned_loss=0.06643, over 3522186.79 frames. ], batch size: 64, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:12:53,183 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=137068.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:13:25,170 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=137093.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:13:45,872 INFO [train.py:903] (3/4) Epoch 21, batch 550, loss[loss=0.176, simple_loss=0.2556, pruned_loss=0.04813, over 19488.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2893, pruned_loss=0.06528, over 3593465.97 frames. ], batch size: 49, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:13:59,997 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=137120.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:14:09,838 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.012e+02 5.365e+02 6.823e+02 8.347e+02 2.113e+03, threshold=1.365e+03, percent-clipped=7.0 +2023-04-02 18:14:48,995 INFO [train.py:903] (3/4) Epoch 21, batch 600, loss[loss=0.2462, simple_loss=0.3264, pruned_loss=0.08295, over 19721.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2894, pruned_loss=0.06513, over 3654151.05 frames. ], batch size: 63, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:14:57,068 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3288, 3.9179, 2.5702, 3.5017, 1.0115, 3.9310, 3.7562, 3.8742], + device='cuda:3'), covar=tensor([0.0713, 0.1179, 0.2048, 0.0881, 0.4132, 0.0694, 0.0906, 0.1249], + device='cuda:3'), in_proj_covar=tensor([0.0492, 0.0403, 0.0486, 0.0338, 0.0401, 0.0423, 0.0418, 0.0453], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:15:00,564 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=137168.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:15:29,957 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 18:15:53,985 INFO [train.py:903] (3/4) Epoch 21, batch 650, loss[loss=0.2231, simple_loss=0.3089, pruned_loss=0.06861, over 19513.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2904, pruned_loss=0.06552, over 3691316.01 frames. ], batch size: 64, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:16:12,098 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0146, 1.6370, 1.9748, 1.7343, 4.5398, 1.1941, 2.5525, 4.9019], + device='cuda:3'), covar=tensor([0.0391, 0.2745, 0.2621, 0.1907, 0.0703, 0.2680, 0.1444, 0.0159], + device='cuda:3'), in_proj_covar=tensor([0.0405, 0.0362, 0.0382, 0.0343, 0.0371, 0.0347, 0.0373, 0.0398], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:16:16,604 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.985e+02 4.775e+02 5.968e+02 8.002e+02 1.696e+03, threshold=1.194e+03, percent-clipped=7.0 +2023-04-02 18:16:56,150 INFO [train.py:903] (3/4) Epoch 21, batch 700, loss[loss=0.2333, simple_loss=0.3051, pruned_loss=0.08075, over 12994.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.29, pruned_loss=0.06592, over 3703347.77 frames. ], batch size: 136, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:17:26,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=137283.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:17:42,550 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6780, 1.7192, 1.6305, 1.4165, 1.3622, 1.4242, 0.3039, 0.6301], + device='cuda:3'), covar=tensor([0.0680, 0.0604, 0.0386, 0.0609, 0.1188, 0.0718, 0.1332, 0.1102], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0352, 0.0355, 0.0380, 0.0458, 0.0388, 0.0335, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:17:54,697 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-02 18:18:00,855 INFO [train.py:903] (3/4) Epoch 21, batch 750, loss[loss=0.1868, simple_loss=0.2791, pruned_loss=0.04728, over 19715.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.29, pruned_loss=0.0657, over 3742912.96 frames. ], batch size: 59, lr: 3.96e-03, grad_scale: 4.0 +2023-04-02 18:18:22,897 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.306e+02 4.851e+02 6.236e+02 7.648e+02 2.101e+03, threshold=1.247e+03, percent-clipped=5.0 +2023-04-02 18:19:01,644 INFO [train.py:903] (3/4) Epoch 21, batch 800, loss[loss=0.2064, simple_loss=0.2841, pruned_loss=0.06437, over 19785.00 frames. ], tot_loss[loss=0.2116, simple_loss=0.2906, pruned_loss=0.06627, over 3746753.48 frames. ], batch size: 47, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:19:07,856 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 18:19:22,454 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=137376.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:19:34,038 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1911, 1.8041, 1.4392, 1.2077, 1.6230, 1.1802, 1.1758, 1.6744], + device='cuda:3'), covar=tensor([0.0743, 0.0778, 0.1070, 0.0816, 0.0534, 0.1264, 0.0593, 0.0426], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0313, 0.0336, 0.0260, 0.0245, 0.0337, 0.0292, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:19:53,527 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=137401.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:20:05,536 INFO [train.py:903] (3/4) Epoch 21, batch 850, loss[loss=0.1605, simple_loss=0.245, pruned_loss=0.03801, over 19328.00 frames. ], tot_loss[loss=0.2117, simple_loss=0.2908, pruned_loss=0.06632, over 3777872.04 frames. ], batch size: 44, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:20:27,117 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.313e+02 5.583e+02 6.647e+02 8.818e+02 2.027e+03, threshold=1.329e+03, percent-clipped=5.0 +2023-04-02 18:20:48,411 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 18:21:06,665 INFO [train.py:903] (3/4) Epoch 21, batch 900, loss[loss=0.2035, simple_loss=0.2874, pruned_loss=0.0598, over 19321.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.291, pruned_loss=0.06642, over 3797590.49 frames. ], batch size: 66, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:21:59,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 18:22:08,624 INFO [train.py:903] (3/4) Epoch 21, batch 950, loss[loss=0.1663, simple_loss=0.245, pruned_loss=0.04375, over 19754.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2913, pruned_loss=0.06628, over 3809387.13 frames. ], batch size: 46, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:22:31,547 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.353e+02 5.256e+02 6.264e+02 7.895e+02 1.664e+03, threshold=1.253e+03, percent-clipped=2.0 +2023-04-02 18:22:34,062 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4597, 1.5499, 1.8217, 1.7713, 2.3476, 2.0926, 2.3529, 1.0008], + device='cuda:3'), covar=tensor([0.2936, 0.4673, 0.3073, 0.2344, 0.1807, 0.2509, 0.1760, 0.5179], + device='cuda:3'), in_proj_covar=tensor([0.0530, 0.0639, 0.0703, 0.0482, 0.0619, 0.0527, 0.0661, 0.0544], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:22:45,797 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=137539.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:23:10,899 INFO [train.py:903] (3/4) Epoch 21, batch 1000, loss[loss=0.1965, simple_loss=0.2881, pruned_loss=0.05242, over 19542.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2889, pruned_loss=0.06503, over 3817460.98 frames. ], batch size: 56, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:23:15,683 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=137564.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:23:55,873 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 18:24:10,009 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.26 vs. limit=5.0 +2023-04-02 18:24:14,003 INFO [train.py:903] (3/4) Epoch 21, batch 1050, loss[loss=0.2083, simple_loss=0.2899, pruned_loss=0.06335, over 19777.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2888, pruned_loss=0.06511, over 3817996.99 frames. ], batch size: 54, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:24:35,335 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.928e+02 5.318e+02 6.486e+02 8.521e+02 3.216e+03, threshold=1.297e+03, percent-clipped=7.0 +2023-04-02 18:24:36,550 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 18:25:17,797 INFO [train.py:903] (3/4) Epoch 21, batch 1100, loss[loss=0.1954, simple_loss=0.2809, pruned_loss=0.05494, over 19589.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2895, pruned_loss=0.06583, over 3800512.35 frames. ], batch size: 61, lr: 3.96e-03, grad_scale: 8.0 +2023-04-02 18:25:36,463 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9037, 1.9576, 2.2355, 2.5234, 1.8889, 2.4762, 2.2950, 2.0412], + device='cuda:3'), covar=tensor([0.4038, 0.3823, 0.1833, 0.2283, 0.3970, 0.1959, 0.4504, 0.3271], + device='cuda:3'), in_proj_covar=tensor([0.0888, 0.0953, 0.0711, 0.0930, 0.0871, 0.0803, 0.0835, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 18:25:47,829 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6298, 1.7021, 1.8962, 2.0286, 1.5271, 1.9850, 1.9326, 1.8082], + device='cuda:3'), covar=tensor([0.4060, 0.3587, 0.1932, 0.2183, 0.3723, 0.1999, 0.4834, 0.3341], + device='cuda:3'), in_proj_covar=tensor([0.0887, 0.0952, 0.0710, 0.0929, 0.0870, 0.0802, 0.0834, 0.0776], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 18:26:19,697 INFO [train.py:903] (3/4) Epoch 21, batch 1150, loss[loss=0.1785, simple_loss=0.2546, pruned_loss=0.05117, over 19409.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2904, pruned_loss=0.06593, over 3802541.98 frames. ], batch size: 48, lr: 3.95e-03, grad_scale: 4.0 +2023-04-02 18:26:43,799 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.088e+02 5.127e+02 6.263e+02 7.580e+02 1.245e+03, threshold=1.253e+03, percent-clipped=0.0 +2023-04-02 18:27:22,178 INFO [train.py:903] (3/4) Epoch 21, batch 1200, loss[loss=0.1736, simple_loss=0.2437, pruned_loss=0.05176, over 19020.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2911, pruned_loss=0.0663, over 3821809.18 frames. ], batch size: 42, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:27:28,769 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-02 18:27:46,793 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 18:28:25,534 INFO [train.py:903] (3/4) Epoch 21, batch 1250, loss[loss=0.1739, simple_loss=0.2532, pruned_loss=0.04733, over 19761.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.2906, pruned_loss=0.06596, over 3822287.03 frames. ], batch size: 47, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:28:36,403 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7786, 1.3478, 1.5222, 1.4950, 3.1984, 1.1086, 2.3304, 3.7958], + device='cuda:3'), covar=tensor([0.0558, 0.3041, 0.2939, 0.2045, 0.0853, 0.2849, 0.1537, 0.0254], + device='cuda:3'), in_proj_covar=tensor([0.0403, 0.0360, 0.0379, 0.0343, 0.0370, 0.0345, 0.0373, 0.0397], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:28:48,828 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.413e+02 5.084e+02 5.991e+02 7.123e+02 1.423e+03, threshold=1.198e+03, percent-clipped=5.0 +2023-04-02 18:29:27,971 INFO [train.py:903] (3/4) Epoch 21, batch 1300, loss[loss=0.219, simple_loss=0.301, pruned_loss=0.06852, over 19288.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2901, pruned_loss=0.06566, over 3833142.62 frames. ], batch size: 66, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:29:36,706 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.29 vs. limit=5.0 +2023-04-02 18:30:30,624 INFO [train.py:903] (3/4) Epoch 21, batch 1350, loss[loss=0.2017, simple_loss=0.2877, pruned_loss=0.05791, over 19546.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2906, pruned_loss=0.06589, over 3835760.07 frames. ], batch size: 54, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:30:30,941 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=137910.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:30:34,446 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=137913.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:30:54,801 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.119e+02 5.098e+02 6.340e+02 8.452e+02 2.491e+03, threshold=1.268e+03, percent-clipped=6.0 +2023-04-02 18:31:33,358 INFO [train.py:903] (3/4) Epoch 21, batch 1400, loss[loss=0.1957, simple_loss=0.2615, pruned_loss=0.06496, over 19753.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2894, pruned_loss=0.06545, over 3840214.95 frames. ], batch size: 46, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:32:28,137 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 18:32:37,232 INFO [train.py:903] (3/4) Epoch 21, batch 1450, loss[loss=0.204, simple_loss=0.2946, pruned_loss=0.05668, over 19653.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.29, pruned_loss=0.06565, over 3841172.71 frames. ], batch size: 58, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:33:01,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.354e+02 4.676e+02 5.543e+02 6.978e+02 2.034e+03, threshold=1.109e+03, percent-clipped=2.0 +2023-04-02 18:33:31,756 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-02 18:33:39,041 INFO [train.py:903] (3/4) Epoch 21, batch 1500, loss[loss=0.2293, simple_loss=0.3146, pruned_loss=0.07207, over 19347.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2897, pruned_loss=0.06551, over 3844623.75 frames. ], batch size: 70, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:33:50,725 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0540, 1.9672, 1.8390, 2.1679, 1.9969, 1.8207, 1.8215, 2.0773], + device='cuda:3'), covar=tensor([0.0955, 0.1410, 0.1331, 0.0963, 0.1173, 0.0541, 0.1254, 0.0649], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0354, 0.0307, 0.0248, 0.0297, 0.0249, 0.0307, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:34:26,155 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1019, 2.0397, 1.6839, 2.0984, 1.9247, 1.7871, 1.7001, 1.9912], + device='cuda:3'), covar=tensor([0.1049, 0.1506, 0.1602, 0.1133, 0.1450, 0.0637, 0.1502, 0.0796], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0355, 0.0309, 0.0249, 0.0298, 0.0251, 0.0309, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:34:42,051 INFO [train.py:903] (3/4) Epoch 21, batch 1550, loss[loss=0.2338, simple_loss=0.3111, pruned_loss=0.07823, over 19671.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.29, pruned_loss=0.06569, over 3843234.24 frames. ], batch size: 58, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:35:05,479 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.935e+02 5.280e+02 6.228e+02 7.726e+02 2.313e+03, threshold=1.246e+03, percent-clipped=5.0 +2023-04-02 18:35:44,859 INFO [train.py:903] (3/4) Epoch 21, batch 1600, loss[loss=0.2258, simple_loss=0.3072, pruned_loss=0.07225, over 19574.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.289, pruned_loss=0.0651, over 3855959.44 frames. ], batch size: 61, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:36:07,166 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 18:36:33,727 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-02 18:36:48,270 INFO [train.py:903] (3/4) Epoch 21, batch 1650, loss[loss=0.1956, simple_loss=0.2764, pruned_loss=0.05742, over 19738.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2889, pruned_loss=0.06492, over 3835141.19 frames. ], batch size: 51, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:36:54,047 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7846, 1.7128, 1.3749, 1.7154, 1.8062, 1.4494, 1.4017, 1.6292], + device='cuda:3'), covar=tensor([0.1227, 0.1610, 0.1880, 0.1228, 0.1413, 0.0994, 0.1955, 0.1038], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0357, 0.0309, 0.0250, 0.0300, 0.0251, 0.0309, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:37:12,981 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.826e+02 4.591e+02 5.832e+02 7.172e+02 1.632e+03, threshold=1.166e+03, percent-clipped=1.0 +2023-04-02 18:37:43,213 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=138254.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:37:46,576 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=138257.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:37:49,813 INFO [train.py:903] (3/4) Epoch 21, batch 1700, loss[loss=0.218, simple_loss=0.3035, pruned_loss=0.06625, over 19388.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2887, pruned_loss=0.06489, over 3837242.31 frames. ], batch size: 70, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:38:25,357 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138288.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:38:28,571 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 18:38:53,713 INFO [train.py:903] (3/4) Epoch 21, batch 1750, loss[loss=0.1939, simple_loss=0.2746, pruned_loss=0.05664, over 19766.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2869, pruned_loss=0.0642, over 3837512.39 frames. ], batch size: 54, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:39:16,069 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.511e+02 4.819e+02 5.989e+02 8.047e+02 2.111e+03, threshold=1.198e+03, percent-clipped=8.0 +2023-04-02 18:39:30,289 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-02 18:39:30,904 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138340.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:39:55,296 INFO [train.py:903] (3/4) Epoch 21, batch 1800, loss[loss=0.1806, simple_loss=0.2606, pruned_loss=0.05033, over 19733.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2873, pruned_loss=0.06447, over 3838133.25 frames. ], batch size: 51, lr: 3.95e-03, grad_scale: 8.0 +2023-04-02 18:40:05,998 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138369.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:40:10,864 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138372.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:40:54,423 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 18:40:57,827 INFO [train.py:903] (3/4) Epoch 21, batch 1850, loss[loss=0.2277, simple_loss=0.2937, pruned_loss=0.08089, over 19468.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2873, pruned_loss=0.06438, over 3844838.93 frames. ], batch size: 49, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:41:22,807 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.428e+02 4.784e+02 5.677e+02 7.800e+02 1.333e+03, threshold=1.135e+03, percent-clipped=2.0 +2023-04-02 18:41:34,221 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 18:41:34,577 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2282, 2.0834, 2.1085, 1.8682, 1.6661, 1.8265, 0.5269, 1.2107], + device='cuda:3'), covar=tensor([0.0660, 0.0662, 0.0389, 0.0789, 0.1101, 0.0949, 0.1342, 0.1059], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0354, 0.0359, 0.0383, 0.0460, 0.0388, 0.0335, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:42:01,831 INFO [train.py:903] (3/4) Epoch 21, batch 1900, loss[loss=0.1894, simple_loss=0.2726, pruned_loss=0.0531, over 19851.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2871, pruned_loss=0.06428, over 3846004.56 frames. ], batch size: 52, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:42:18,810 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 18:42:23,575 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 18:42:48,430 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 18:43:04,472 INFO [train.py:903] (3/4) Epoch 21, batch 1950, loss[loss=0.2392, simple_loss=0.3236, pruned_loss=0.07741, over 19679.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2877, pruned_loss=0.06416, over 3849369.85 frames. ], batch size: 55, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:43:27,750 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.334e+02 4.783e+02 6.007e+02 7.405e+02 3.008e+03, threshold=1.201e+03, percent-clipped=4.0 +2023-04-02 18:43:49,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1343, 1.9965, 1.8978, 1.6728, 1.5693, 1.6855, 0.4752, 1.0010], + device='cuda:3'), covar=tensor([0.0656, 0.0610, 0.0425, 0.0722, 0.1116, 0.0809, 0.1276, 0.1057], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0354, 0.0360, 0.0383, 0.0461, 0.0389, 0.0336, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:44:06,909 INFO [train.py:903] (3/4) Epoch 21, batch 2000, loss[loss=0.2259, simple_loss=0.3193, pruned_loss=0.06618, over 19591.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2877, pruned_loss=0.06383, over 3856701.02 frames. ], batch size: 61, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:44:09,466 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.9185, 5.3767, 2.9272, 4.7648, 1.4221, 5.4445, 5.2675, 5.5185], + device='cuda:3'), covar=tensor([0.0407, 0.0814, 0.1989, 0.0736, 0.3529, 0.0569, 0.0753, 0.1103], + device='cuda:3'), in_proj_covar=tensor([0.0497, 0.0407, 0.0493, 0.0344, 0.0402, 0.0429, 0.0421, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:45:03,888 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 18:45:08,462 INFO [train.py:903] (3/4) Epoch 21, batch 2050, loss[loss=0.2105, simple_loss=0.2951, pruned_loss=0.06294, over 17236.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2875, pruned_loss=0.06404, over 3831641.90 frames. ], batch size: 101, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:45:22,115 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 18:45:23,304 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 18:45:28,193 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138625.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:45:32,705 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138628.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:45:33,534 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.887e+02 4.818e+02 6.151e+02 8.119e+02 1.355e+03, threshold=1.230e+03, percent-clipped=5.0 +2023-04-02 18:45:38,226 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=138632.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:45:45,206 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 18:46:00,112 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138650.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:46:03,382 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9689, 4.5158, 2.7588, 3.9848, 0.9623, 4.4442, 4.3246, 4.5134], + device='cuda:3'), covar=tensor([0.0506, 0.0855, 0.1898, 0.0764, 0.3888, 0.0599, 0.0778, 0.1017], + device='cuda:3'), in_proj_covar=tensor([0.0493, 0.0404, 0.0487, 0.0342, 0.0398, 0.0425, 0.0417, 0.0454], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:46:03,580 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138653.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:46:12,219 INFO [train.py:903] (3/4) Epoch 21, batch 2100, loss[loss=0.2069, simple_loss=0.286, pruned_loss=0.06384, over 19763.00 frames. ], tot_loss[loss=0.2089, simple_loss=0.2888, pruned_loss=0.06448, over 3819653.92 frames. ], batch size: 54, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:46:29,064 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1016, 2.8444, 2.2374, 2.1998, 2.0717, 2.4588, 1.0306, 2.0484], + device='cuda:3'), covar=tensor([0.0747, 0.0592, 0.0707, 0.1086, 0.1078, 0.1186, 0.1407, 0.1071], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0353, 0.0359, 0.0382, 0.0460, 0.0388, 0.0335, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:46:33,620 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5246, 1.0711, 1.2895, 1.2085, 2.1931, 0.9757, 1.9008, 2.4008], + device='cuda:3'), covar=tensor([0.0728, 0.2894, 0.2958, 0.1709, 0.0881, 0.2253, 0.1234, 0.0517], + device='cuda:3'), in_proj_covar=tensor([0.0408, 0.0365, 0.0384, 0.0346, 0.0373, 0.0349, 0.0376, 0.0401], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:46:39,345 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 18:46:41,766 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=138684.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:47:01,045 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138699.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:47:02,063 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 18:47:06,986 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138704.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:47:14,547 INFO [train.py:903] (3/4) Epoch 21, batch 2150, loss[loss=0.2222, simple_loss=0.3034, pruned_loss=0.0705, over 19741.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2893, pruned_loss=0.06492, over 3824947.66 frames. ], batch size: 51, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:47:21,597 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138715.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:47:26,591 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0159, 2.0945, 2.3795, 2.6875, 1.9621, 2.5658, 2.4581, 2.1557], + device='cuda:3'), covar=tensor([0.4251, 0.3831, 0.1817, 0.2386, 0.4337, 0.2153, 0.4485, 0.3165], + device='cuda:3'), in_proj_covar=tensor([0.0883, 0.0946, 0.0707, 0.0924, 0.0865, 0.0797, 0.0827, 0.0772], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 18:47:37,627 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.976e+02 5.033e+02 5.895e+02 7.227e+02 1.459e+03, threshold=1.179e+03, percent-clipped=3.0 +2023-04-02 18:48:00,981 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138747.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:48:18,144 INFO [train.py:903] (3/4) Epoch 21, batch 2200, loss[loss=0.2451, simple_loss=0.3279, pruned_loss=0.08114, over 19414.00 frames. ], tot_loss[loss=0.208, simple_loss=0.288, pruned_loss=0.06398, over 3841591.23 frames. ], batch size: 70, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:49:08,123 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138799.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:49:20,488 INFO [train.py:903] (3/4) Epoch 21, batch 2250, loss[loss=0.2323, simple_loss=0.319, pruned_loss=0.07277, over 19608.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2884, pruned_loss=0.06443, over 3841383.54 frames. ], batch size: 61, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:49:44,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.352e+02 5.230e+02 6.732e+02 8.271e+02 2.316e+03, threshold=1.346e+03, percent-clipped=7.0 +2023-04-02 18:50:23,834 INFO [train.py:903] (3/4) Epoch 21, batch 2300, loss[loss=0.187, simple_loss=0.2683, pruned_loss=0.05283, over 19613.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2884, pruned_loss=0.06484, over 3834279.31 frames. ], batch size: 50, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:50:38,383 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 18:51:27,176 INFO [train.py:903] (3/4) Epoch 21, batch 2350, loss[loss=0.228, simple_loss=0.2997, pruned_loss=0.07816, over 13187.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2885, pruned_loss=0.06479, over 3833126.06 frames. ], batch size: 136, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:51:48,954 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:51:50,726 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.927e+02 4.807e+02 6.646e+02 8.268e+02 1.737e+03, threshold=1.329e+03, percent-clipped=3.0 +2023-04-02 18:52:08,317 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 18:52:24,449 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 18:52:31,433 INFO [train.py:903] (3/4) Epoch 21, batch 2400, loss[loss=0.2242, simple_loss=0.2992, pruned_loss=0.07456, over 19773.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2883, pruned_loss=0.06484, over 3826748.05 frames. ], batch size: 46, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:53:25,710 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7160, 1.8418, 1.3799, 1.6755, 1.7260, 1.2856, 1.3503, 1.5062], + device='cuda:3'), covar=tensor([0.1325, 0.1622, 0.2022, 0.1270, 0.1516, 0.1215, 0.2086, 0.1205], + device='cuda:3'), in_proj_covar=tensor([0.0266, 0.0354, 0.0308, 0.0248, 0.0298, 0.0249, 0.0307, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:53:26,798 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139003.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:53:34,427 INFO [train.py:903] (3/4) Epoch 21, batch 2450, loss[loss=0.2003, simple_loss=0.2723, pruned_loss=0.06421, over 19095.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2886, pruned_loss=0.06481, over 3826076.43 frames. ], batch size: 42, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:53:59,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139028.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:53:59,946 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.106e+02 5.027e+02 6.383e+02 8.090e+02 1.476e+03, threshold=1.277e+03, percent-clipped=1.0 +2023-04-02 18:54:17,224 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139043.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:54:22,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139048.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:54:32,631 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139055.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:54:38,306 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139059.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:54:39,184 INFO [train.py:903] (3/4) Epoch 21, batch 2500, loss[loss=0.2478, simple_loss=0.3192, pruned_loss=0.08814, over 19660.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2896, pruned_loss=0.06532, over 3831436.80 frames. ], batch size: 58, lr: 3.94e-03, grad_scale: 8.0 +2023-04-02 18:55:05,390 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139080.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:55:35,400 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9781, 1.7339, 1.6520, 1.9101, 1.6036, 1.5955, 1.5389, 1.8068], + device='cuda:3'), covar=tensor([0.1022, 0.1443, 0.1401, 0.1116, 0.1373, 0.0573, 0.1514, 0.0739], + device='cuda:3'), in_proj_covar=tensor([0.0265, 0.0353, 0.0307, 0.0247, 0.0296, 0.0248, 0.0306, 0.0252], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 18:55:42,986 INFO [train.py:903] (3/4) Epoch 21, batch 2550, loss[loss=0.1874, simple_loss=0.2737, pruned_loss=0.05053, over 19676.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2884, pruned_loss=0.06499, over 3831341.98 frames. ], batch size: 53, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 18:56:06,423 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.656e+02 4.926e+02 6.170e+02 7.459e+02 2.294e+03, threshold=1.234e+03, percent-clipped=3.0 +2023-04-02 18:56:35,495 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 18:56:43,491 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139158.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:56:45,332 INFO [train.py:903] (3/4) Epoch 21, batch 2600, loss[loss=0.2135, simple_loss=0.2943, pruned_loss=0.06642, over 19711.00 frames. ], tot_loss[loss=0.2099, simple_loss=0.2893, pruned_loss=0.0652, over 3827934.65 frames. ], batch size: 51, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 18:56:49,215 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139163.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:57:02,184 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139174.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:57:47,166 INFO [train.py:903] (3/4) Epoch 21, batch 2650, loss[loss=0.2298, simple_loss=0.3018, pruned_loss=0.07889, over 19767.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2887, pruned_loss=0.06495, over 3828740.72 frames. ], batch size: 54, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 18:58:08,403 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 18:58:13,099 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.645e+02 4.962e+02 5.879e+02 7.846e+02 2.263e+03, threshold=1.176e+03, percent-clipped=6.0 +2023-04-02 18:58:29,464 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5853, 1.4567, 1.4942, 2.2050, 1.7237, 1.9073, 2.0754, 1.6673], + device='cuda:3'), covar=tensor([0.0823, 0.0968, 0.1021, 0.0766, 0.0822, 0.0771, 0.0817, 0.0731], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0225, 0.0241, 0.0225, 0.0211, 0.0187, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 18:58:39,736 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4363, 1.4996, 1.8424, 1.6685, 3.0220, 2.5613, 3.3708, 1.6202], + device='cuda:3'), covar=tensor([0.2542, 0.4408, 0.2809, 0.2034, 0.1662, 0.2039, 0.1492, 0.4164], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0638, 0.0705, 0.0482, 0.0619, 0.0525, 0.0659, 0.0546], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 18:58:45,357 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139256.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:58:49,502 INFO [train.py:903] (3/4) Epoch 21, batch 2700, loss[loss=0.2192, simple_loss=0.2992, pruned_loss=0.06962, over 19474.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2884, pruned_loss=0.06448, over 3835839.46 frames. ], batch size: 64, lr: 3.93e-03, grad_scale: 4.0 +2023-04-02 18:59:04,235 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139271.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:59:33,133 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139294.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:59:49,906 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139307.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 18:59:52,994 INFO [train.py:903] (3/4) Epoch 21, batch 2750, loss[loss=0.2087, simple_loss=0.2832, pruned_loss=0.06713, over 19404.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2899, pruned_loss=0.06533, over 3818453.69 frames. ], batch size: 48, lr: 3.93e-03, grad_scale: 4.0 +2023-04-02 19:00:18,058 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.927e+02 4.705e+02 5.796e+02 7.407e+02 1.811e+03, threshold=1.159e+03, percent-clipped=2.0 +2023-04-02 19:00:55,541 INFO [train.py:903] (3/4) Epoch 21, batch 2800, loss[loss=0.2193, simple_loss=0.2925, pruned_loss=0.07309, over 19660.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2887, pruned_loss=0.06465, over 3819208.61 frames. ], batch size: 53, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:01:28,012 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139386.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:01:58,454 INFO [train.py:903] (3/4) Epoch 21, batch 2850, loss[loss=0.212, simple_loss=0.292, pruned_loss=0.06601, over 19532.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2891, pruned_loss=0.06491, over 3818211.26 frames. ], batch size: 54, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:02:03,753 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139414.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:09,495 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139419.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:10,544 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9340, 5.0683, 5.7773, 5.7707, 2.1644, 5.4379, 4.7007, 5.3750], + device='cuda:3'), covar=tensor([0.1600, 0.0956, 0.0518, 0.0559, 0.5625, 0.0724, 0.0570, 0.1127], + device='cuda:3'), in_proj_covar=tensor([0.0778, 0.0730, 0.0942, 0.0820, 0.0826, 0.0693, 0.0568, 0.0870], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 19:02:23,967 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.945e+02 5.282e+02 6.184e+02 7.725e+02 1.552e+03, threshold=1.237e+03, percent-clipped=4.0 +2023-04-02 19:02:24,365 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139430.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:35,340 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139439.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:42,196 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139444.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:54,852 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139455.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:02:57,812 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 19:03:00,100 INFO [train.py:903] (3/4) Epoch 21, batch 2900, loss[loss=0.2003, simple_loss=0.2782, pruned_loss=0.0612, over 19607.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2907, pruned_loss=0.06588, over 3812300.90 frames. ], batch size: 50, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:04:04,508 INFO [train.py:903] (3/4) Epoch 21, batch 2950, loss[loss=0.2551, simple_loss=0.3234, pruned_loss=0.09338, over 13120.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2906, pruned_loss=0.06558, over 3817953.33 frames. ], batch size: 135, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:04:28,834 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.677e+02 4.684e+02 5.947e+02 7.442e+02 1.403e+03, threshold=1.189e+03, percent-clipped=5.0 +2023-04-02 19:05:06,918 INFO [train.py:903] (3/4) Epoch 21, batch 3000, loss[loss=0.1956, simple_loss=0.2793, pruned_loss=0.056, over 19735.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2899, pruned_loss=0.0654, over 3827991.06 frames. ], batch size: 51, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:05:06,918 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 19:05:20,610 INFO [train.py:937] (3/4) Epoch 21, validation: loss=0.1693, simple_loss=0.2693, pruned_loss=0.03465, over 944034.00 frames. +2023-04-02 19:05:20,613 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 19:05:24,360 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 19:05:28,248 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8757, 1.9830, 2.2427, 2.4956, 1.9075, 2.3929, 2.2714, 2.0465], + device='cuda:3'), covar=tensor([0.4121, 0.3510, 0.1720, 0.2215, 0.3722, 0.1990, 0.4411, 0.3172], + device='cuda:3'), in_proj_covar=tensor([0.0891, 0.0954, 0.0710, 0.0929, 0.0869, 0.0804, 0.0835, 0.0776], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 19:05:41,465 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-02 19:05:57,513 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139589.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:06:12,383 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139600.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:06:12,520 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139600.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:06:23,831 INFO [train.py:903] (3/4) Epoch 21, batch 3050, loss[loss=0.1935, simple_loss=0.2788, pruned_loss=0.05408, over 19660.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2907, pruned_loss=0.06589, over 3830609.37 frames. ], batch size: 55, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:06:48,011 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.185e+02 5.171e+02 6.119e+02 7.965e+02 1.649e+03, threshold=1.224e+03, percent-clipped=3.0 +2023-04-02 19:06:57,081 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139638.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:07:01,742 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 19:07:03,673 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139642.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:07:14,485 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139651.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:07:24,509 INFO [train.py:903] (3/4) Epoch 21, batch 3100, loss[loss=0.2177, simple_loss=0.3022, pruned_loss=0.06655, over 19700.00 frames. ], tot_loss[loss=0.2123, simple_loss=0.2916, pruned_loss=0.06652, over 3839652.42 frames. ], batch size: 59, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:07:34,139 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139667.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:08:05,845 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8232, 1.9090, 2.2151, 2.3958, 1.7841, 2.2790, 2.2039, 1.9515], + device='cuda:3'), covar=tensor([0.4188, 0.4100, 0.1933, 0.2425, 0.4276, 0.2283, 0.4836, 0.3515], + device='cuda:3'), in_proj_covar=tensor([0.0893, 0.0956, 0.0713, 0.0929, 0.0871, 0.0805, 0.0836, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 19:08:25,946 INFO [train.py:903] (3/4) Epoch 21, batch 3150, loss[loss=0.2012, simple_loss=0.2873, pruned_loss=0.05758, over 19533.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.292, pruned_loss=0.06691, over 3824909.57 frames. ], batch size: 56, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:08:29,912 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8599, 0.9246, 0.8830, 0.7713, 0.7497, 0.8377, 0.1178, 0.3023], + device='cuda:3'), covar=tensor([0.0500, 0.0490, 0.0334, 0.0430, 0.0842, 0.0481, 0.1045, 0.0787], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0354, 0.0359, 0.0383, 0.0459, 0.0388, 0.0334, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 19:08:32,271 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139715.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:08:51,321 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.442e+02 5.084e+02 6.643e+02 9.166e+02 2.493e+03, threshold=1.329e+03, percent-clipped=12.0 +2023-04-02 19:08:52,531 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 19:08:57,931 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-02 19:09:13,969 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-04-02 19:09:19,228 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139753.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:09:28,042 INFO [train.py:903] (3/4) Epoch 21, batch 3200, loss[loss=0.1675, simple_loss=0.2448, pruned_loss=0.0451, over 19351.00 frames. ], tot_loss[loss=0.213, simple_loss=0.292, pruned_loss=0.06706, over 3810360.77 frames. ], batch size: 47, lr: 3.93e-03, grad_scale: 8.0 +2023-04-02 19:09:36,915 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139766.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:10:01,973 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139787.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:10:27,849 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139807.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:10:30,928 INFO [train.py:903] (3/4) Epoch 21, batch 3250, loss[loss=0.2556, simple_loss=0.3288, pruned_loss=0.09115, over 12919.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2917, pruned_loss=0.06698, over 3808063.28 frames. ], batch size: 138, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:10:46,075 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139822.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:10:48,411 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139824.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:10:55,283 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.059e+02 5.030e+02 6.558e+02 8.674e+02 2.471e+03, threshold=1.312e+03, percent-clipped=9.0 +2023-04-02 19:11:32,335 INFO [train.py:903] (3/4) Epoch 21, batch 3300, loss[loss=0.2451, simple_loss=0.3326, pruned_loss=0.07878, over 17592.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2917, pruned_loss=0.06656, over 3812833.62 frames. ], batch size: 101, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:11:35,823 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 19:11:43,202 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139868.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:12:34,957 INFO [train.py:903] (3/4) Epoch 21, batch 3350, loss[loss=0.2194, simple_loss=0.2997, pruned_loss=0.06954, over 19525.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2923, pruned_loss=0.06668, over 3815803.61 frames. ], batch size: 56, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:12:44,268 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2776, 1.2042, 1.6169, 1.0999, 2.3783, 3.3378, 3.0428, 3.4938], + device='cuda:3'), covar=tensor([0.1498, 0.3938, 0.3507, 0.2558, 0.0621, 0.0212, 0.0223, 0.0271], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0321, 0.0350, 0.0263, 0.0242, 0.0185, 0.0216, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 19:13:00,886 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.203e+02 4.734e+02 5.660e+02 7.256e+02 1.171e+03, threshold=1.132e+03, percent-clipped=0.0 +2023-04-02 19:13:04,675 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139933.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:13:17,290 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=139944.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:13:37,533 INFO [train.py:903] (3/4) Epoch 21, batch 3400, loss[loss=0.1684, simple_loss=0.2468, pruned_loss=0.04498, over 19362.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2921, pruned_loss=0.06687, over 3815032.41 frames. ], batch size: 47, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:13:42,168 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9150, 1.3121, 1.5207, 1.4532, 3.4537, 1.0728, 2.3293, 3.9049], + device='cuda:3'), covar=tensor([0.0472, 0.2945, 0.3000, 0.2044, 0.0727, 0.2656, 0.1366, 0.0233], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0363, 0.0382, 0.0343, 0.0369, 0.0346, 0.0373, 0.0398], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:13:52,367 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139971.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:14:21,017 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8083, 1.6156, 1.4187, 1.7723, 1.3947, 1.5152, 1.4110, 1.6641], + device='cuda:3'), covar=tensor([0.1118, 0.1206, 0.1541, 0.1026, 0.1326, 0.0655, 0.1546, 0.0803], + device='cuda:3'), in_proj_covar=tensor([0.0268, 0.0356, 0.0309, 0.0249, 0.0299, 0.0250, 0.0309, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:14:22,097 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139996.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:14:41,270 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140009.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:14:42,107 INFO [train.py:903] (3/4) Epoch 21, batch 3450, loss[loss=0.2102, simple_loss=0.2918, pruned_loss=0.06428, over 17453.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2914, pruned_loss=0.06646, over 3805786.24 frames. ], batch size: 101, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:14:43,261 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 19:14:46,244 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.39 vs. limit=5.0 +2023-04-02 19:14:56,584 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140022.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:15:06,287 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.793e+02 5.068e+02 6.577e+02 8.644e+02 2.362e+03, threshold=1.315e+03, percent-clipped=9.0 +2023-04-02 19:15:11,108 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140034.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:15:27,238 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140047.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:15:28,360 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140048.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:15:41,624 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140059.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:15:42,383 INFO [train.py:903] (3/4) Epoch 21, batch 3500, loss[loss=0.2572, simple_loss=0.3214, pruned_loss=0.09651, over 12802.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2919, pruned_loss=0.06645, over 3796480.91 frames. ], batch size: 137, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:16:45,563 INFO [train.py:903] (3/4) Epoch 21, batch 3550, loss[loss=0.2244, simple_loss=0.3081, pruned_loss=0.07034, over 19671.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2917, pruned_loss=0.06653, over 3803274.89 frames. ], batch size: 59, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:16:59,827 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140121.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:17:08,454 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140128.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:17:10,317 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.090e+02 5.095e+02 6.549e+02 8.089e+02 2.006e+03, threshold=1.310e+03, percent-clipped=2.0 +2023-04-02 19:17:12,586 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140131.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:17:34,926 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9357, 4.9839, 5.6754, 5.6674, 1.9211, 5.3609, 4.5184, 5.3439], + device='cuda:3'), covar=tensor([0.1586, 0.0910, 0.0545, 0.0590, 0.6026, 0.0670, 0.0581, 0.1125], + device='cuda:3'), in_proj_covar=tensor([0.0774, 0.0732, 0.0935, 0.0822, 0.0824, 0.0694, 0.0567, 0.0870], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 19:17:36,901 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140151.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:17:36,974 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3170, 3.0057, 2.1921, 2.6550, 0.7178, 2.9787, 2.8774, 2.9383], + device='cuda:3'), covar=tensor([0.1097, 0.1412, 0.2090, 0.1136, 0.3906, 0.0985, 0.1135, 0.1389], + device='cuda:3'), in_proj_covar=tensor([0.0501, 0.0412, 0.0493, 0.0348, 0.0404, 0.0431, 0.0425, 0.0462], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:17:47,035 INFO [train.py:903] (3/4) Epoch 21, batch 3600, loss[loss=0.2388, simple_loss=0.3189, pruned_loss=0.07933, over 19671.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2918, pruned_loss=0.06649, over 3818021.28 frames. ], batch size: 58, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:17:56,437 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140166.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:17:56,723 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3069, 2.0332, 1.5629, 1.3944, 1.8673, 1.2905, 1.3380, 1.7763], + device='cuda:3'), covar=tensor([0.0859, 0.0749, 0.1040, 0.0779, 0.0485, 0.1182, 0.0589, 0.0427], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0314, 0.0337, 0.0262, 0.0246, 0.0336, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:17:58,699 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140168.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:18:34,548 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0438, 1.9431, 1.7849, 2.1344, 1.8733, 1.7570, 1.7050, 1.9863], + device='cuda:3'), covar=tensor([0.1054, 0.1486, 0.1335, 0.0975, 0.1288, 0.0551, 0.1396, 0.0681], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0360, 0.0311, 0.0251, 0.0301, 0.0252, 0.0310, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:18:51,571 INFO [train.py:903] (3/4) Epoch 21, batch 3650, loss[loss=0.2173, simple_loss=0.2865, pruned_loss=0.07404, over 19474.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2921, pruned_loss=0.06677, over 3820793.10 frames. ], batch size: 49, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:18:54,128 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140212.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:19:15,567 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.814e+02 5.123e+02 6.079e+02 7.474e+02 1.635e+03, threshold=1.216e+03, percent-clipped=1.0 +2023-04-02 19:19:37,431 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140246.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:19:54,178 INFO [train.py:903] (3/4) Epoch 21, batch 3700, loss[loss=0.1956, simple_loss=0.2709, pruned_loss=0.06017, over 19487.00 frames. ], tot_loss[loss=0.2126, simple_loss=0.2915, pruned_loss=0.06685, over 3822047.33 frames. ], batch size: 49, lr: 3.92e-03, grad_scale: 8.0 +2023-04-02 19:20:01,738 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140266.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:20:15,897 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140277.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:20:20,535 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140281.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:20:23,884 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140283.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:20:24,931 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2616, 3.8291, 3.9391, 3.9374, 1.5294, 3.7065, 3.2291, 3.6996], + device='cuda:3'), covar=tensor([0.1706, 0.0810, 0.0630, 0.0778, 0.5776, 0.0948, 0.0741, 0.1128], + device='cuda:3'), in_proj_covar=tensor([0.0765, 0.0723, 0.0923, 0.0811, 0.0812, 0.0687, 0.0560, 0.0859], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 19:20:49,130 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140304.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:20:55,397 INFO [train.py:903] (3/4) Epoch 21, batch 3750, loss[loss=0.1933, simple_loss=0.272, pruned_loss=0.05733, over 19500.00 frames. ], tot_loss[loss=0.2131, simple_loss=0.292, pruned_loss=0.06709, over 3816803.42 frames. ], batch size: 49, lr: 3.92e-03, grad_scale: 4.0 +2023-04-02 19:21:02,665 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140315.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:21:17,923 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140327.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:21:20,277 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140329.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:21:22,189 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.949e+02 4.554e+02 6.128e+02 7.118e+02 1.255e+03, threshold=1.226e+03, percent-clipped=1.0 +2023-04-02 19:21:33,873 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140340.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:21:40,806 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4144, 1.7543, 2.2056, 1.7576, 3.2626, 4.8444, 4.7293, 5.2214], + device='cuda:3'), covar=tensor([0.1614, 0.3475, 0.3070, 0.2241, 0.0545, 0.0172, 0.0156, 0.0162], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0320, 0.0348, 0.0262, 0.0242, 0.0184, 0.0215, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 19:21:57,852 INFO [train.py:903] (3/4) Epoch 21, batch 3800, loss[loss=0.2697, simple_loss=0.3305, pruned_loss=0.1044, over 13447.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2899, pruned_loss=0.06605, over 3806897.47 frames. ], batch size: 138, lr: 3.92e-03, grad_scale: 4.0 +2023-04-02 19:22:30,457 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 19:22:54,726 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140405.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:23:01,180 INFO [train.py:903] (3/4) Epoch 21, batch 3850, loss[loss=0.2429, simple_loss=0.3329, pruned_loss=0.07643, over 19305.00 frames. ], tot_loss[loss=0.2119, simple_loss=0.2907, pruned_loss=0.06652, over 3806562.83 frames. ], batch size: 66, lr: 3.92e-03, grad_scale: 4.0 +2023-04-02 19:23:02,817 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4265, 1.5184, 1.7476, 1.6874, 2.5640, 2.3116, 2.6781, 1.1018], + device='cuda:3'), covar=tensor([0.2342, 0.4022, 0.2583, 0.1789, 0.1454, 0.1928, 0.1432, 0.4162], + device='cuda:3'), in_proj_covar=tensor([0.0531, 0.0639, 0.0705, 0.0483, 0.0616, 0.0528, 0.0660, 0.0545], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 19:23:25,867 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.131e+02 5.035e+02 6.153e+02 7.922e+02 1.662e+03, threshold=1.231e+03, percent-clipped=3.0 +2023-04-02 19:23:33,491 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.93 vs. limit=5.0 +2023-04-02 19:24:03,190 INFO [train.py:903] (3/4) Epoch 21, batch 3900, loss[loss=0.2031, simple_loss=0.295, pruned_loss=0.05563, over 19681.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2893, pruned_loss=0.06545, over 3819282.43 frames. ], batch size: 59, lr: 3.92e-03, grad_scale: 4.0 +2023-04-02 19:24:04,950 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0651, 1.8387, 1.7168, 1.9828, 1.7019, 1.7026, 1.6797, 1.8836], + device='cuda:3'), covar=tensor([0.1037, 0.1378, 0.1490, 0.1033, 0.1397, 0.0609, 0.1472, 0.0777], + device='cuda:3'), in_proj_covar=tensor([0.0271, 0.0360, 0.0311, 0.0250, 0.0301, 0.0252, 0.0310, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:24:09,119 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140465.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:24:10,724 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7798, 1.8877, 2.1661, 2.2990, 1.6983, 2.2067, 2.1825, 2.0192], + device='cuda:3'), covar=tensor([0.4099, 0.3682, 0.1820, 0.2308, 0.3947, 0.2029, 0.4712, 0.3201], + device='cuda:3'), in_proj_covar=tensor([0.0891, 0.0953, 0.0714, 0.0928, 0.0871, 0.0808, 0.0836, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 19:24:17,262 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140472.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:24:55,683 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140502.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:01,348 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140507.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:04,340 INFO [train.py:903] (3/4) Epoch 21, batch 3950, loss[loss=0.2448, simple_loss=0.3091, pruned_loss=0.09031, over 13707.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.29, pruned_loss=0.06575, over 3827214.41 frames. ], batch size: 137, lr: 3.91e-03, grad_scale: 4.0 +2023-04-02 19:25:10,277 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 19:25:20,279 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140522.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:26,824 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140527.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:26,932 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140527.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:31,861 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.897e+02 4.763e+02 5.824e+02 7.544e+02 1.413e+03, threshold=1.165e+03, percent-clipped=2.0 +2023-04-02 19:25:40,198 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140537.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:42,492 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140539.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:25:51,855 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140547.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:07,106 INFO [train.py:903] (3/4) Epoch 21, batch 4000, loss[loss=0.2264, simple_loss=0.3072, pruned_loss=0.0728, over 19588.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2904, pruned_loss=0.06596, over 3819056.80 frames. ], batch size: 61, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:26:09,736 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140562.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:13,074 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140564.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:33,130 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140580.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:36,738 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:39,037 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4128, 1.3864, 1.5662, 1.6750, 3.0295, 1.2782, 2.3548, 3.4203], + device='cuda:3'), covar=tensor([0.0521, 0.2594, 0.2720, 0.1653, 0.0694, 0.2241, 0.1175, 0.0245], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0364, 0.0384, 0.0344, 0.0371, 0.0347, 0.0373, 0.0399], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:26:41,328 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140587.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:26:55,193 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 19:27:09,320 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140608.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:27:11,209 INFO [train.py:903] (3/4) Epoch 21, batch 4050, loss[loss=0.1845, simple_loss=0.2641, pruned_loss=0.05248, over 19385.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2889, pruned_loss=0.06528, over 3819120.67 frames. ], batch size: 48, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:27:25,055 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140621.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:27:36,528 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.169e+02 5.165e+02 6.710e+02 8.332e+02 1.443e+03, threshold=1.342e+03, percent-clipped=3.0 +2023-04-02 19:27:37,926 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0302, 1.9022, 1.7048, 2.0840, 1.8060, 1.7330, 1.6097, 1.8823], + device='cuda:3'), covar=tensor([0.1068, 0.1430, 0.1499, 0.0997, 0.1289, 0.0570, 0.1489, 0.0773], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0359, 0.0311, 0.0249, 0.0300, 0.0252, 0.0310, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:28:13,404 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-02 19:28:13,821 INFO [train.py:903] (3/4) Epoch 21, batch 4100, loss[loss=0.2204, simple_loss=0.2997, pruned_loss=0.07058, over 18081.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2876, pruned_loss=0.06464, over 3825620.81 frames. ], batch size: 83, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:28:52,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 19:29:08,488 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.04 vs. limit=5.0 +2023-04-02 19:29:15,738 INFO [train.py:903] (3/4) Epoch 21, batch 4150, loss[loss=0.2318, simple_loss=0.3081, pruned_loss=0.07781, over 19587.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2869, pruned_loss=0.06452, over 3822786.08 frames. ], batch size: 61, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:29:42,685 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.451e+02 5.318e+02 6.390e+02 7.957e+02 1.686e+03, threshold=1.278e+03, percent-clipped=3.0 +2023-04-02 19:29:49,995 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140736.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:29:55,919 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140741.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:30:05,109 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140749.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:30:17,540 INFO [train.py:903] (3/4) Epoch 21, batch 4200, loss[loss=0.2007, simple_loss=0.2897, pruned_loss=0.05587, over 19683.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.2876, pruned_loss=0.06464, over 3825629.55 frames. ], batch size: 55, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:30:24,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 19:31:21,340 INFO [train.py:903] (3/4) Epoch 21, batch 4250, loss[loss=0.187, simple_loss=0.2628, pruned_loss=0.05558, over 19357.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2883, pruned_loss=0.06532, over 3820716.96 frames. ], batch size: 47, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:31:40,147 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 19:31:42,805 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140827.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:31:47,368 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.155e+02 5.291e+02 6.401e+02 8.566e+02 1.506e+03, threshold=1.280e+03, percent-clipped=6.0 +2023-04-02 19:31:50,845 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 19:31:53,739 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140836.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:01,580 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140843.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:13,503 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140851.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:24,303 INFO [train.py:903] (3/4) Epoch 21, batch 4300, loss[loss=0.216, simple_loss=0.2866, pruned_loss=0.07273, over 19603.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2875, pruned_loss=0.06526, over 3816029.18 frames. ], batch size: 50, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:32:25,828 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140861.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:29,371 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140864.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:33,870 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140868.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:32:37,200 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=140871.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:33:18,593 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 19:33:24,440 INFO [train.py:903] (3/4) Epoch 21, batch 4350, loss[loss=0.2024, simple_loss=0.2877, pruned_loss=0.05859, over 19675.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2896, pruned_loss=0.06634, over 3816414.44 frames. ], batch size: 59, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:33:51,510 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.395e+02 4.814e+02 5.846e+02 6.905e+02 1.613e+03, threshold=1.169e+03, percent-clipped=2.0 +2023-04-02 19:33:57,271 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140935.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:34:25,324 INFO [train.py:903] (3/4) Epoch 21, batch 4400, loss[loss=0.2522, simple_loss=0.3226, pruned_loss=0.09092, over 13144.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2889, pruned_loss=0.06557, over 3813991.60 frames. ], batch size: 135, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:34:32,100 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140965.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:34:33,402 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140966.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:34:53,850 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 19:34:58,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140986.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:35:02,881 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 19:35:05,698 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140992.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:35:23,744 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 19:35:27,466 INFO [train.py:903] (3/4) Epoch 21, batch 4450, loss[loss=0.2499, simple_loss=0.3314, pruned_loss=0.08426, over 19603.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2889, pruned_loss=0.06557, over 3817637.02 frames. ], batch size: 61, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:35:37,895 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141017.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:35:54,629 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.268e+02 5.226e+02 6.450e+02 8.222e+02 2.218e+03, threshold=1.290e+03, percent-clipped=9.0 +2023-04-02 19:36:32,492 INFO [train.py:903] (3/4) Epoch 21, batch 4500, loss[loss=0.1882, simple_loss=0.2678, pruned_loss=0.05434, over 19688.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2892, pruned_loss=0.06578, over 3824636.71 frames. ], batch size: 53, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:36:38,415 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7961, 4.3810, 2.8168, 3.8957, 1.2073, 4.2817, 4.2149, 4.3263], + device='cuda:3'), covar=tensor([0.0562, 0.0817, 0.1833, 0.0766, 0.3661, 0.0648, 0.0895, 0.1187], + device='cuda:3'), in_proj_covar=tensor([0.0494, 0.0404, 0.0484, 0.0342, 0.0394, 0.0425, 0.0417, 0.0453], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:37:01,396 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141085.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:37:34,670 INFO [train.py:903] (3/4) Epoch 21, batch 4550, loss[loss=0.1633, simple_loss=0.2411, pruned_loss=0.04279, over 19729.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2878, pruned_loss=0.06507, over 3838981.27 frames. ], batch size: 45, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:37:37,502 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2340, 2.2686, 2.5346, 2.9149, 2.1104, 2.7848, 2.5206, 2.2096], + device='cuda:3'), covar=tensor([0.4118, 0.3922, 0.1766, 0.2840, 0.4522, 0.2208, 0.4780, 0.3392], + device='cuda:3'), in_proj_covar=tensor([0.0891, 0.0957, 0.0713, 0.0930, 0.0872, 0.0809, 0.0838, 0.0777], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 19:37:46,072 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 19:37:46,505 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141120.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:37:59,729 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.292e+02 5.246e+02 6.195e+02 7.478e+02 1.454e+03, threshold=1.239e+03, percent-clipped=4.0 +2023-04-02 19:38:10,835 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 19:38:18,684 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141145.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:38:35,232 INFO [train.py:903] (3/4) Epoch 21, batch 4600, loss[loss=0.225, simple_loss=0.3089, pruned_loss=0.07058, over 19332.00 frames. ], tot_loss[loss=0.209, simple_loss=0.288, pruned_loss=0.06499, over 3842152.37 frames. ], batch size: 66, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:38:36,685 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141161.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:38:49,788 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141171.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:39:25,219 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141200.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:39:36,149 INFO [train.py:903] (3/4) Epoch 21, batch 4650, loss[loss=0.1938, simple_loss=0.2777, pruned_loss=0.05494, over 19587.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2878, pruned_loss=0.06486, over 3837888.39 frames. ], batch size: 52, lr: 3.91e-03, grad_scale: 8.0 +2023-04-02 19:39:53,023 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141222.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:39:56,097 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 19:40:02,331 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.173e+02 4.513e+02 5.853e+02 7.712e+02 1.984e+03, threshold=1.171e+03, percent-clipped=6.0 +2023-04-02 19:40:05,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 19:40:15,514 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141242.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:40:21,273 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141247.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:40:37,900 INFO [train.py:903] (3/4) Epoch 21, batch 4700, loss[loss=0.1742, simple_loss=0.2459, pruned_loss=0.05123, over 19735.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2892, pruned_loss=0.06559, over 3834652.20 frames. ], batch size: 45, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:40:47,274 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141267.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:41:00,553 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 19:41:00,660 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141279.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:41:08,818 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141286.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:41:38,505 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141309.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:41:39,456 INFO [train.py:903] (3/4) Epoch 21, batch 4750, loss[loss=0.1727, simple_loss=0.2609, pruned_loss=0.04222, over 19759.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2887, pruned_loss=0.06535, over 3838436.30 frames. ], batch size: 47, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:42:03,068 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.546e+02 5.235e+02 6.308e+02 8.195e+02 2.468e+03, threshold=1.262e+03, percent-clipped=8.0 +2023-04-02 19:42:25,003 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:42:39,819 INFO [train.py:903] (3/4) Epoch 21, batch 4800, loss[loss=0.2436, simple_loss=0.3187, pruned_loss=0.08425, over 18796.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2892, pruned_loss=0.06548, over 3833020.31 frames. ], batch size: 74, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:43:23,106 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141394.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:43:24,864 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.05 vs. limit=5.0 +2023-04-02 19:43:41,808 INFO [train.py:903] (3/4) Epoch 21, batch 4850, loss[loss=0.2113, simple_loss=0.2957, pruned_loss=0.06347, over 19408.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2886, pruned_loss=0.06525, over 3827927.47 frames. ], batch size: 70, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:44:01,436 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141424.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:44:07,736 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 19:44:08,841 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.291e+02 5.192e+02 6.534e+02 8.985e+02 2.500e+03, threshold=1.307e+03, percent-clipped=12.0 +2023-04-02 19:44:13,611 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141435.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 19:44:28,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 19:44:33,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 19:44:34,458 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 19:44:37,275 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8774, 1.9802, 2.1505, 2.5019, 1.8548, 2.3881, 2.2101, 2.0306], + device='cuda:3'), covar=tensor([0.4096, 0.3713, 0.1881, 0.2271, 0.3879, 0.2051, 0.4535, 0.3169], + device='cuda:3'), in_proj_covar=tensor([0.0888, 0.0952, 0.0710, 0.0928, 0.0871, 0.0809, 0.0834, 0.0774], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 19:44:40,772 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141456.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:44:44,244 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141459.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:44:45,062 INFO [train.py:903] (3/4) Epoch 21, batch 4900, loss[loss=0.1923, simple_loss=0.2835, pruned_loss=0.05057, over 19627.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2885, pruned_loss=0.06513, over 3830460.01 frames. ], batch size: 57, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:44:47,115 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 19:44:59,109 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0668, 0.8948, 0.8726, 1.0652, 0.7889, 0.9419, 0.8613, 0.9854], + device='cuda:3'), covar=tensor([0.0872, 0.1083, 0.1119, 0.0702, 0.1016, 0.0501, 0.1104, 0.0661], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0362, 0.0313, 0.0252, 0.0302, 0.0254, 0.0311, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 19:45:05,765 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 19:45:10,383 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141481.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:45:19,553 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9732, 5.0022, 5.7038, 5.7228, 2.2381, 5.4134, 4.5303, 5.3211], + device='cuda:3'), covar=tensor([0.1628, 0.0899, 0.0614, 0.0614, 0.5602, 0.0728, 0.0689, 0.1264], + device='cuda:3'), in_proj_covar=tensor([0.0787, 0.0736, 0.0951, 0.0832, 0.0831, 0.0704, 0.0573, 0.0878], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 19:45:40,794 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141505.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:45:46,105 INFO [train.py:903] (3/4) Epoch 21, batch 4950, loss[loss=0.2167, simple_loss=0.3059, pruned_loss=0.06371, over 17953.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2887, pruned_loss=0.06504, over 3828575.33 frames. ], batch size: 83, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:46:03,684 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 19:46:10,475 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.127e+02 5.116e+02 6.031e+02 7.721e+02 1.403e+03, threshold=1.206e+03, percent-clipped=1.0 +2023-04-02 19:46:26,570 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141542.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:46:28,480 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 19:46:46,930 INFO [train.py:903] (3/4) Epoch 21, batch 5000, loss[loss=0.175, simple_loss=0.2592, pruned_loss=0.04541, over 19742.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2883, pruned_loss=0.0646, over 3824986.70 frames. ], batch size: 51, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:46:53,586 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 19:46:55,140 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141567.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:47:08,297 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 19:47:48,201 INFO [train.py:903] (3/4) Epoch 21, batch 5050, loss[loss=0.2381, simple_loss=0.3221, pruned_loss=0.07707, over 19345.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2875, pruned_loss=0.06408, over 3828590.68 frames. ], batch size: 66, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:48:03,144 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141620.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:48:04,285 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7418, 4.1566, 4.4311, 4.4384, 1.7132, 4.1619, 3.6226, 4.1392], + device='cuda:3'), covar=tensor([0.1605, 0.1091, 0.0626, 0.0658, 0.6154, 0.0961, 0.0701, 0.1191], + device='cuda:3'), in_proj_covar=tensor([0.0791, 0.0742, 0.0959, 0.0836, 0.0839, 0.0708, 0.0578, 0.0883], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 19:48:16,455 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.325e+02 4.754e+02 5.832e+02 6.826e+02 1.423e+03, threshold=1.166e+03, percent-clipped=2.0 +2023-04-02 19:48:25,906 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 19:48:39,270 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141650.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:48:53,396 INFO [train.py:903] (3/4) Epoch 21, batch 5100, loss[loss=0.2277, simple_loss=0.3078, pruned_loss=0.07376, over 18629.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2866, pruned_loss=0.06389, over 3827698.20 frames. ], batch size: 74, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:49:04,595 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 19:49:07,002 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 19:49:11,625 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 19:49:11,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141675.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:49:17,717 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141680.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:49:30,078 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141691.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:49:50,713 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141705.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:49:56,350 INFO [train.py:903] (3/4) Epoch 21, batch 5150, loss[loss=0.1937, simple_loss=0.2673, pruned_loss=0.06003, over 19411.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2879, pruned_loss=0.06478, over 3834048.43 frames. ], batch size: 48, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:50:09,418 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 19:50:20,928 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.369e+02 4.996e+02 6.424e+02 8.131e+02 1.633e+03, threshold=1.285e+03, percent-clipped=6.0 +2023-04-02 19:50:33,517 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3936, 1.4610, 1.7047, 1.6053, 2.5655, 2.2410, 2.7091, 1.1872], + device='cuda:3'), covar=tensor([0.2543, 0.4325, 0.2783, 0.2010, 0.1534, 0.2200, 0.1499, 0.4402], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0641, 0.0711, 0.0485, 0.0619, 0.0533, 0.0662, 0.0550], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 19:50:46,539 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 19:50:58,093 INFO [train.py:903] (3/4) Epoch 21, batch 5200, loss[loss=0.1834, simple_loss=0.2726, pruned_loss=0.04711, over 19833.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2894, pruned_loss=0.06558, over 3838770.41 frames. ], batch size: 52, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:51:14,053 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 19:51:23,420 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141779.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 19:51:51,427 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=141803.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:51:54,966 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141806.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:51:58,274 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 19:51:59,438 INFO [train.py:903] (3/4) Epoch 21, batch 5250, loss[loss=0.2046, simple_loss=0.2775, pruned_loss=0.06584, over 19712.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2903, pruned_loss=0.06605, over 3838727.47 frames. ], batch size: 51, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:52:28,865 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.823e+02 4.884e+02 6.185e+02 7.574e+02 1.457e+03, threshold=1.237e+03, percent-clipped=1.0 +2023-04-02 19:53:02,552 INFO [train.py:903] (3/4) Epoch 21, batch 5300, loss[loss=0.1874, simple_loss=0.2758, pruned_loss=0.04947, over 19586.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2897, pruned_loss=0.06579, over 3829805.69 frames. ], batch size: 52, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:53:23,415 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141876.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:53:24,178 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 19:53:36,407 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7169, 1.7279, 1.5952, 1.4280, 1.3509, 1.4720, 0.2550, 0.6587], + device='cuda:3'), covar=tensor([0.0639, 0.0576, 0.0408, 0.0618, 0.1203, 0.0728, 0.1249, 0.1078], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0352, 0.0355, 0.0380, 0.0458, 0.0384, 0.0333, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 19:53:45,666 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141894.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 19:53:55,961 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141901.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:54:06,888 INFO [train.py:903] (3/4) Epoch 21, batch 5350, loss[loss=0.2071, simple_loss=0.2853, pruned_loss=0.06447, over 19760.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2897, pruned_loss=0.06546, over 3828002.68 frames. ], batch size: 48, lr: 3.90e-03, grad_scale: 8.0 +2023-04-02 19:54:16,602 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141918.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:54:32,496 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.261e+02 4.997e+02 5.942e+02 7.628e+02 1.099e+03, threshold=1.188e+03, percent-clipped=0.0 +2023-04-02 19:54:35,114 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141934.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:54:43,734 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 19:54:51,527 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141946.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:55:07,469 INFO [train.py:903] (3/4) Epoch 21, batch 5400, loss[loss=0.1842, simple_loss=0.2658, pruned_loss=0.05129, over 19853.00 frames. ], tot_loss[loss=0.2109, simple_loss=0.2902, pruned_loss=0.06578, over 3824281.75 frames. ], batch size: 52, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 19:56:10,567 INFO [train.py:903] (3/4) Epoch 21, batch 5450, loss[loss=0.1851, simple_loss=0.2678, pruned_loss=0.05124, over 19646.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2896, pruned_loss=0.06561, over 3815374.41 frames. ], batch size: 53, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 19:56:39,854 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.702e+02 5.354e+02 6.392e+02 8.231e+02 1.329e+03, threshold=1.278e+03, percent-clipped=1.0 +2023-04-02 19:57:14,046 INFO [train.py:903] (3/4) Epoch 21, batch 5500, loss[loss=0.1799, simple_loss=0.2639, pruned_loss=0.04794, over 19587.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2891, pruned_loss=0.065, over 3831952.46 frames. ], batch size: 52, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 19:57:17,998 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=142062.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:57:42,014 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 19:57:48,307 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=142087.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:58:17,746 INFO [train.py:903] (3/4) Epoch 21, batch 5550, loss[loss=0.2435, simple_loss=0.3042, pruned_loss=0.09141, over 13668.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2884, pruned_loss=0.06495, over 3822632.56 frames. ], batch size: 137, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 19:58:26,202 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 19:58:43,632 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.159e+02 4.904e+02 5.778e+02 7.569e+02 2.193e+03, threshold=1.156e+03, percent-clipped=4.0 +2023-04-02 19:59:07,863 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=142150.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 19:59:15,748 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 19:59:19,357 INFO [train.py:903] (3/4) Epoch 21, batch 5600, loss[loss=0.2458, simple_loss=0.3215, pruned_loss=0.08501, over 19314.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2892, pruned_loss=0.06557, over 3820783.22 frames. ], batch size: 66, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 19:59:36,307 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=142174.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 19:59:37,386 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=142175.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:00:08,592 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=142199.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:00:20,415 INFO [train.py:903] (3/4) Epoch 21, batch 5650, loss[loss=0.2859, simple_loss=0.3572, pruned_loss=0.1073, over 19735.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.29, pruned_loss=0.06643, over 3827351.05 frames. ], batch size: 63, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:00:36,097 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5221, 2.1749, 1.5601, 1.2646, 2.1306, 1.1011, 1.3010, 1.9532], + device='cuda:3'), covar=tensor([0.1253, 0.0865, 0.1205, 0.1118, 0.0544, 0.1516, 0.0891, 0.0531], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0315, 0.0338, 0.0263, 0.0246, 0.0338, 0.0291, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:00:49,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.717e+02 4.896e+02 5.835e+02 7.915e+02 1.772e+03, threshold=1.167e+03, percent-clipped=3.0 +2023-04-02 20:01:10,304 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 20:01:22,637 INFO [train.py:903] (3/4) Epoch 21, batch 5700, loss[loss=0.2007, simple_loss=0.2847, pruned_loss=0.05835, over 19666.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2897, pruned_loss=0.06572, over 3828239.83 frames. ], batch size: 60, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:01:45,802 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=142278.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:01:59,759 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=142290.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:02:26,067 INFO [train.py:903] (3/4) Epoch 21, batch 5750, loss[loss=0.1755, simple_loss=0.2493, pruned_loss=0.05086, over 19790.00 frames. ], tot_loss[loss=0.2105, simple_loss=0.2897, pruned_loss=0.06567, over 3817075.14 frames. ], batch size: 49, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:02:28,359 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 20:02:36,341 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 20:02:41,026 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 20:02:51,479 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.086e+02 5.170e+02 6.700e+02 8.460e+02 1.665e+03, threshold=1.340e+03, percent-clipped=6.0 +2023-04-02 20:03:26,602 INFO [train.py:903] (3/4) Epoch 21, batch 5800, loss[loss=0.2044, simple_loss=0.2786, pruned_loss=0.06514, over 19427.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2892, pruned_loss=0.06541, over 3812362.01 frames. ], batch size: 48, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:03:59,394 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-04-02 20:04:08,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=142393.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:04:22,734 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=142405.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:04:28,041 INFO [train.py:903] (3/4) Epoch 21, batch 5850, loss[loss=0.2188, simple_loss=0.304, pruned_loss=0.0668, over 17254.00 frames. ], tot_loss[loss=0.2101, simple_loss=0.2895, pruned_loss=0.06531, over 3813801.56 frames. ], batch size: 101, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:04:37,184 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.71 vs. limit=5.0 +2023-04-02 20:04:57,529 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.411e+02 4.867e+02 6.014e+02 7.504e+02 1.855e+03, threshold=1.203e+03, percent-clipped=4.0 +2023-04-02 20:05:31,489 INFO [train.py:903] (3/4) Epoch 21, batch 5900, loss[loss=0.2369, simple_loss=0.313, pruned_loss=0.08035, over 13928.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2883, pruned_loss=0.06485, over 3816924.26 frames. ], batch size: 135, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:05:35,065 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 20:05:36,581 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1535, 1.2879, 1.5640, 1.4185, 2.7466, 1.1626, 2.1742, 3.1074], + device='cuda:3'), covar=tensor([0.0587, 0.2712, 0.2612, 0.1718, 0.0769, 0.2214, 0.1183, 0.0305], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0363, 0.0383, 0.0346, 0.0373, 0.0347, 0.0374, 0.0398], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:05:58,763 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 20:06:36,831 INFO [train.py:903] (3/4) Epoch 21, batch 5950, loss[loss=0.1913, simple_loss=0.271, pruned_loss=0.05584, over 17697.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2891, pruned_loss=0.06511, over 3806877.97 frames. ], batch size: 101, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:07:02,051 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.165e+02 4.752e+02 5.691e+02 7.550e+02 2.003e+03, threshold=1.138e+03, percent-clipped=5.0 +2023-04-02 20:07:37,097 INFO [train.py:903] (3/4) Epoch 21, batch 6000, loss[loss=0.211, simple_loss=0.2948, pruned_loss=0.06355, over 19027.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2897, pruned_loss=0.06576, over 3802065.74 frames. ], batch size: 75, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:07:37,098 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 20:07:50,389 INFO [train.py:937] (3/4) Epoch 21, validation: loss=0.1692, simple_loss=0.2693, pruned_loss=0.03459, over 944034.00 frames. +2023-04-02 20:07:50,390 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 20:08:27,748 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=142591.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:08:52,466 INFO [train.py:903] (3/4) Epoch 21, batch 6050, loss[loss=0.1947, simple_loss=0.2817, pruned_loss=0.05383, over 19607.00 frames. ], tot_loss[loss=0.2111, simple_loss=0.2901, pruned_loss=0.06604, over 3806005.46 frames. ], batch size: 57, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:09:03,420 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8030, 1.8927, 2.2020, 2.3525, 1.7218, 2.2903, 2.2729, 2.0513], + device='cuda:3'), covar=tensor([0.4226, 0.3925, 0.1850, 0.2263, 0.3960, 0.2076, 0.4684, 0.3331], + device='cuda:3'), in_proj_covar=tensor([0.0893, 0.0955, 0.0713, 0.0927, 0.0872, 0.0809, 0.0835, 0.0776], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 20:09:18,890 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.159e+02 4.812e+02 5.741e+02 7.692e+02 1.541e+03, threshold=1.148e+03, percent-clipped=3.0 +2023-04-02 20:09:41,455 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=142649.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:09:53,720 INFO [train.py:903] (3/4) Epoch 21, batch 6100, loss[loss=0.197, simple_loss=0.2832, pruned_loss=0.05543, over 19527.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2896, pruned_loss=0.06539, over 3812456.89 frames. ], batch size: 54, lr: 3.89e-03, grad_scale: 8.0 +2023-04-02 20:09:55,353 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=142661.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:10:12,469 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=142674.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:10:28,104 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=142686.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:10:55,895 INFO [train.py:903] (3/4) Epoch 21, batch 6150, loss[loss=0.2011, simple_loss=0.291, pruned_loss=0.05559, over 19661.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.29, pruned_loss=0.06581, over 3792935.34 frames. ], batch size: 58, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:11:05,637 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 20:11:25,007 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.643e+02 5.511e+02 6.918e+02 9.659e+02 2.206e+03, threshold=1.384e+03, percent-clipped=13.0 +2023-04-02 20:11:26,180 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 20:11:36,770 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=142742.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:11:59,801 INFO [train.py:903] (3/4) Epoch 21, batch 6200, loss[loss=0.189, simple_loss=0.2826, pruned_loss=0.0477, over 19668.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2888, pruned_loss=0.06521, over 3810501.78 frames. ], batch size: 53, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:12:03,977 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.49 vs. limit=2.0 +2023-04-02 20:12:15,045 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-04-02 20:13:02,737 INFO [train.py:903] (3/4) Epoch 21, batch 6250, loss[loss=0.2298, simple_loss=0.3001, pruned_loss=0.07971, over 19617.00 frames. ], tot_loss[loss=0.21, simple_loss=0.289, pruned_loss=0.06555, over 3800003.70 frames. ], batch size: 61, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:13:28,466 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.899e+02 5.305e+02 6.117e+02 7.859e+02 2.157e+03, threshold=1.223e+03, percent-clipped=2.0 +2023-04-02 20:13:30,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 20:14:05,184 INFO [train.py:903] (3/4) Epoch 21, batch 6300, loss[loss=0.1669, simple_loss=0.2461, pruned_loss=0.04382, over 19734.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2872, pruned_loss=0.06445, over 3814000.44 frames. ], batch size: 45, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:15:07,141 INFO [train.py:903] (3/4) Epoch 21, batch 6350, loss[loss=0.1943, simple_loss=0.2749, pruned_loss=0.05684, over 19488.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2866, pruned_loss=0.06362, over 3826028.94 frames. ], batch size: 49, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:15:36,285 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.232e+02 4.679e+02 5.550e+02 7.220e+02 1.923e+03, threshold=1.110e+03, percent-clipped=1.0 +2023-04-02 20:15:40,034 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=142935.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:15:52,189 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-02 20:16:11,259 INFO [train.py:903] (3/4) Epoch 21, batch 6400, loss[loss=0.1977, simple_loss=0.2951, pruned_loss=0.05019, over 19710.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2872, pruned_loss=0.06353, over 3834753.11 frames. ], batch size: 59, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:17:14,723 INFO [train.py:903] (3/4) Epoch 21, batch 6450, loss[loss=0.1958, simple_loss=0.2702, pruned_loss=0.06068, over 19363.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.287, pruned_loss=0.06341, over 3836400.17 frames. ], batch size: 47, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:17:40,516 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 4.995e+02 6.270e+02 8.275e+02 2.312e+03, threshold=1.254e+03, percent-clipped=6.0 +2023-04-02 20:18:01,479 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 20:18:04,907 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=143050.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:18:16,205 INFO [train.py:903] (3/4) Epoch 21, batch 6500, loss[loss=0.213, simple_loss=0.3017, pruned_loss=0.06216, over 17564.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2872, pruned_loss=0.06333, over 3850568.56 frames. ], batch size: 101, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:18:23,412 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 20:18:48,310 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.29 vs. limit=5.0 +2023-04-02 20:18:48,998 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=143086.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:19:16,578 INFO [train.py:903] (3/4) Epoch 21, batch 6550, loss[loss=0.2056, simple_loss=0.2921, pruned_loss=0.05954, over 17335.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2872, pruned_loss=0.06353, over 3831577.75 frames. ], batch size: 101, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:19:44,581 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.797e+02 5.073e+02 6.169e+02 7.633e+02 2.146e+03, threshold=1.234e+03, percent-clipped=4.0 +2023-04-02 20:19:46,053 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2634, 5.6391, 3.0738, 4.9224, 1.1336, 5.8763, 5.6847, 5.8877], + device='cuda:3'), covar=tensor([0.0369, 0.0815, 0.1810, 0.0692, 0.4135, 0.0479, 0.0767, 0.0816], + device='cuda:3'), in_proj_covar=tensor([0.0501, 0.0406, 0.0491, 0.0345, 0.0400, 0.0429, 0.0421, 0.0456], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:20:19,908 INFO [train.py:903] (3/4) Epoch 21, batch 6600, loss[loss=0.2055, simple_loss=0.2895, pruned_loss=0.06078, over 19662.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2874, pruned_loss=0.06369, over 3818693.63 frames. ], batch size: 60, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:21:11,156 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=143201.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:21:22,423 INFO [train.py:903] (3/4) Epoch 21, batch 6650, loss[loss=0.2407, simple_loss=0.3111, pruned_loss=0.08522, over 12905.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2863, pruned_loss=0.06316, over 3815794.45 frames. ], batch size: 135, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:21:46,321 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.72 vs. limit=5.0 +2023-04-02 20:21:47,859 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.522e+02 4.940e+02 5.672e+02 7.065e+02 1.538e+03, threshold=1.134e+03, percent-clipped=2.0 +2023-04-02 20:22:23,656 INFO [train.py:903] (3/4) Epoch 21, batch 6700, loss[loss=0.1801, simple_loss=0.2629, pruned_loss=0.04858, over 19846.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2863, pruned_loss=0.06316, over 3823547.14 frames. ], batch size: 52, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:23:19,534 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=143306.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:23:23,647 INFO [train.py:903] (3/4) Epoch 21, batch 6750, loss[loss=0.1696, simple_loss=0.2547, pruned_loss=0.0423, over 19370.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2865, pruned_loss=0.06352, over 3829422.01 frames. ], batch size: 48, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:23:48,083 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=143331.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:23:48,851 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.668e+02 4.995e+02 6.197e+02 7.772e+02 2.067e+03, threshold=1.239e+03, percent-clipped=6.0 +2023-04-02 20:24:20,249 INFO [train.py:903] (3/4) Epoch 21, batch 6800, loss[loss=0.1938, simple_loss=0.2768, pruned_loss=0.05538, over 19599.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2869, pruned_loss=0.06404, over 3835978.03 frames. ], batch size: 50, lr: 3.88e-03, grad_scale: 8.0 +2023-04-02 20:24:21,784 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9785, 4.4947, 2.8300, 3.9187, 1.0047, 4.4474, 4.3888, 4.4742], + device='cuda:3'), covar=tensor([0.0534, 0.0892, 0.1880, 0.0810, 0.3852, 0.0572, 0.0791, 0.0980], + device='cuda:3'), in_proj_covar=tensor([0.0494, 0.0401, 0.0486, 0.0341, 0.0395, 0.0423, 0.0418, 0.0451], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:25:05,895 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 20:25:07,119 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 20:25:09,792 INFO [train.py:903] (3/4) Epoch 22, batch 0, loss[loss=0.2183, simple_loss=0.2984, pruned_loss=0.06906, over 19545.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.2984, pruned_loss=0.06906, over 19545.00 frames. ], batch size: 56, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:25:09,792 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 20:25:18,055 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4130, 1.3568, 1.3656, 1.7504, 1.4190, 1.6211, 1.5628, 1.5060], + device='cuda:3'), covar=tensor([0.0759, 0.0925, 0.0909, 0.0631, 0.0907, 0.0776, 0.0883, 0.0699], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0228, 0.0240, 0.0227, 0.0213, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 20:25:20,456 INFO [train.py:937] (3/4) Epoch 22, validation: loss=0.1683, simple_loss=0.2691, pruned_loss=0.03373, over 944034.00 frames. +2023-04-02 20:25:20,457 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 20:25:29,870 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9749, 4.3434, 4.6219, 4.6789, 1.7021, 4.3626, 3.8277, 4.3127], + device='cuda:3'), covar=tensor([0.1532, 0.0890, 0.0594, 0.0606, 0.6347, 0.0921, 0.0668, 0.1163], + device='cuda:3'), in_proj_covar=tensor([0.0776, 0.0734, 0.0943, 0.0827, 0.0825, 0.0701, 0.0568, 0.0874], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 20:25:31,891 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 20:25:55,302 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=143418.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 20:26:14,249 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 4.908e+02 5.891e+02 8.006e+02 1.582e+03, threshold=1.178e+03, percent-clipped=4.0 +2023-04-02 20:26:21,014 INFO [train.py:903] (3/4) Epoch 22, batch 50, loss[loss=0.1696, simple_loss=0.2498, pruned_loss=0.04468, over 19378.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2882, pruned_loss=0.06305, over 866096.02 frames. ], batch size: 47, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:26:42,108 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=143457.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:26:47,952 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0667, 1.9049, 1.8051, 2.0585, 1.8128, 1.8031, 1.7182, 2.0007], + device='cuda:3'), covar=tensor([0.0936, 0.1381, 0.1331, 0.0979, 0.1222, 0.0498, 0.1350, 0.0630], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0354, 0.0308, 0.0247, 0.0296, 0.0248, 0.0307, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:26:53,919 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 20:27:13,769 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=143482.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:27:19,994 INFO [train.py:903] (3/4) Epoch 22, batch 100, loss[loss=0.2447, simple_loss=0.3167, pruned_loss=0.08638, over 18206.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.284, pruned_loss=0.06107, over 1539061.47 frames. ], batch size: 83, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:27:23,819 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=143491.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 20:27:31,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 20:28:12,239 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.507e+02 5.227e+02 6.391e+02 8.671e+02 1.540e+03, threshold=1.278e+03, percent-clipped=3.0 +2023-04-02 20:28:19,037 INFO [train.py:903] (3/4) Epoch 22, batch 150, loss[loss=0.2274, simple_loss=0.3051, pruned_loss=0.07489, over 19578.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2871, pruned_loss=0.06296, over 2051844.78 frames. ], batch size: 61, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:28:45,136 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5020, 2.2893, 1.7193, 1.4017, 2.0671, 1.3288, 1.3668, 1.9546], + device='cuda:3'), covar=tensor([0.1081, 0.0794, 0.0991, 0.0919, 0.0532, 0.1308, 0.0762, 0.0518], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0313, 0.0334, 0.0260, 0.0245, 0.0335, 0.0289, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:29:18,916 INFO [train.py:903] (3/4) Epoch 22, batch 200, loss[loss=0.2228, simple_loss=0.3015, pruned_loss=0.07203, over 18811.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2879, pruned_loss=0.06363, over 2448494.53 frames. ], batch size: 74, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:29:18,969 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 20:29:32,292 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-04-02 20:30:12,489 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.534e+02 5.144e+02 6.083e+02 7.742e+02 1.350e+03, threshold=1.217e+03, percent-clipped=1.0 +2023-04-02 20:30:20,919 INFO [train.py:903] (3/4) Epoch 22, batch 250, loss[loss=0.1881, simple_loss=0.2593, pruned_loss=0.05845, over 18259.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2871, pruned_loss=0.06333, over 2744845.76 frames. ], batch size: 40, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:30:33,624 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.66 vs. limit=5.0 +2023-04-02 20:31:20,923 INFO [train.py:903] (3/4) Epoch 22, batch 300, loss[loss=0.1964, simple_loss=0.2647, pruned_loss=0.06405, over 19750.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2867, pruned_loss=0.06357, over 2982588.43 frames. ], batch size: 47, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:32:15,079 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.028e+02 5.065e+02 6.247e+02 8.237e+02 1.383e+03, threshold=1.249e+03, percent-clipped=3.0 +2023-04-02 20:32:22,218 INFO [train.py:903] (3/4) Epoch 22, batch 350, loss[loss=0.1529, simple_loss=0.2366, pruned_loss=0.03457, over 19366.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2847, pruned_loss=0.06261, over 3174051.09 frames. ], batch size: 47, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:32:29,130 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 20:32:51,106 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=143762.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:33:01,966 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7503, 4.2310, 4.4355, 4.4503, 1.8628, 4.2084, 3.6452, 4.1854], + device='cuda:3'), covar=tensor([0.1533, 0.0958, 0.0588, 0.0644, 0.5496, 0.0883, 0.0627, 0.1016], + device='cuda:3'), in_proj_covar=tensor([0.0773, 0.0729, 0.0936, 0.0819, 0.0821, 0.0692, 0.0564, 0.0863], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 20:33:20,943 INFO [train.py:903] (3/4) Epoch 22, batch 400, loss[loss=0.2424, simple_loss=0.323, pruned_loss=0.08088, over 19671.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2865, pruned_loss=0.06324, over 3328559.65 frames. ], batch size: 59, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:34:10,388 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-02 20:34:15,304 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.174e+02 5.215e+02 6.557e+02 8.093e+02 2.351e+03, threshold=1.311e+03, percent-clipped=8.0 +2023-04-02 20:34:17,790 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=143835.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:34:20,913 INFO [train.py:903] (3/4) Epoch 22, batch 450, loss[loss=0.1814, simple_loss=0.2541, pruned_loss=0.05429, over 19766.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2875, pruned_loss=0.06411, over 3437628.93 frames. ], batch size: 45, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:34:37,220 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.52 vs. limit=5.0 +2023-04-02 20:34:57,893 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 20:34:58,985 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 20:35:06,270 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3963, 1.4934, 1.7933, 1.6701, 2.7015, 2.1729, 2.7715, 1.2083], + device='cuda:3'), covar=tensor([0.2414, 0.4054, 0.2515, 0.1849, 0.1359, 0.2150, 0.1389, 0.4249], + device='cuda:3'), in_proj_covar=tensor([0.0530, 0.0639, 0.0710, 0.0482, 0.0616, 0.0530, 0.0661, 0.0545], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 20:35:08,556 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=143877.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:35:22,926 INFO [train.py:903] (3/4) Epoch 22, batch 500, loss[loss=0.1697, simple_loss=0.2462, pruned_loss=0.04656, over 19136.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2879, pruned_loss=0.06413, over 3533046.14 frames. ], batch size: 42, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:36:17,499 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.793e+02 5.123e+02 6.359e+02 8.434e+02 1.804e+03, threshold=1.272e+03, percent-clipped=4.0 +2023-04-02 20:36:23,290 INFO [train.py:903] (3/4) Epoch 22, batch 550, loss[loss=0.224, simple_loss=0.2958, pruned_loss=0.07606, over 19769.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.288, pruned_loss=0.06443, over 3585547.73 frames. ], batch size: 54, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:36:37,276 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=143950.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:37:23,301 INFO [train.py:903] (3/4) Epoch 22, batch 600, loss[loss=0.2235, simple_loss=0.3023, pruned_loss=0.07238, over 18852.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2871, pruned_loss=0.06396, over 3644156.14 frames. ], batch size: 74, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:37:41,216 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5996, 1.4975, 1.5474, 1.9431, 1.5343, 1.8703, 1.9486, 1.7103], + device='cuda:3'), covar=tensor([0.0801, 0.0919, 0.0957, 0.0732, 0.0833, 0.0694, 0.0764, 0.0686], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0227, 0.0240, 0.0226, 0.0212, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 20:38:06,438 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-04-02 20:38:06,636 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 20:38:13,658 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9433, 2.7183, 2.3934, 2.8279, 2.5056, 2.3588, 2.0940, 2.6677], + device='cuda:3'), covar=tensor([0.0854, 0.1490, 0.1398, 0.1057, 0.1380, 0.0487, 0.1477, 0.0676], + device='cuda:3'), in_proj_covar=tensor([0.0267, 0.0354, 0.0310, 0.0249, 0.0298, 0.0249, 0.0307, 0.0256], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:38:17,797 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.931e+02 4.914e+02 6.190e+02 8.004e+02 1.732e+03, threshold=1.238e+03, percent-clipped=3.0 +2023-04-02 20:38:23,571 INFO [train.py:903] (3/4) Epoch 22, batch 650, loss[loss=0.1741, simple_loss=0.2525, pruned_loss=0.04786, over 19749.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2874, pruned_loss=0.06391, over 3684678.47 frames. ], batch size: 45, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:38:24,336 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.52 vs. limit=2.0 +2023-04-02 20:39:25,482 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0110, 4.3004, 4.6693, 4.7137, 2.0064, 4.4028, 3.8314, 4.3694], + device='cuda:3'), covar=tensor([0.1648, 0.1491, 0.0618, 0.0679, 0.5837, 0.0943, 0.0675, 0.1182], + device='cuda:3'), in_proj_covar=tensor([0.0774, 0.0733, 0.0937, 0.0821, 0.0826, 0.0695, 0.0565, 0.0867], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 20:39:26,362 INFO [train.py:903] (3/4) Epoch 22, batch 700, loss[loss=0.2235, simple_loss=0.3072, pruned_loss=0.06992, over 18246.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2869, pruned_loss=0.06314, over 3715472.42 frames. ], batch size: 83, lr: 3.78e-03, grad_scale: 8.0 +2023-04-02 20:40:19,657 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.965e+02 4.796e+02 6.107e+02 7.975e+02 1.533e+03, threshold=1.221e+03, percent-clipped=5.0 +2023-04-02 20:40:20,134 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=144133.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 20:40:26,351 INFO [train.py:903] (3/4) Epoch 22, batch 750, loss[loss=0.1931, simple_loss=0.2658, pruned_loss=0.06019, over 19424.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2871, pruned_loss=0.06361, over 3734242.95 frames. ], batch size: 48, lr: 3.77e-03, grad_scale: 8.0 +2023-04-02 20:40:49,266 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=144158.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:41:26,350 INFO [train.py:903] (3/4) Epoch 22, batch 800, loss[loss=0.22, simple_loss=0.2982, pruned_loss=0.07084, over 18764.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2862, pruned_loss=0.06314, over 3759808.84 frames. ], batch size: 74, lr: 3.77e-03, grad_scale: 8.0 +2023-04-02 20:41:44,756 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 20:41:48,073 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=144206.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 20:42:19,066 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=144231.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 20:42:20,959 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.163e+02 5.021e+02 6.351e+02 8.019e+02 1.751e+03, threshold=1.270e+03, percent-clipped=5.0 +2023-04-02 20:42:26,009 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4194, 1.4824, 1.7335, 1.6596, 2.3817, 2.0955, 2.4456, 1.0414], + device='cuda:3'), covar=tensor([0.2687, 0.4657, 0.2976, 0.2230, 0.1773, 0.2464, 0.1683, 0.5000], + device='cuda:3'), in_proj_covar=tensor([0.0530, 0.0642, 0.0711, 0.0482, 0.0619, 0.0529, 0.0661, 0.0547], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 20:42:26,670 INFO [train.py:903] (3/4) Epoch 22, batch 850, loss[loss=0.1972, simple_loss=0.28, pruned_loss=0.05722, over 19855.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2863, pruned_loss=0.06334, over 3784689.83 frames. ], batch size: 52, lr: 3.77e-03, grad_scale: 8.0 +2023-04-02 20:43:19,878 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 20:43:26,414 INFO [train.py:903] (3/4) Epoch 22, batch 900, loss[loss=0.2149, simple_loss=0.2907, pruned_loss=0.06956, over 19565.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2872, pruned_loss=0.06378, over 3782968.20 frames. ], batch size: 52, lr: 3.77e-03, grad_scale: 8.0 +2023-04-02 20:44:03,937 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8683, 1.8634, 1.3731, 1.8020, 1.8499, 1.4207, 1.5019, 1.6468], + device='cuda:3'), covar=tensor([0.1192, 0.1478, 0.1955, 0.1233, 0.1379, 0.0977, 0.1816, 0.1077], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0356, 0.0312, 0.0249, 0.0299, 0.0249, 0.0308, 0.0256], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:44:21,523 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.404e+02 5.111e+02 6.361e+02 7.451e+02 1.172e+03, threshold=1.272e+03, percent-clipped=0.0 +2023-04-02 20:44:26,103 INFO [train.py:903] (3/4) Epoch 22, batch 950, loss[loss=0.2002, simple_loss=0.2713, pruned_loss=0.06459, over 19711.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2869, pruned_loss=0.06372, over 3782103.69 frames. ], batch size: 51, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:44:30,638 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 20:45:22,972 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6613, 1.4965, 1.4969, 2.0694, 1.5617, 1.8791, 1.9918, 1.6632], + device='cuda:3'), covar=tensor([0.0842, 0.0955, 0.1025, 0.0759, 0.0870, 0.0818, 0.0824, 0.0719], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0221, 0.0225, 0.0238, 0.0225, 0.0211, 0.0186, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 20:45:27,344 INFO [train.py:903] (3/4) Epoch 22, batch 1000, loss[loss=0.1948, simple_loss=0.2804, pruned_loss=0.05464, over 19747.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2882, pruned_loss=0.06451, over 3767066.00 frames. ], batch size: 51, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:45:29,233 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-02 20:45:53,083 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-02 20:46:17,111 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=144429.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:46:17,952 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 20:46:22,205 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.871e+02 5.215e+02 6.539e+02 8.059e+02 1.779e+03, threshold=1.308e+03, percent-clipped=4.0 +2023-04-02 20:46:26,888 INFO [train.py:903] (3/4) Epoch 22, batch 1050, loss[loss=0.1859, simple_loss=0.2667, pruned_loss=0.05255, over 19728.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2877, pruned_loss=0.06434, over 3789937.50 frames. ], batch size: 51, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:47:00,704 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 20:47:22,075 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-04-02 20:47:26,634 INFO [train.py:903] (3/4) Epoch 22, batch 1100, loss[loss=0.2615, simple_loss=0.3261, pruned_loss=0.09848, over 13630.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2876, pruned_loss=0.06442, over 3794014.81 frames. ], batch size: 135, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:48:21,826 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.380e+02 5.103e+02 6.169e+02 7.547e+02 2.403e+03, threshold=1.234e+03, percent-clipped=2.0 +2023-04-02 20:48:27,962 INFO [train.py:903] (3/4) Epoch 22, batch 1150, loss[loss=0.1813, simple_loss=0.2621, pruned_loss=0.05021, over 19576.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2867, pruned_loss=0.06339, over 3814204.19 frames. ], batch size: 52, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:49:28,353 INFO [train.py:903] (3/4) Epoch 22, batch 1200, loss[loss=0.2009, simple_loss=0.2813, pruned_loss=0.06026, over 19847.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2869, pruned_loss=0.06394, over 3807501.74 frames. ], batch size: 52, lr: 3.77e-03, grad_scale: 8.0 +2023-04-02 20:50:00,834 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 20:50:12,312 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5270, 1.5584, 1.8425, 1.7724, 2.7273, 2.4349, 2.9003, 1.3666], + device='cuda:3'), covar=tensor([0.2333, 0.4186, 0.2534, 0.1800, 0.1505, 0.1936, 0.1418, 0.4149], + device='cuda:3'), in_proj_covar=tensor([0.0531, 0.0643, 0.0712, 0.0484, 0.0621, 0.0529, 0.0662, 0.0549], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 20:50:23,743 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.253e+02 4.877e+02 6.112e+02 7.869e+02 2.071e+03, threshold=1.222e+03, percent-clipped=4.0 +2023-04-02 20:50:27,109 INFO [train.py:903] (3/4) Epoch 22, batch 1250, loss[loss=0.24, simple_loss=0.3165, pruned_loss=0.08174, over 19361.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2877, pruned_loss=0.06438, over 3823701.27 frames. ], batch size: 66, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:51:16,185 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-04-02 20:51:28,180 INFO [train.py:903] (3/4) Epoch 22, batch 1300, loss[loss=0.1864, simple_loss=0.2713, pruned_loss=0.05078, over 19830.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2869, pruned_loss=0.06396, over 3811329.78 frames. ], batch size: 52, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:52:26,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.265e+02 4.679e+02 5.951e+02 8.140e+02 2.957e+03, threshold=1.190e+03, percent-clipped=7.0 +2023-04-02 20:52:30,257 INFO [train.py:903] (3/4) Epoch 22, batch 1350, loss[loss=0.2035, simple_loss=0.2765, pruned_loss=0.06529, over 19617.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2864, pruned_loss=0.06396, over 3816600.68 frames. ], batch size: 50, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:52:48,517 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0191, 1.2207, 1.7355, 1.1766, 2.6303, 3.5318, 3.1909, 3.7464], + device='cuda:3'), covar=tensor([0.1791, 0.3988, 0.3313, 0.2469, 0.0605, 0.0183, 0.0223, 0.0266], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0320, 0.0349, 0.0263, 0.0241, 0.0184, 0.0215, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 20:53:13,406 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=144773.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:53:31,334 INFO [train.py:903] (3/4) Epoch 22, batch 1400, loss[loss=0.2126, simple_loss=0.2766, pruned_loss=0.0743, over 19003.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2865, pruned_loss=0.0637, over 3797760.10 frames. ], batch size: 42, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:54:11,100 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-02 20:54:28,429 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.383e+02 4.808e+02 5.945e+02 7.380e+02 1.517e+03, threshold=1.189e+03, percent-clipped=2.0 +2023-04-02 20:54:29,518 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 20:54:31,678 INFO [train.py:903] (3/4) Epoch 22, batch 1450, loss[loss=0.21, simple_loss=0.2829, pruned_loss=0.06858, over 19604.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2874, pruned_loss=0.06391, over 3798924.62 frames. ], batch size: 50, lr: 3.77e-03, grad_scale: 4.0 +2023-04-02 20:54:47,899 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1107, 1.1147, 1.3480, 1.3223, 2.6942, 1.0448, 2.1475, 3.0109], + device='cuda:3'), covar=tensor([0.0616, 0.3032, 0.3183, 0.1923, 0.0806, 0.2549, 0.1235, 0.0376], + device='cuda:3'), in_proj_covar=tensor([0.0405, 0.0363, 0.0383, 0.0345, 0.0371, 0.0348, 0.0375, 0.0401], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:55:30,803 INFO [train.py:903] (3/4) Epoch 22, batch 1500, loss[loss=0.2432, simple_loss=0.3257, pruned_loss=0.08028, over 18804.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.288, pruned_loss=0.06387, over 3806789.12 frames. ], batch size: 74, lr: 3.76e-03, grad_scale: 4.0 +2023-04-02 20:55:31,173 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=144888.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:55:49,810 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7082, 1.7513, 1.6274, 1.3837, 1.3895, 1.4275, 0.2212, 0.6454], + device='cuda:3'), covar=tensor([0.0679, 0.0633, 0.0438, 0.0667, 0.1356, 0.0769, 0.1350, 0.1155], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0355, 0.0356, 0.0381, 0.0457, 0.0385, 0.0334, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 20:56:27,870 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.316e+02 4.957e+02 5.987e+02 8.036e+02 1.770e+03, threshold=1.197e+03, percent-clipped=5.0 +2023-04-02 20:56:31,408 INFO [train.py:903] (3/4) Epoch 22, batch 1550, loss[loss=0.2029, simple_loss=0.2868, pruned_loss=0.05952, over 19590.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2881, pruned_loss=0.06435, over 3782139.82 frames. ], batch size: 61, lr: 3.76e-03, grad_scale: 4.0 +2023-04-02 20:57:08,473 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0805, 5.1254, 5.8907, 5.9105, 1.9315, 5.5538, 4.7654, 5.5378], + device='cuda:3'), covar=tensor([0.1688, 0.0877, 0.0586, 0.0569, 0.6324, 0.0816, 0.0595, 0.1214], + device='cuda:3'), in_proj_covar=tensor([0.0781, 0.0740, 0.0944, 0.0830, 0.0830, 0.0702, 0.0568, 0.0876], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 20:57:21,881 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-02 20:57:30,413 INFO [train.py:903] (3/4) Epoch 22, batch 1600, loss[loss=0.2214, simple_loss=0.3027, pruned_loss=0.07001, over 19605.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2881, pruned_loss=0.06418, over 3802898.41 frames. ], batch size: 57, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 20:57:50,805 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 20:58:02,389 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145014.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:58:27,678 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.160e+02 4.884e+02 5.870e+02 7.908e+02 1.403e+03, threshold=1.174e+03, percent-clipped=3.0 +2023-04-02 20:58:31,145 INFO [train.py:903] (3/4) Epoch 22, batch 1650, loss[loss=0.1974, simple_loss=0.2909, pruned_loss=0.052, over 19633.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2891, pruned_loss=0.06451, over 3794448.88 frames. ], batch size: 57, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 20:58:39,518 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145045.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:58:48,578 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8336, 1.3120, 1.4527, 1.7071, 3.4055, 1.2709, 2.5098, 3.8575], + device='cuda:3'), covar=tensor([0.0504, 0.2846, 0.2983, 0.1869, 0.0750, 0.2512, 0.1310, 0.0258], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0364, 0.0386, 0.0348, 0.0373, 0.0349, 0.0377, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:58:49,782 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9654, 1.6186, 1.8197, 1.7475, 4.4951, 1.0814, 2.6723, 4.9429], + device='cuda:3'), covar=tensor([0.0427, 0.2781, 0.2919, 0.2007, 0.0760, 0.2798, 0.1379, 0.0161], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0364, 0.0386, 0.0348, 0.0373, 0.0349, 0.0377, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 20:59:27,094 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145084.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 20:59:31,377 INFO [train.py:903] (3/4) Epoch 22, batch 1700, loss[loss=0.2145, simple_loss=0.3014, pruned_loss=0.06374, over 19341.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2884, pruned_loss=0.06403, over 3803763.38 frames. ], batch size: 70, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:00:08,578 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 21:00:28,006 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.426e+02 4.845e+02 6.228e+02 7.810e+02 2.223e+03, threshold=1.246e+03, percent-clipped=2.0 +2023-04-02 21:00:33,034 INFO [train.py:903] (3/4) Epoch 22, batch 1750, loss[loss=0.3032, simple_loss=0.3595, pruned_loss=0.1234, over 13442.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2879, pruned_loss=0.06407, over 3800927.93 frames. ], batch size: 137, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:00:40,320 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=145144.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:00:50,966 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.7983, 5.2616, 3.0409, 4.5887, 1.2527, 5.3190, 5.1964, 5.3642], + device='cuda:3'), covar=tensor([0.0425, 0.0829, 0.1910, 0.0740, 0.3964, 0.0492, 0.0765, 0.1041], + device='cuda:3'), in_proj_covar=tensor([0.0503, 0.0414, 0.0495, 0.0346, 0.0406, 0.0434, 0.0428, 0.0462], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:01:09,191 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=145169.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:01:31,687 INFO [train.py:903] (3/4) Epoch 22, batch 1800, loss[loss=0.1906, simple_loss=0.2623, pruned_loss=0.05942, over 18695.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2882, pruned_loss=0.06416, over 3816203.40 frames. ], batch size: 41, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:01:56,775 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.40 vs. limit=5.0 +2023-04-02 21:01:57,538 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6469, 1.2379, 1.2607, 1.5060, 1.0808, 1.4289, 1.2807, 1.4778], + device='cuda:3'), covar=tensor([0.1129, 0.1252, 0.1591, 0.1044, 0.1365, 0.0603, 0.1506, 0.0857], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0359, 0.0314, 0.0251, 0.0302, 0.0252, 0.0311, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:02:27,947 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.845e+02 5.086e+02 5.993e+02 7.804e+02 1.410e+03, threshold=1.199e+03, percent-clipped=2.0 +2023-04-02 21:02:27,971 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 21:02:31,524 INFO [train.py:903] (3/4) Epoch 22, batch 1850, loss[loss=0.1949, simple_loss=0.2816, pruned_loss=0.0541, over 19660.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2873, pruned_loss=0.06422, over 3821717.05 frames. ], batch size: 55, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:03:04,102 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 21:03:30,818 INFO [train.py:903] (3/4) Epoch 22, batch 1900, loss[loss=0.1987, simple_loss=0.2847, pruned_loss=0.0564, over 19451.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2877, pruned_loss=0.06392, over 3828819.08 frames. ], batch size: 64, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:03:48,271 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 21:03:52,770 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 21:04:15,275 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 21:04:26,520 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.830e+02 5.286e+02 6.049e+02 6.874e+02 1.450e+03, threshold=1.210e+03, percent-clipped=2.0 +2023-04-02 21:04:30,789 INFO [train.py:903] (3/4) Epoch 22, batch 1950, loss[loss=0.1811, simple_loss=0.2562, pruned_loss=0.05304, over 19333.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2883, pruned_loss=0.06411, over 3815106.00 frames. ], batch size: 44, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:04:38,867 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-04-02 21:04:55,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=145358.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:05:31,551 INFO [train.py:903] (3/4) Epoch 22, batch 2000, loss[loss=0.1718, simple_loss=0.2589, pruned_loss=0.04233, over 19848.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2879, pruned_loss=0.06364, over 3806926.22 frames. ], batch size: 52, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:05:32,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=145389.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:05:33,000 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145389.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:06:19,686 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=145428.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:06:27,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.585e+02 4.801e+02 6.060e+02 7.916e+02 1.266e+03, threshold=1.212e+03, percent-clipped=1.0 +2023-04-02 21:06:27,635 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 21:06:30,885 INFO [train.py:903] (3/4) Epoch 22, batch 2050, loss[loss=0.1966, simple_loss=0.2682, pruned_loss=0.06245, over 19744.00 frames. ], tot_loss[loss=0.207, simple_loss=0.2873, pruned_loss=0.06333, over 3820909.59 frames. ], batch size: 46, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:06:46,540 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 21:06:46,569 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 21:07:06,389 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 21:07:13,320 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=145473.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:07:30,917 INFO [train.py:903] (3/4) Epoch 22, batch 2100, loss[loss=0.2107, simple_loss=0.2733, pruned_loss=0.07407, over 19725.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2875, pruned_loss=0.06355, over 3819295.27 frames. ], batch size: 46, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:07:51,410 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=145504.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:08:01,441 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 21:08:22,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 21:08:27,088 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.088e+02 4.926e+02 6.113e+02 7.931e+02 1.598e+03, threshold=1.223e+03, percent-clipped=5.0 +2023-04-02 21:08:30,650 INFO [train.py:903] (3/4) Epoch 22, batch 2150, loss[loss=0.1875, simple_loss=0.2636, pruned_loss=0.05567, over 19754.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2874, pruned_loss=0.06363, over 3835448.63 frames. ], batch size: 46, lr: 3.76e-03, grad_scale: 8.0 +2023-04-02 21:08:38,981 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=145543.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:09:32,291 INFO [train.py:903] (3/4) Epoch 22, batch 2200, loss[loss=0.2179, simple_loss=0.2827, pruned_loss=0.07653, over 19336.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2888, pruned_loss=0.06456, over 3819558.91 frames. ], batch size: 44, lr: 3.76e-03, grad_scale: 4.0 +2023-04-02 21:10:00,321 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145612.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:10:29,862 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.244e+02 5.056e+02 6.179e+02 8.064e+02 2.249e+03, threshold=1.236e+03, percent-clipped=3.0 +2023-04-02 21:10:32,066 INFO [train.py:903] (3/4) Epoch 22, batch 2250, loss[loss=0.2301, simple_loss=0.3085, pruned_loss=0.07584, over 19337.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2882, pruned_loss=0.06423, over 3827077.63 frames. ], batch size: 66, lr: 3.76e-03, grad_scale: 4.0 +2023-04-02 21:10:50,655 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3682, 1.1935, 1.2321, 1.8048, 1.3325, 1.5232, 1.4937, 1.4115], + device='cuda:3'), covar=tensor([0.1033, 0.1227, 0.1184, 0.0708, 0.0909, 0.0917, 0.0955, 0.0902], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0221, 0.0225, 0.0240, 0.0226, 0.0211, 0.0186, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 21:11:01,371 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145661.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:11:33,804 INFO [train.py:903] (3/4) Epoch 22, batch 2300, loss[loss=0.2556, simple_loss=0.329, pruned_loss=0.0911, over 19358.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2878, pruned_loss=0.06411, over 3831828.14 frames. ], batch size: 66, lr: 3.75e-03, grad_scale: 4.0 +2023-04-02 21:11:45,979 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 21:12:22,731 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=145729.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:12:27,056 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=145733.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:12:27,584 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-02 21:12:30,302 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.518e+02 4.899e+02 6.109e+02 7.402e+02 2.135e+03, threshold=1.222e+03, percent-clipped=2.0 +2023-04-02 21:12:32,754 INFO [train.py:903] (3/4) Epoch 22, batch 2350, loss[loss=0.1803, simple_loss=0.253, pruned_loss=0.0538, over 19741.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2879, pruned_loss=0.06402, over 3834222.55 frames. ], batch size: 45, lr: 3.75e-03, grad_scale: 4.0 +2023-04-02 21:12:54,218 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=145754.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:13:01,136 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=145760.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:13:14,112 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 21:13:31,591 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 21:13:31,939 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=145785.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:13:35,673 INFO [train.py:903] (3/4) Epoch 22, batch 2400, loss[loss=0.2605, simple_loss=0.3344, pruned_loss=0.09332, over 19544.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2881, pruned_loss=0.06393, over 3821880.76 frames. ], batch size: 54, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:13:48,799 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=145799.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:14:20,082 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=145824.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:14:34,014 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.210e+02 4.826e+02 5.747e+02 7.009e+02 1.532e+03, threshold=1.149e+03, percent-clipped=5.0 +2023-04-02 21:14:36,522 INFO [train.py:903] (3/4) Epoch 22, batch 2450, loss[loss=0.2106, simple_loss=0.2999, pruned_loss=0.06062, over 19599.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2871, pruned_loss=0.06326, over 3824539.83 frames. ], batch size: 57, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:14:38,006 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=145839.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 21:14:49,281 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=145848.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:15:37,466 INFO [train.py:903] (3/4) Epoch 22, batch 2500, loss[loss=0.2928, simple_loss=0.349, pruned_loss=0.1183, over 13237.00 frames. ], tot_loss[loss=0.207, simple_loss=0.2875, pruned_loss=0.06327, over 3829375.11 frames. ], batch size: 136, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:16:01,264 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6326, 1.3758, 1.4624, 1.4658, 3.2064, 1.1718, 2.3715, 3.6475], + device='cuda:3'), covar=tensor([0.0511, 0.2823, 0.3057, 0.1976, 0.0689, 0.2483, 0.1291, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0410, 0.0367, 0.0388, 0.0351, 0.0376, 0.0350, 0.0381, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:16:34,428 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.066e+02 4.836e+02 5.791e+02 7.519e+02 1.267e+03, threshold=1.158e+03, percent-clipped=1.0 +2023-04-02 21:16:36,601 INFO [train.py:903] (3/4) Epoch 22, batch 2550, loss[loss=0.2355, simple_loss=0.3173, pruned_loss=0.07689, over 18116.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2882, pruned_loss=0.064, over 3828756.78 frames. ], batch size: 83, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:16:59,330 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=145956.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:17:33,989 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 21:17:38,147 INFO [train.py:903] (3/4) Epoch 22, batch 2600, loss[loss=0.1999, simple_loss=0.2914, pruned_loss=0.0542, over 19522.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.288, pruned_loss=0.0641, over 3825071.26 frames. ], batch size: 54, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:17:59,423 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=146005.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:18:38,070 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.365e+02 5.022e+02 6.231e+02 7.783e+02 1.698e+03, threshold=1.246e+03, percent-clipped=5.0 +2023-04-02 21:18:40,363 INFO [train.py:903] (3/4) Epoch 22, batch 2650, loss[loss=0.1888, simple_loss=0.2694, pruned_loss=0.05404, over 19732.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2885, pruned_loss=0.06457, over 3813536.76 frames. ], batch size: 51, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:19:00,431 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 21:19:21,403 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=146071.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:19:25,863 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3466, 1.3702, 1.5645, 1.4952, 1.7455, 1.8535, 1.8126, 0.5708], + device='cuda:3'), covar=tensor([0.2433, 0.4285, 0.2562, 0.1961, 0.1626, 0.2292, 0.1395, 0.4791], + device='cuda:3'), in_proj_covar=tensor([0.0536, 0.0644, 0.0714, 0.0483, 0.0620, 0.0531, 0.0663, 0.0551], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 21:19:41,310 INFO [train.py:903] (3/4) Epoch 22, batch 2700, loss[loss=0.181, simple_loss=0.2601, pruned_loss=0.05095, over 19760.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2871, pruned_loss=0.06384, over 3817605.63 frames. ], batch size: 45, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:20:01,757 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=146104.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:20:20,683 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=146120.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:20:32,016 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=146129.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:20:39,324 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.068e+02 4.787e+02 6.288e+02 8.148e+02 2.582e+03, threshold=1.258e+03, percent-clipped=4.0 +2023-04-02 21:20:41,736 INFO [train.py:903] (3/4) Epoch 22, batch 2750, loss[loss=0.215, simple_loss=0.2996, pruned_loss=0.06522, over 19485.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2864, pruned_loss=0.06348, over 3830635.82 frames. ], batch size: 64, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:20:44,451 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2220, 2.2659, 2.4352, 2.8932, 2.1490, 2.7661, 2.4247, 2.2840], + device='cuda:3'), covar=tensor([0.4659, 0.4351, 0.2069, 0.2782, 0.4671, 0.2378, 0.5189, 0.3524], + device='cuda:3'), in_proj_covar=tensor([0.0899, 0.0963, 0.0718, 0.0934, 0.0880, 0.0816, 0.0845, 0.0782], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 21:21:18,244 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=146167.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:21:29,665 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5070, 1.4271, 2.1580, 1.6155, 2.9807, 4.7122, 4.6083, 5.0766], + device='cuda:3'), covar=tensor([0.1564, 0.3774, 0.3047, 0.2338, 0.0628, 0.0204, 0.0165, 0.0185], + device='cuda:3'), in_proj_covar=tensor([0.0271, 0.0319, 0.0351, 0.0264, 0.0241, 0.0186, 0.0214, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 21:21:37,445 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=146183.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 21:21:43,824 INFO [train.py:903] (3/4) Epoch 22, batch 2800, loss[loss=0.1781, simple_loss=0.2573, pruned_loss=0.04941, over 19393.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.286, pruned_loss=0.06304, over 3839976.28 frames. ], batch size: 47, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:22:04,640 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6539, 1.4258, 1.5189, 2.1322, 1.5436, 1.8785, 1.9296, 1.6329], + device='cuda:3'), covar=tensor([0.0897, 0.1123, 0.1090, 0.0799, 0.0985, 0.0853, 0.0970, 0.0833], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0220, 0.0223, 0.0238, 0.0224, 0.0209, 0.0186, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 21:22:19,574 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-02 21:22:42,904 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.980e+02 4.555e+02 5.863e+02 7.335e+02 1.249e+03, threshold=1.173e+03, percent-clipped=1.0 +2023-04-02 21:22:45,126 INFO [train.py:903] (3/4) Epoch 22, batch 2850, loss[loss=0.198, simple_loss=0.2803, pruned_loss=0.05786, over 19353.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2871, pruned_loss=0.06314, over 3824199.38 frames. ], batch size: 70, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:23:42,925 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 21:23:45,159 INFO [train.py:903] (3/4) Epoch 22, batch 2900, loss[loss=0.1868, simple_loss=0.2732, pruned_loss=0.05022, over 19592.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2878, pruned_loss=0.06346, over 3821538.20 frames. ], batch size: 52, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:23:57,327 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=146298.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 21:24:33,232 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=146327.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:24:43,683 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.365e+02 4.789e+02 5.854e+02 7.393e+02 1.532e+03, threshold=1.171e+03, percent-clipped=5.0 +2023-04-02 21:24:45,869 INFO [train.py:903] (3/4) Epoch 22, batch 2950, loss[loss=0.224, simple_loss=0.3026, pruned_loss=0.07268, over 19663.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2885, pruned_loss=0.06404, over 3826735.79 frames. ], batch size: 59, lr: 3.75e-03, grad_scale: 8.0 +2023-04-02 21:25:04,151 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=146352.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:25:31,426 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1404, 1.9833, 1.6102, 1.9166, 1.8765, 1.5788, 1.4318, 1.8825], + device='cuda:3'), covar=tensor([0.1055, 0.1562, 0.1746, 0.1165, 0.1471, 0.0775, 0.1807, 0.0826], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0354, 0.0309, 0.0248, 0.0298, 0.0249, 0.0306, 0.0253], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:25:32,592 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=146376.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:25:46,769 INFO [train.py:903] (3/4) Epoch 22, batch 3000, loss[loss=0.2448, simple_loss=0.3168, pruned_loss=0.08644, over 12784.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2873, pruned_loss=0.06378, over 3826800.29 frames. ], batch size: 136, lr: 3.75e-03, grad_scale: 4.0 +2023-04-02 21:25:46,770 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 21:25:59,181 INFO [train.py:937] (3/4) Epoch 22, validation: loss=0.1687, simple_loss=0.2687, pruned_loss=0.0344, over 944034.00 frames. +2023-04-02 21:25:59,182 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 21:26:02,607 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 21:26:16,170 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=146401.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:26:37,349 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1132, 3.4147, 2.0819, 2.2513, 3.0844, 1.8757, 1.5973, 2.1972], + device='cuda:3'), covar=tensor([0.1301, 0.0573, 0.1045, 0.0757, 0.0469, 0.1150, 0.0899, 0.0685], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0318, 0.0340, 0.0267, 0.0249, 0.0338, 0.0293, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:26:58,610 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.135e+02 5.066e+02 6.686e+02 8.533e+02 1.871e+03, threshold=1.337e+03, percent-clipped=6.0 +2023-04-02 21:26:59,734 INFO [train.py:903] (3/4) Epoch 22, batch 3050, loss[loss=0.2123, simple_loss=0.2786, pruned_loss=0.07299, over 19783.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2873, pruned_loss=0.06386, over 3836213.66 frames. ], batch size: 46, lr: 3.74e-03, grad_scale: 4.0 +2023-04-02 21:27:20,073 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8319, 4.2913, 4.5039, 4.5424, 1.7964, 4.2320, 3.7406, 4.2499], + device='cuda:3'), covar=tensor([0.1697, 0.0799, 0.0603, 0.0712, 0.5655, 0.0812, 0.0664, 0.1148], + device='cuda:3'), in_proj_covar=tensor([0.0785, 0.0743, 0.0948, 0.0832, 0.0833, 0.0708, 0.0570, 0.0882], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 21:28:00,898 INFO [train.py:903] (3/4) Epoch 22, batch 3100, loss[loss=0.2241, simple_loss=0.3101, pruned_loss=0.06902, over 19587.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2874, pruned_loss=0.06411, over 3818927.00 frames. ], batch size: 61, lr: 3.74e-03, grad_scale: 4.0 +2023-04-02 21:28:27,624 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=146511.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:28:59,180 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.441e+02 5.077e+02 6.322e+02 8.082e+02 1.628e+03, threshold=1.264e+03, percent-clipped=2.0 +2023-04-02 21:29:00,359 INFO [train.py:903] (3/4) Epoch 22, batch 3150, loss[loss=0.2044, simple_loss=0.2862, pruned_loss=0.06124, over 19762.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2875, pruned_loss=0.06408, over 3820335.38 frames. ], batch size: 56, lr: 3.74e-03, grad_scale: 4.0 +2023-04-02 21:29:20,747 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=146554.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 21:29:29,191 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 21:29:51,077 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=146579.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 21:30:00,731 INFO [train.py:903] (3/4) Epoch 22, batch 3200, loss[loss=0.2226, simple_loss=0.304, pruned_loss=0.07062, over 19561.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2866, pruned_loss=0.06356, over 3834203.59 frames. ], batch size: 61, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:30:04,651 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3623, 1.4372, 1.7496, 1.6045, 2.4748, 2.1483, 2.6041, 1.0335], + device='cuda:3'), covar=tensor([0.2553, 0.4292, 0.2570, 0.1925, 0.1505, 0.2155, 0.1388, 0.4450], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0639, 0.0707, 0.0479, 0.0615, 0.0526, 0.0660, 0.0547], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 21:30:20,494 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5236, 2.2766, 1.6263, 1.5529, 1.9972, 1.2359, 1.4711, 1.8943], + device='cuda:3'), covar=tensor([0.1037, 0.0775, 0.1162, 0.0879, 0.0667, 0.1430, 0.0749, 0.0562], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0318, 0.0340, 0.0267, 0.0249, 0.0338, 0.0293, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:30:28,042 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=146609.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:30:40,791 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6801, 1.5088, 1.6333, 1.4886, 3.2944, 1.1991, 2.4886, 3.6871], + device='cuda:3'), covar=tensor([0.0452, 0.2576, 0.2617, 0.1927, 0.0691, 0.2396, 0.1132, 0.0235], + device='cuda:3'), in_proj_covar=tensor([0.0408, 0.0364, 0.0386, 0.0349, 0.0372, 0.0348, 0.0380, 0.0404], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:30:47,443 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=146626.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:31:01,672 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.447e+02 4.719e+02 5.878e+02 7.469e+02 1.229e+03, threshold=1.176e+03, percent-clipped=0.0 +2023-04-02 21:31:02,824 INFO [train.py:903] (3/4) Epoch 22, batch 3250, loss[loss=0.1961, simple_loss=0.282, pruned_loss=0.05506, over 17477.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2871, pruned_loss=0.06405, over 3822772.42 frames. ], batch size: 101, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:31:10,883 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=146644.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:31:25,803 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1774, 2.8677, 2.2732, 2.2597, 2.0480, 2.5376, 1.0520, 2.0628], + device='cuda:3'), covar=tensor([0.0616, 0.0584, 0.0615, 0.0960, 0.0974, 0.0957, 0.1270, 0.0984], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0356, 0.0358, 0.0382, 0.0457, 0.0386, 0.0336, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 21:32:03,196 INFO [train.py:903] (3/4) Epoch 22, batch 3300, loss[loss=0.1914, simple_loss=0.2772, pruned_loss=0.05281, over 19775.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2868, pruned_loss=0.06424, over 3822323.48 frames. ], batch size: 56, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:32:09,858 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 21:32:11,322 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0489, 1.7631, 1.9167, 2.6095, 2.0514, 2.2498, 2.3324, 2.1138], + device='cuda:3'), covar=tensor([0.0719, 0.0859, 0.0888, 0.0721, 0.0831, 0.0713, 0.0801, 0.0625], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0225, 0.0240, 0.0226, 0.0210, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 21:32:59,758 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 5.268e+02 6.638e+02 8.509e+02 1.642e+03, threshold=1.328e+03, percent-clipped=7.0 +2023-04-02 21:33:00,760 INFO [train.py:903] (3/4) Epoch 22, batch 3350, loss[loss=0.2778, simple_loss=0.3385, pruned_loss=0.1086, over 19700.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2883, pruned_loss=0.06491, over 3820137.38 frames. ], batch size: 63, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:33:41,153 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8313, 4.4285, 2.8587, 3.8830, 1.0449, 4.3082, 4.2459, 4.3733], + device='cuda:3'), covar=tensor([0.0555, 0.0999, 0.1751, 0.0742, 0.3897, 0.0619, 0.0831, 0.0963], + device='cuda:3'), in_proj_covar=tensor([0.0504, 0.0411, 0.0492, 0.0344, 0.0402, 0.0431, 0.0424, 0.0459], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:34:00,057 INFO [train.py:903] (3/4) Epoch 22, batch 3400, loss[loss=0.2644, simple_loss=0.348, pruned_loss=0.09043, over 19719.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2877, pruned_loss=0.06411, over 3810186.29 frames. ], batch size: 59, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:34:59,710 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.964e+02 5.013e+02 6.159e+02 8.066e+02 2.491e+03, threshold=1.232e+03, percent-clipped=4.0 +2023-04-02 21:35:00,913 INFO [train.py:903] (3/4) Epoch 22, batch 3450, loss[loss=0.2454, simple_loss=0.3211, pruned_loss=0.08487, over 19539.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2886, pruned_loss=0.06472, over 3789411.92 frames. ], batch size: 56, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:35:04,225 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 21:35:28,808 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=146862.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:35:54,065 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=146882.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:35:55,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8772, 4.2745, 4.6066, 4.6173, 1.7292, 4.3487, 3.7832, 4.2787], + device='cuda:3'), covar=tensor([0.1661, 0.0903, 0.0619, 0.0663, 0.6096, 0.0912, 0.0672, 0.1224], + device='cuda:3'), in_proj_covar=tensor([0.0778, 0.0739, 0.0944, 0.0827, 0.0826, 0.0707, 0.0567, 0.0878], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 21:35:59,851 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 21:36:01,458 INFO [train.py:903] (3/4) Epoch 22, batch 3500, loss[loss=0.2569, simple_loss=0.3273, pruned_loss=0.09323, over 17301.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2891, pruned_loss=0.06507, over 3780752.04 frames. ], batch size: 101, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:36:23,421 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=146907.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:37:00,108 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.850e+02 4.615e+02 6.325e+02 8.235e+02 2.059e+03, threshold=1.265e+03, percent-clipped=6.0 +2023-04-02 21:37:01,340 INFO [train.py:903] (3/4) Epoch 22, batch 3550, loss[loss=0.1927, simple_loss=0.2609, pruned_loss=0.06226, over 19322.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2883, pruned_loss=0.065, over 3780646.32 frames. ], batch size: 44, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:37:18,272 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=146953.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:38:02,336 INFO [train.py:903] (3/4) Epoch 22, batch 3600, loss[loss=0.2091, simple_loss=0.2958, pruned_loss=0.06122, over 19673.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2882, pruned_loss=0.06476, over 3800233.71 frames. ], batch size: 60, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:38:02,530 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=146988.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:39:01,661 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.391e+02 4.948e+02 6.297e+02 8.667e+02 2.605e+03, threshold=1.259e+03, percent-clipped=8.0 +2023-04-02 21:39:02,712 INFO [train.py:903] (3/4) Epoch 22, batch 3650, loss[loss=0.2066, simple_loss=0.2891, pruned_loss=0.06203, over 19735.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2886, pruned_loss=0.06514, over 3800472.66 frames. ], batch size: 63, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:39:07,563 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6007, 1.5280, 1.4582, 2.0560, 1.4700, 2.0120, 1.9422, 1.7547], + device='cuda:3'), covar=tensor([0.0855, 0.0897, 0.1027, 0.0733, 0.0907, 0.0681, 0.0792, 0.0663], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0222, 0.0226, 0.0241, 0.0227, 0.0211, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-02 21:39:25,249 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2751, 2.1725, 1.9823, 1.8358, 1.6255, 1.8638, 0.6159, 1.2825], + device='cuda:3'), covar=tensor([0.0631, 0.0645, 0.0529, 0.0875, 0.1282, 0.0945, 0.1409, 0.1055], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0356, 0.0360, 0.0383, 0.0459, 0.0387, 0.0337, 0.0342], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 21:39:39,058 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147068.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:40:03,857 INFO [train.py:903] (3/4) Epoch 22, batch 3700, loss[loss=0.2268, simple_loss=0.31, pruned_loss=0.07177, over 19504.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2896, pruned_loss=0.06561, over 3789821.95 frames. ], batch size: 64, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:40:20,504 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-02 21:40:21,169 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147103.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:40:25,103 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.93 vs. limit=5.0 +2023-04-02 21:41:02,895 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.261e+02 4.816e+02 5.960e+02 7.144e+02 1.653e+03, threshold=1.192e+03, percent-clipped=4.0 +2023-04-02 21:41:04,071 INFO [train.py:903] (3/4) Epoch 22, batch 3750, loss[loss=0.2003, simple_loss=0.2887, pruned_loss=0.05594, over 19056.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2895, pruned_loss=0.06524, over 3800247.49 frames. ], batch size: 69, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:42:04,509 INFO [train.py:903] (3/4) Epoch 22, batch 3800, loss[loss=0.2106, simple_loss=0.2867, pruned_loss=0.06727, over 19674.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2874, pruned_loss=0.06442, over 3810319.03 frames. ], batch size: 53, lr: 3.74e-03, grad_scale: 8.0 +2023-04-02 21:42:26,702 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=147206.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:42:38,890 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-02 21:42:40,377 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4381, 1.2313, 1.5771, 1.7240, 2.9911, 1.3357, 2.2132, 3.3258], + device='cuda:3'), covar=tensor([0.0553, 0.2978, 0.2827, 0.1618, 0.0723, 0.2212, 0.1339, 0.0309], + device='cuda:3'), in_proj_covar=tensor([0.0405, 0.0363, 0.0384, 0.0347, 0.0372, 0.0346, 0.0378, 0.0401], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:43:02,578 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.867e+02 5.077e+02 5.970e+02 7.548e+02 1.289e+03, threshold=1.194e+03, percent-clipped=1.0 +2023-04-02 21:43:03,536 INFO [train.py:903] (3/4) Epoch 22, batch 3850, loss[loss=0.2336, simple_loss=0.31, pruned_loss=0.07863, over 19596.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2874, pruned_loss=0.0642, over 3827729.27 frames. ], batch size: 57, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:43:31,150 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147259.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:44:05,363 INFO [train.py:903] (3/4) Epoch 22, batch 3900, loss[loss=0.2257, simple_loss=0.3044, pruned_loss=0.07352, over 13385.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2879, pruned_loss=0.06437, over 3819136.97 frames. ], batch size: 135, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:44:42,217 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1711, 2.0560, 1.9432, 2.2036, 2.0456, 1.8372, 1.9086, 2.1094], + device='cuda:3'), covar=tensor([0.0797, 0.1175, 0.1176, 0.0900, 0.1063, 0.0512, 0.1097, 0.0590], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0354, 0.0310, 0.0250, 0.0301, 0.0251, 0.0308, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:44:46,178 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147321.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:44:49,712 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=147324.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:44:55,066 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147328.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:45:04,849 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.696e+02 4.890e+02 6.799e+02 8.609e+02 1.784e+03, threshold=1.360e+03, percent-clipped=9.0 +2023-04-02 21:45:05,883 INFO [train.py:903] (3/4) Epoch 22, batch 3950, loss[loss=0.256, simple_loss=0.3289, pruned_loss=0.09159, over 19773.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2889, pruned_loss=0.06505, over 3794052.01 frames. ], batch size: 56, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:45:08,151 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-02 21:45:18,232 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=147349.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:45:30,101 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=147359.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:46:00,887 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=147384.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:46:04,830 INFO [train.py:903] (3/4) Epoch 22, batch 4000, loss[loss=0.2191, simple_loss=0.2948, pruned_loss=0.07166, over 19591.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2881, pruned_loss=0.0647, over 3791870.42 frames. ], batch size: 57, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:46:50,160 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147425.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:46:52,234 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-02 21:47:03,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.354e+02 4.920e+02 5.925e+02 7.748e+02 1.160e+03, threshold=1.185e+03, percent-clipped=0.0 +2023-04-02 21:47:05,863 INFO [train.py:903] (3/4) Epoch 22, batch 4050, loss[loss=0.2177, simple_loss=0.2926, pruned_loss=0.07138, over 17472.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2884, pruned_loss=0.06455, over 3799384.97 frames. ], batch size: 101, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:47:24,617 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147452.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:47:40,272 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1908, 1.4497, 1.8883, 1.6350, 3.0706, 4.5900, 4.4141, 5.0299], + device='cuda:3'), covar=tensor([0.1753, 0.3794, 0.3516, 0.2317, 0.0581, 0.0203, 0.0162, 0.0156], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0324, 0.0356, 0.0267, 0.0245, 0.0188, 0.0217, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 21:47:51,679 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2294, 2.2996, 2.4810, 2.9497, 2.2365, 2.8572, 2.4603, 2.2883], + device='cuda:3'), covar=tensor([0.4085, 0.4017, 0.1887, 0.2673, 0.4516, 0.2187, 0.4752, 0.3257], + device='cuda:3'), in_proj_covar=tensor([0.0899, 0.0964, 0.0714, 0.0932, 0.0877, 0.0815, 0.0842, 0.0779], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 21:48:07,526 INFO [train.py:903] (3/4) Epoch 22, batch 4100, loss[loss=0.2339, simple_loss=0.3106, pruned_loss=0.07859, over 18733.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.289, pruned_loss=0.06485, over 3802502.40 frames. ], batch size: 74, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:48:44,884 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-02 21:49:04,890 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1369, 1.8202, 1.6672, 2.0352, 1.7607, 1.7890, 1.6370, 2.0168], + device='cuda:3'), covar=tensor([0.0990, 0.1447, 0.1492, 0.0993, 0.1318, 0.0559, 0.1482, 0.0704], + device='cuda:3'), in_proj_covar=tensor([0.0269, 0.0355, 0.0311, 0.0250, 0.0301, 0.0251, 0.0309, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:49:07,917 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.284e+02 5.254e+02 6.349e+02 7.493e+02 1.711e+03, threshold=1.270e+03, percent-clipped=4.0 +2023-04-02 21:49:09,113 INFO [train.py:903] (3/4) Epoch 22, batch 4150, loss[loss=0.18, simple_loss=0.2657, pruned_loss=0.0471, over 19413.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2886, pruned_loss=0.06438, over 3792173.13 frames. ], batch size: 48, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:49:56,983 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=147577.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:50:09,104 INFO [train.py:903] (3/4) Epoch 22, batch 4200, loss[loss=0.2437, simple_loss=0.3105, pruned_loss=0.08845, over 13196.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2889, pruned_loss=0.06478, over 3787490.67 frames. ], batch size: 136, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:50:13,779 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-02 21:50:15,176 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6069, 1.3087, 1.5058, 1.4153, 3.2064, 1.0115, 2.2797, 3.5786], + device='cuda:3'), covar=tensor([0.0498, 0.2843, 0.2948, 0.1994, 0.0772, 0.2607, 0.1315, 0.0269], + device='cuda:3'), in_proj_covar=tensor([0.0407, 0.0365, 0.0387, 0.0349, 0.0373, 0.0346, 0.0380, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:50:26,629 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=147602.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:50:27,412 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=147603.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:50:59,156 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7431, 1.6502, 1.8459, 1.8865, 3.3766, 1.3886, 2.5575, 3.7272], + device='cuda:3'), covar=tensor([0.0484, 0.2614, 0.2547, 0.1682, 0.0702, 0.2331, 0.1400, 0.0250], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0364, 0.0385, 0.0348, 0.0373, 0.0346, 0.0380, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:51:09,771 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.286e+02 4.933e+02 5.671e+02 7.351e+02 2.024e+03, threshold=1.134e+03, percent-clipped=5.0 +2023-04-02 21:51:10,939 INFO [train.py:903] (3/4) Epoch 22, batch 4250, loss[loss=0.196, simple_loss=0.2745, pruned_loss=0.05875, over 19483.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2888, pruned_loss=0.06475, over 3784992.06 frames. ], batch size: 49, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:51:25,902 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-02 21:51:38,186 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-02 21:51:45,138 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2066, 1.2694, 1.7375, 1.3288, 2.6165, 3.5236, 3.2010, 3.6960], + device='cuda:3'), covar=tensor([0.1711, 0.3950, 0.3450, 0.2448, 0.0616, 0.0182, 0.0214, 0.0264], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0324, 0.0355, 0.0267, 0.0245, 0.0188, 0.0216, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 21:51:51,748 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=147672.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:51:57,375 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147677.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:52:11,231 INFO [train.py:903] (3/4) Epoch 22, batch 4300, loss[loss=0.1882, simple_loss=0.2702, pruned_loss=0.05311, over 19624.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2875, pruned_loss=0.06386, over 3799346.24 frames. ], batch size: 50, lr: 3.73e-03, grad_scale: 4.0 +2023-04-02 21:52:47,218 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147718.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:52:52,679 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5802, 1.5099, 1.4482, 1.8985, 1.4744, 1.7461, 1.8175, 1.6884], + device='cuda:3'), covar=tensor([0.0856, 0.0952, 0.1060, 0.0763, 0.0863, 0.0817, 0.0852, 0.0709], + device='cuda:3'), in_proj_covar=tensor([0.0209, 0.0219, 0.0222, 0.0237, 0.0224, 0.0211, 0.0184, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 21:53:02,985 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-02 21:53:11,622 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.973e+02 5.055e+02 6.219e+02 8.115e+02 2.735e+03, threshold=1.244e+03, percent-clipped=11.0 +2023-04-02 21:53:11,644 INFO [train.py:903] (3/4) Epoch 22, batch 4350, loss[loss=0.196, simple_loss=0.2775, pruned_loss=0.05724, over 19609.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2889, pruned_loss=0.06485, over 3789527.13 frames. ], batch size: 50, lr: 3.73e-03, grad_scale: 4.0 +2023-04-02 21:53:48,737 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=147769.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:54:11,201 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147787.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:54:12,001 INFO [train.py:903] (3/4) Epoch 22, batch 4400, loss[loss=0.2259, simple_loss=0.3182, pruned_loss=0.06676, over 19263.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2887, pruned_loss=0.06502, over 3784693.42 frames. ], batch size: 66, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:54:20,677 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=147796.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:54:37,131 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-02 21:54:46,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-02 21:55:06,308 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147833.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:55:12,512 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.305e+02 5.137e+02 6.374e+02 7.743e+02 1.534e+03, threshold=1.275e+03, percent-clipped=3.0 +2023-04-02 21:55:12,529 INFO [train.py:903] (3/4) Epoch 22, batch 4450, loss[loss=0.1824, simple_loss=0.2664, pruned_loss=0.04923, over 19619.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.2887, pruned_loss=0.06514, over 3771869.93 frames. ], batch size: 50, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:55:57,402 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147875.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:56:08,754 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147884.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:56:08,809 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3649, 2.1363, 1.5836, 1.4026, 1.9727, 1.2644, 1.2086, 1.7923], + device='cuda:3'), covar=tensor([0.1117, 0.0788, 0.1085, 0.0925, 0.0545, 0.1327, 0.0762, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0336, 0.0263, 0.0248, 0.0336, 0.0289, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:56:12,918 INFO [train.py:903] (3/4) Epoch 22, batch 4500, loss[loss=0.2017, simple_loss=0.2895, pruned_loss=0.05696, over 19787.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2881, pruned_loss=0.06477, over 3784589.89 frames. ], batch size: 56, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:56:41,882 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=147911.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:57:03,801 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=147929.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:57:15,412 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.490e+02 4.535e+02 5.624e+02 7.235e+02 1.683e+03, threshold=1.125e+03, percent-clipped=3.0 +2023-04-02 21:57:15,434 INFO [train.py:903] (3/4) Epoch 22, batch 4550, loss[loss=0.2346, simple_loss=0.3149, pruned_loss=0.07719, over 19520.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2895, pruned_loss=0.06507, over 3800845.15 frames. ], batch size: 54, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:57:23,466 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-02 21:57:46,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-02 21:57:58,529 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=147974.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:58:15,825 INFO [train.py:903] (3/4) Epoch 22, batch 4600, loss[loss=0.2152, simple_loss=0.2965, pruned_loss=0.06698, over 19662.00 frames. ], tot_loss[loss=0.2095, simple_loss=0.289, pruned_loss=0.06496, over 3810633.56 frames. ], batch size: 55, lr: 3.73e-03, grad_scale: 8.0 +2023-04-02 21:58:28,678 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=147999.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:58:56,686 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148021.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:59:14,041 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6137, 1.2847, 1.2170, 1.4898, 1.1622, 1.3329, 1.2239, 1.4026], + device='cuda:3'), covar=tensor([0.1087, 0.1164, 0.1556, 0.1030, 0.1288, 0.0632, 0.1500, 0.0821], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0354, 0.0311, 0.0250, 0.0301, 0.0250, 0.0308, 0.0254], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 21:59:16,085 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.327e+02 4.857e+02 6.032e+02 8.227e+02 1.754e+03, threshold=1.206e+03, percent-clipped=4.0 +2023-04-02 21:59:16,102 INFO [train.py:903] (3/4) Epoch 22, batch 4650, loss[loss=0.1883, simple_loss=0.26, pruned_loss=0.05835, over 19747.00 frames. ], tot_loss[loss=0.2094, simple_loss=0.2888, pruned_loss=0.06494, over 3814268.20 frames. ], batch size: 46, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 21:59:22,687 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148043.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 21:59:32,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-02 21:59:43,982 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-02 21:59:53,353 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148068.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:00:16,032 INFO [train.py:903] (3/4) Epoch 22, batch 4700, loss[loss=0.2219, simple_loss=0.3086, pruned_loss=0.06756, over 19650.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2891, pruned_loss=0.06511, over 3818210.70 frames. ], batch size: 58, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:00:39,956 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-02 22:01:15,659 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148136.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:01:17,614 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.017e+02 5.448e+02 6.322e+02 7.572e+02 1.580e+03, threshold=1.264e+03, percent-clipped=4.0 +2023-04-02 22:01:17,632 INFO [train.py:903] (3/4) Epoch 22, batch 4750, loss[loss=0.236, simple_loss=0.3098, pruned_loss=0.08105, over 19688.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2886, pruned_loss=0.06486, over 3826504.83 frames. ], batch size: 59, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:01:21,299 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148140.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:01:50,465 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:01:52,792 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148167.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:02:05,586 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148177.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:02:18,759 INFO [train.py:903] (3/4) Epoch 22, batch 4800, loss[loss=0.1915, simple_loss=0.27, pruned_loss=0.05646, over 19834.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2889, pruned_loss=0.06512, over 3830127.52 frames. ], batch size: 52, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:02:23,669 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148192.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:02:57,716 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148219.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:03:14,872 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5263, 4.1280, 2.6725, 3.6297, 1.2123, 4.0160, 3.9576, 4.0383], + device='cuda:3'), covar=tensor([0.0633, 0.0908, 0.2007, 0.0856, 0.3684, 0.0763, 0.0939, 0.1219], + device='cuda:3'), in_proj_covar=tensor([0.0505, 0.0409, 0.0493, 0.0345, 0.0398, 0.0433, 0.0424, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:03:18,978 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.136e+02 4.799e+02 5.785e+02 7.279e+02 1.291e+03, threshold=1.157e+03, percent-clipped=1.0 +2023-04-02 22:03:18,995 INFO [train.py:903] (3/4) Epoch 22, batch 4850, loss[loss=0.1965, simple_loss=0.2781, pruned_loss=0.05747, over 19590.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2887, pruned_loss=0.0646, over 3837959.17 frames. ], batch size: 52, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:03:44,130 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-02 22:03:47,705 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9294, 1.3061, 1.6452, 1.5299, 3.4889, 1.1699, 2.5265, 3.8943], + device='cuda:3'), covar=tensor([0.0436, 0.2906, 0.2753, 0.1967, 0.0715, 0.2510, 0.1247, 0.0231], + device='cuda:3'), in_proj_covar=tensor([0.0406, 0.0364, 0.0385, 0.0346, 0.0373, 0.0347, 0.0380, 0.0402], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:04:01,911 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148273.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:04:02,927 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-02 22:04:08,079 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-02 22:04:09,252 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-02 22:04:18,710 INFO [train.py:903] (3/4) Epoch 22, batch 4900, loss[loss=0.2124, simple_loss=0.2682, pruned_loss=0.07828, over 19736.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2887, pruned_loss=0.0648, over 3842255.61 frames. ], batch size: 45, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:04:18,724 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-02 22:04:24,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148292.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:04:37,221 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4311, 1.2846, 1.4899, 1.3842, 3.0319, 1.1112, 2.3395, 3.3587], + device='cuda:3'), covar=tensor([0.0488, 0.2790, 0.2913, 0.1950, 0.0667, 0.2498, 0.1231, 0.0297], + device='cuda:3'), in_proj_covar=tensor([0.0408, 0.0365, 0.0385, 0.0347, 0.0374, 0.0348, 0.0381, 0.0403], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:04:39,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-02 22:04:40,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=148305.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:05:14,441 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148334.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:05:19,520 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.031e+02 5.396e+02 6.572e+02 8.433e+02 1.736e+03, threshold=1.314e+03, percent-clipped=6.0 +2023-04-02 22:05:19,538 INFO [train.py:903] (3/4) Epoch 22, batch 4950, loss[loss=0.1809, simple_loss=0.2644, pruned_loss=0.0487, over 19531.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2893, pruned_loss=0.0651, over 3833523.62 frames. ], batch size: 54, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:05:33,636 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-04-02 22:05:35,934 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-02 22:06:01,184 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-02 22:06:10,281 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8663, 1.9264, 2.0891, 2.3849, 1.9166, 2.2853, 2.1621, 1.9913], + device='cuda:3'), covar=tensor([0.3512, 0.3133, 0.1533, 0.1982, 0.3207, 0.1712, 0.3729, 0.2651], + device='cuda:3'), in_proj_covar=tensor([0.0897, 0.0963, 0.0716, 0.0929, 0.0879, 0.0815, 0.0841, 0.0781], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 22:06:15,530 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1707, 3.6175, 3.7650, 3.7534, 1.7083, 3.5315, 3.1571, 3.5410], + device='cuda:3'), covar=tensor([0.1656, 0.1434, 0.0728, 0.0842, 0.5569, 0.1228, 0.0703, 0.1161], + device='cuda:3'), in_proj_covar=tensor([0.0772, 0.0739, 0.0939, 0.0820, 0.0823, 0.0703, 0.0562, 0.0874], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 22:06:20,895 INFO [train.py:903] (3/4) Epoch 22, batch 5000, loss[loss=0.1729, simple_loss=0.2457, pruned_loss=0.05005, over 19742.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2883, pruned_loss=0.06447, over 3834158.45 frames. ], batch size: 45, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:06:21,237 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148388.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:06:25,709 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148392.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:06:29,621 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-02 22:06:40,558 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-02 22:06:55,461 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148417.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:07:19,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.597e+02 4.828e+02 6.337e+02 7.923e+02 1.456e+03, threshold=1.267e+03, percent-clipped=1.0 +2023-04-02 22:07:19,262 INFO [train.py:903] (3/4) Epoch 22, batch 5050, loss[loss=0.2127, simple_loss=0.297, pruned_loss=0.0642, over 19526.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2887, pruned_loss=0.06467, over 3842384.32 frames. ], batch size: 56, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:07:30,120 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=148447.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:07:54,481 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-02 22:08:14,134 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9302, 1.3337, 1.0662, 1.0049, 1.1850, 1.0019, 0.9748, 1.1976], + device='cuda:3'), covar=tensor([0.0661, 0.0885, 0.1222, 0.0749, 0.0686, 0.1373, 0.0600, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0317, 0.0337, 0.0265, 0.0249, 0.0338, 0.0291, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:08:19,393 INFO [train.py:903] (3/4) Epoch 22, batch 5100, loss[loss=0.213, simple_loss=0.2868, pruned_loss=0.06962, over 19621.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2893, pruned_loss=0.06493, over 3834976.13 frames. ], batch size: 50, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:08:30,462 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-02 22:08:33,784 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-02 22:08:39,206 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-02 22:08:47,624 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8614, 1.3318, 1.0722, 1.0164, 1.1845, 1.0235, 0.9160, 1.2091], + device='cuda:3'), covar=tensor([0.0673, 0.0919, 0.1155, 0.0740, 0.0580, 0.1315, 0.0594, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0334, 0.0263, 0.0248, 0.0336, 0.0289, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:08:53,294 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=148516.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:09:04,405 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0630, 4.4645, 4.8387, 4.8202, 1.8571, 4.5210, 3.9539, 4.5434], + device='cuda:3'), covar=tensor([0.1629, 0.0802, 0.0569, 0.0630, 0.5722, 0.0866, 0.0646, 0.1050], + device='cuda:3'), in_proj_covar=tensor([0.0777, 0.0740, 0.0943, 0.0823, 0.0828, 0.0706, 0.0567, 0.0877], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 22:09:19,526 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.928e+02 5.468e+02 6.941e+02 9.893e+02 2.948e+03, threshold=1.388e+03, percent-clipped=12.0 +2023-04-02 22:09:19,544 INFO [train.py:903] (3/4) Epoch 22, batch 5150, loss[loss=0.2095, simple_loss=0.2898, pruned_loss=0.0646, over 19548.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2886, pruned_loss=0.06486, over 3819781.03 frames. ], batch size: 54, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:09:31,357 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-02 22:09:32,953 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148548.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:10:02,810 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148573.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:10:05,888 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 22:10:20,793 INFO [train.py:903] (3/4) Epoch 22, batch 5200, loss[loss=0.2374, simple_loss=0.3147, pruned_loss=0.08008, over 13502.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2887, pruned_loss=0.06484, over 3813829.20 frames. ], batch size: 136, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:10:23,564 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148590.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:10:33,156 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-02 22:10:53,704 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148615.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:11:17,524 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-02 22:11:21,001 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.230e+02 4.558e+02 5.776e+02 7.251e+02 2.001e+03, threshold=1.155e+03, percent-clipped=2.0 +2023-04-02 22:11:21,019 INFO [train.py:903] (3/4) Epoch 22, batch 5250, loss[loss=0.1754, simple_loss=0.248, pruned_loss=0.05139, over 19763.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2871, pruned_loss=0.0638, over 3817823.41 frames. ], batch size: 46, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:11:28,745 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=148644.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:11:33,740 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148649.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:11:58,326 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=148669.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:11:58,441 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=148669.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:12:20,276 INFO [train.py:903] (3/4) Epoch 22, batch 5300, loss[loss=0.2331, simple_loss=0.3101, pruned_loss=0.07807, over 19480.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2883, pruned_loss=0.0646, over 3824860.80 frames. ], batch size: 64, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:12:39,144 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-02 22:13:17,770 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=148734.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:13:22,172 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.679e+02 5.309e+02 6.457e+02 8.011e+02 2.116e+03, threshold=1.291e+03, percent-clipped=5.0 +2023-04-02 22:13:22,190 INFO [train.py:903] (3/4) Epoch 22, batch 5350, loss[loss=0.2209, simple_loss=0.2992, pruned_loss=0.07124, over 17249.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2877, pruned_loss=0.06431, over 3812870.66 frames. ], batch size: 101, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:13:54,749 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148764.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:13:55,483 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-02 22:14:16,567 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-02 22:14:24,407 INFO [train.py:903] (3/4) Epoch 22, batch 5400, loss[loss=0.2481, simple_loss=0.3224, pruned_loss=0.08687, over 19678.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.2882, pruned_loss=0.06446, over 3803591.62 frames. ], batch size: 60, lr: 3.72e-03, grad_scale: 8.0 +2023-04-02 22:14:28,067 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148791.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:14:34,166 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-02 22:15:24,093 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.599e+02 4.882e+02 5.931e+02 8.184e+02 2.288e+03, threshold=1.186e+03, percent-clipped=7.0 +2023-04-02 22:15:24,111 INFO [train.py:903] (3/4) Epoch 22, batch 5450, loss[loss=0.2049, simple_loss=0.2838, pruned_loss=0.06296, over 19758.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2886, pruned_loss=0.06437, over 3825866.23 frames. ], batch size: 54, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:15:50,186 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=148860.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:16:24,046 INFO [train.py:903] (3/4) Epoch 22, batch 5500, loss[loss=0.208, simple_loss=0.2875, pruned_loss=0.06426, over 19505.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2885, pruned_loss=0.06447, over 3832364.55 frames. ], batch size: 64, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:16:47,587 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148906.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:16:49,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-02 22:17:25,270 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.265e+02 4.956e+02 6.031e+02 8.410e+02 1.981e+03, threshold=1.206e+03, percent-clipped=11.0 +2023-04-02 22:17:25,288 INFO [train.py:903] (3/4) Epoch 22, batch 5550, loss[loss=0.2152, simple_loss=0.3009, pruned_loss=0.06477, over 19742.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.2886, pruned_loss=0.06421, over 3839613.57 frames. ], batch size: 63, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:17:33,832 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-02 22:18:10,834 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=148975.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:18:11,970 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4627, 2.1187, 1.5633, 1.5037, 1.9722, 1.3166, 1.4003, 1.7494], + device='cuda:3'), covar=tensor([0.1015, 0.0800, 0.1109, 0.0798, 0.0553, 0.1299, 0.0711, 0.0564], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0318, 0.0339, 0.0265, 0.0249, 0.0339, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:18:21,590 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-02 22:18:27,005 INFO [train.py:903] (3/4) Epoch 22, batch 5600, loss[loss=0.2059, simple_loss=0.297, pruned_loss=0.05738, over 19757.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2887, pruned_loss=0.06434, over 3823010.29 frames. ], batch size: 63, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:18:41,576 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1694, 2.8809, 2.2853, 2.2349, 2.0950, 2.5294, 1.1080, 2.0731], + device='cuda:3'), covar=tensor([0.0696, 0.0641, 0.0683, 0.1256, 0.1167, 0.1062, 0.1394, 0.1042], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0355, 0.0357, 0.0381, 0.0460, 0.0388, 0.0335, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 22:18:56,784 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=149013.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:18:59,351 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-04-02 22:19:05,995 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=149020.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:19:27,592 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.511e+02 5.096e+02 6.059e+02 8.103e+02 1.757e+03, threshold=1.212e+03, percent-clipped=10.0 +2023-04-02 22:19:27,609 INFO [train.py:903] (3/4) Epoch 22, batch 5650, loss[loss=0.2188, simple_loss=0.2992, pruned_loss=0.06924, over 19525.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.289, pruned_loss=0.06461, over 3810086.71 frames. ], batch size: 56, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:19:29,171 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4562, 1.4151, 1.3739, 1.6673, 1.2252, 1.6390, 1.5980, 1.5050], + device='cuda:3'), covar=tensor([0.0907, 0.1022, 0.1095, 0.0823, 0.0989, 0.0828, 0.0931, 0.0791], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0224, 0.0240, 0.0227, 0.0213, 0.0186, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 22:19:35,829 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=149045.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:20:15,157 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-02 22:20:16,299 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=149078.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:20:27,395 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4913, 3.6680, 4.0122, 4.0126, 2.3345, 3.7463, 3.3982, 3.8047], + device='cuda:3'), covar=tensor([0.1514, 0.2900, 0.0686, 0.0787, 0.4705, 0.1369, 0.0674, 0.1093], + device='cuda:3'), in_proj_covar=tensor([0.0789, 0.0747, 0.0955, 0.0838, 0.0845, 0.0720, 0.0571, 0.0888], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 22:20:28,279 INFO [train.py:903] (3/4) Epoch 22, batch 5700, loss[loss=0.2294, simple_loss=0.3008, pruned_loss=0.079, over 13335.00 frames. ], tot_loss[loss=0.2089, simple_loss=0.2888, pruned_loss=0.06455, over 3797995.00 frames. ], batch size: 138, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:21:17,818 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=149128.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:21:29,598 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.928e+02 4.618e+02 6.077e+02 7.635e+02 1.470e+03, threshold=1.215e+03, percent-clipped=6.0 +2023-04-02 22:21:29,617 INFO [train.py:903] (3/4) Epoch 22, batch 5750, loss[loss=0.1856, simple_loss=0.2763, pruned_loss=0.04743, over 19771.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2886, pruned_loss=0.06431, over 3800674.77 frames. ], batch size: 56, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:21:30,804 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-02 22:21:39,626 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-02 22:21:46,345 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-02 22:21:59,276 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=149162.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:22:29,865 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=149187.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:22:30,652 INFO [train.py:903] (3/4) Epoch 22, batch 5800, loss[loss=0.1825, simple_loss=0.2642, pruned_loss=0.05039, over 19619.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2887, pruned_loss=0.06471, over 3789287.95 frames. ], batch size: 50, lr: 3.71e-03, grad_scale: 4.0 +2023-04-02 22:22:37,212 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=149193.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:23:23,102 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=149231.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:23:30,416 INFO [train.py:903] (3/4) Epoch 22, batch 5850, loss[loss=0.2675, simple_loss=0.3283, pruned_loss=0.1033, over 13621.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2891, pruned_loss=0.06515, over 3789511.56 frames. ], batch size: 135, lr: 3.71e-03, grad_scale: 4.0 +2023-04-02 22:23:31,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.414e+02 5.174e+02 6.346e+02 7.936e+02 1.645e+03, threshold=1.269e+03, percent-clipped=7.0 +2023-04-02 22:23:52,875 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=149256.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:24:20,905 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-02 22:24:30,287 INFO [train.py:903] (3/4) Epoch 22, batch 5900, loss[loss=0.1873, simple_loss=0.2614, pruned_loss=0.05659, over 19748.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2884, pruned_loss=0.06457, over 3805197.23 frames. ], batch size: 46, lr: 3.71e-03, grad_scale: 4.0 +2023-04-02 22:24:35,574 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-02 22:24:56,377 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-02 22:25:06,676 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1561, 1.3289, 1.5630, 1.4324, 2.7789, 1.0663, 2.2065, 3.0595], + device='cuda:3'), covar=tensor([0.0529, 0.2762, 0.2773, 0.1694, 0.0686, 0.2393, 0.1146, 0.0332], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0368, 0.0388, 0.0348, 0.0376, 0.0352, 0.0384, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:25:12,709 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 22:25:31,732 INFO [train.py:903] (3/4) Epoch 22, batch 5950, loss[loss=0.2536, simple_loss=0.3238, pruned_loss=0.09164, over 19664.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.2891, pruned_loss=0.06502, over 3796167.37 frames. ], batch size: 58, lr: 3.71e-03, grad_scale: 4.0 +2023-04-02 22:25:32,879 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.792e+02 4.957e+02 5.985e+02 7.132e+02 1.534e+03, threshold=1.197e+03, percent-clipped=1.0 +2023-04-02 22:25:48,140 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-02 22:26:28,512 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=149384.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:26:32,557 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0896, 2.8728, 1.8481, 1.6214, 2.8073, 1.6372, 1.4706, 2.3624], + device='cuda:3'), covar=tensor([0.1042, 0.0745, 0.0950, 0.1025, 0.0440, 0.1238, 0.0947, 0.0553], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0317, 0.0338, 0.0265, 0.0247, 0.0337, 0.0290, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:26:33,225 INFO [train.py:903] (3/4) Epoch 22, batch 6000, loss[loss=0.1971, simple_loss=0.2868, pruned_loss=0.05371, over 17384.00 frames. ], tot_loss[loss=0.2089, simple_loss=0.2889, pruned_loss=0.06446, over 3791408.18 frames. ], batch size: 101, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:26:33,225 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 22:26:46,897 INFO [train.py:937] (3/4) Epoch 22, validation: loss=0.1681, simple_loss=0.2682, pruned_loss=0.03398, over 944034.00 frames. +2023-04-02 22:26:46,898 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 22:27:13,636 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=149409.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:27:48,670 INFO [train.py:903] (3/4) Epoch 22, batch 6050, loss[loss=0.205, simple_loss=0.2909, pruned_loss=0.05953, over 19301.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2885, pruned_loss=0.06439, over 3806831.98 frames. ], batch size: 66, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:27:49,811 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.518e+02 4.880e+02 5.766e+02 7.280e+02 1.810e+03, threshold=1.153e+03, percent-clipped=3.0 +2023-04-02 22:28:02,676 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=149449.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:28:15,554 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-04-02 22:28:32,630 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=149474.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:28:50,321 INFO [train.py:903] (3/4) Epoch 22, batch 6100, loss[loss=0.2168, simple_loss=0.3045, pruned_loss=0.06455, over 19482.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.2885, pruned_loss=0.06485, over 3811353.70 frames. ], batch size: 64, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:29:42,734 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=149532.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:29:48,979 INFO [train.py:903] (3/4) Epoch 22, batch 6150, loss[loss=0.2515, simple_loss=0.3308, pruned_loss=0.08608, over 19326.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2881, pruned_loss=0.06458, over 3817316.28 frames. ], batch size: 66, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:29:50,021 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.191e+02 4.761e+02 6.063e+02 7.648e+02 1.908e+03, threshold=1.213e+03, percent-clipped=8.0 +2023-04-02 22:30:19,809 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-02 22:30:44,104 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=149583.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:30:49,421 INFO [train.py:903] (3/4) Epoch 22, batch 6200, loss[loss=0.1826, simple_loss=0.2699, pruned_loss=0.04759, over 19753.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2882, pruned_loss=0.06422, over 3815773.69 frames. ], batch size: 51, lr: 3.71e-03, grad_scale: 8.0 +2023-04-02 22:31:24,694 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-02 22:31:36,837 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-02 22:31:51,302 INFO [train.py:903] (3/4) Epoch 22, batch 6250, loss[loss=0.181, simple_loss=0.2724, pruned_loss=0.04483, over 19481.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2876, pruned_loss=0.06378, over 3825179.41 frames. ], batch size: 64, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:31:52,392 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.896e+02 5.068e+02 6.178e+02 8.268e+02 1.694e+03, threshold=1.236e+03, percent-clipped=5.0 +2023-04-02 22:32:21,863 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-02 22:32:39,423 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1141, 1.9085, 1.8197, 2.9487, 2.1831, 2.3449, 2.5610, 2.1813], + device='cuda:3'), covar=tensor([0.0833, 0.0905, 0.0967, 0.0676, 0.0765, 0.0732, 0.0832, 0.0651], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0221, 0.0223, 0.0239, 0.0226, 0.0212, 0.0187, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 22:32:52,474 INFO [train.py:903] (3/4) Epoch 22, batch 6300, loss[loss=0.1904, simple_loss=0.2824, pruned_loss=0.04917, over 19661.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2867, pruned_loss=0.06324, over 3814780.35 frames. ], batch size: 55, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:33:51,828 INFO [train.py:903] (3/4) Epoch 22, batch 6350, loss[loss=0.1855, simple_loss=0.2653, pruned_loss=0.05287, over 19606.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2878, pruned_loss=0.06367, over 3826188.65 frames. ], batch size: 50, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:33:52,935 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.101e+02 5.551e+02 6.531e+02 8.044e+02 1.579e+03, threshold=1.306e+03, percent-clipped=6.0 +2023-04-02 22:34:23,954 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=149764.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 22:34:37,667 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8065, 1.7255, 2.0176, 1.7762, 4.3571, 1.0041, 2.5075, 4.6803], + device='cuda:3'), covar=tensor([0.0478, 0.2646, 0.2606, 0.1864, 0.0742, 0.2910, 0.1603, 0.0198], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0368, 0.0388, 0.0347, 0.0375, 0.0353, 0.0383, 0.0405], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:34:52,501 INFO [train.py:903] (3/4) Epoch 22, batch 6400, loss[loss=0.2474, simple_loss=0.3128, pruned_loss=0.09102, over 13538.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2876, pruned_loss=0.06359, over 3819958.67 frames. ], batch size: 137, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:35:37,226 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9911, 3.6483, 2.5816, 3.3157, 1.0163, 3.5940, 3.4577, 3.5763], + device='cuda:3'), covar=tensor([0.0779, 0.1166, 0.1827, 0.0935, 0.3683, 0.0777, 0.1014, 0.1277], + device='cuda:3'), in_proj_covar=tensor([0.0499, 0.0408, 0.0493, 0.0343, 0.0398, 0.0429, 0.0422, 0.0457], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:35:54,292 INFO [train.py:903] (3/4) Epoch 22, batch 6450, loss[loss=0.1786, simple_loss=0.2537, pruned_loss=0.05176, over 19811.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2874, pruned_loss=0.06356, over 3807003.22 frames. ], batch size: 48, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:35:55,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.826e+02 4.788e+02 5.699e+02 7.070e+02 1.580e+03, threshold=1.140e+03, percent-clipped=2.0 +2023-04-02 22:36:39,138 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-02 22:36:40,519 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=149876.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:36:50,720 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9496, 2.0259, 2.2862, 2.6793, 1.9908, 2.6183, 2.2838, 2.0715], + device='cuda:3'), covar=tensor([0.4329, 0.4009, 0.1854, 0.2322, 0.4219, 0.2041, 0.4867, 0.3385], + device='cuda:3'), in_proj_covar=tensor([0.0899, 0.0967, 0.0719, 0.0932, 0.0880, 0.0818, 0.0841, 0.0782], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 22:36:54,745 INFO [train.py:903] (3/4) Epoch 22, batch 6500, loss[loss=0.2464, simple_loss=0.337, pruned_loss=0.07792, over 18551.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2879, pruned_loss=0.06336, over 3821444.14 frames. ], batch size: 84, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:37:00,219 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-02 22:37:02,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0328, 5.0658, 5.8306, 5.8409, 1.9227, 5.5189, 4.6491, 5.4729], + device='cuda:3'), covar=tensor([0.1628, 0.0965, 0.0548, 0.0622, 0.6280, 0.0816, 0.0625, 0.1163], + device='cuda:3'), in_proj_covar=tensor([0.0783, 0.0743, 0.0948, 0.0826, 0.0834, 0.0708, 0.0564, 0.0878], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 22:37:31,259 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-04-02 22:37:41,923 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=149927.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:37:55,093 INFO [train.py:903] (3/4) Epoch 22, batch 6550, loss[loss=0.1937, simple_loss=0.2642, pruned_loss=0.06162, over 19610.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2882, pruned_loss=0.06344, over 3825202.46 frames. ], batch size: 50, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:37:56,255 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.115e+02 4.654e+02 5.933e+02 7.304e+02 1.667e+03, threshold=1.187e+03, percent-clipped=4.0 +2023-04-02 22:38:08,356 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-04-02 22:38:55,831 INFO [train.py:903] (3/4) Epoch 22, batch 6600, loss[loss=0.1564, simple_loss=0.2368, pruned_loss=0.03806, over 19731.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2881, pruned_loss=0.06354, over 3804953.02 frames. ], batch size: 46, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:38:58,358 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7630, 4.3499, 2.7754, 3.7649, 1.1977, 4.2881, 4.1669, 4.3115], + device='cuda:3'), covar=tensor([0.0575, 0.0936, 0.1839, 0.0798, 0.3722, 0.0645, 0.0920, 0.1212], + device='cuda:3'), in_proj_covar=tensor([0.0501, 0.0409, 0.0493, 0.0344, 0.0399, 0.0431, 0.0423, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:38:59,630 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=149991.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:39:39,196 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2346, 1.2804, 1.6590, 1.2679, 2.6856, 3.5748, 3.2861, 3.7920], + device='cuda:3'), covar=tensor([0.1649, 0.3932, 0.3620, 0.2581, 0.0653, 0.0218, 0.0227, 0.0282], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0323, 0.0354, 0.0264, 0.0244, 0.0188, 0.0215, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 22:39:59,904 INFO [train.py:903] (3/4) Epoch 22, batch 6650, loss[loss=0.2303, simple_loss=0.3138, pruned_loss=0.0734, over 19785.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2876, pruned_loss=0.06348, over 3820753.37 frames. ], batch size: 56, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:40:01,056 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.517e+02 4.673e+02 5.867e+02 7.414e+02 1.313e+03, threshold=1.173e+03, percent-clipped=2.0 +2023-04-02 22:40:04,782 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=150042.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:40:59,358 INFO [train.py:903] (3/4) Epoch 22, batch 6700, loss[loss=0.1842, simple_loss=0.2567, pruned_loss=0.05585, over 19066.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2871, pruned_loss=0.06371, over 3807258.23 frames. ], batch size: 42, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:41:10,398 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=150097.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:41:23,665 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=150108.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 22:41:57,254 INFO [train.py:903] (3/4) Epoch 22, batch 6750, loss[loss=0.1932, simple_loss=0.2802, pruned_loss=0.05312, over 19670.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2871, pruned_loss=0.06395, over 3795064.39 frames. ], batch size: 55, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:41:58,371 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.145e+02 4.879e+02 6.504e+02 7.654e+02 1.720e+03, threshold=1.301e+03, percent-clipped=5.0 +2023-04-02 22:42:53,123 INFO [train.py:903] (3/4) Epoch 22, batch 6800, loss[loss=0.2143, simple_loss=0.2873, pruned_loss=0.07065, over 19713.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2872, pruned_loss=0.06401, over 3799168.33 frames. ], batch size: 51, lr: 3.70e-03, grad_scale: 8.0 +2023-04-02 22:43:38,162 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-02 22:43:39,167 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-02 22:43:42,617 INFO [train.py:903] (3/4) Epoch 23, batch 0, loss[loss=0.234, simple_loss=0.3035, pruned_loss=0.08231, over 13122.00 frames. ], tot_loss[loss=0.234, simple_loss=0.3035, pruned_loss=0.08231, over 13122.00 frames. ], batch size: 137, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:43:42,617 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 22:43:54,249 INFO [train.py:937] (3/4) Epoch 23, validation: loss=0.1688, simple_loss=0.2693, pruned_loss=0.03418, over 944034.00 frames. +2023-04-02 22:43:54,250 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 22:44:03,534 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=150223.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 22:44:06,612 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-02 22:44:21,434 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.019e+02 4.848e+02 5.561e+02 7.527e+02 1.735e+03, threshold=1.112e+03, percent-clipped=5.0 +2023-04-02 22:44:31,829 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=150247.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:44:35,195 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6047, 2.2512, 1.6885, 1.4925, 2.0910, 1.3078, 1.4050, 1.9683], + device='cuda:3'), covar=tensor([0.1127, 0.0849, 0.1219, 0.0929, 0.0627, 0.1390, 0.0854, 0.0611], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0320, 0.0342, 0.0269, 0.0250, 0.0339, 0.0294, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:44:55,826 INFO [train.py:903] (3/4) Epoch 23, batch 50, loss[loss=0.2187, simple_loss=0.2973, pruned_loss=0.07005, over 18842.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2873, pruned_loss=0.06226, over 860254.24 frames. ], batch size: 74, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:45:03,036 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=150272.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:45:27,192 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=150292.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:45:30,274 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-02 22:45:36,373 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=150298.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:45:57,952 INFO [train.py:903] (3/4) Epoch 23, batch 100, loss[loss=0.191, simple_loss=0.2768, pruned_loss=0.05266, over 19375.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2874, pruned_loss=0.06235, over 1507317.57 frames. ], batch size: 70, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:46:06,469 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=150323.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:46:07,254 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-02 22:46:26,514 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.179e+02 4.910e+02 5.630e+02 7.676e+02 1.557e+03, threshold=1.126e+03, percent-clipped=7.0 +2023-04-02 22:46:48,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5206, 2.2085, 1.6768, 1.3289, 2.0633, 1.2834, 1.3179, 1.9937], + device='cuda:3'), covar=tensor([0.1219, 0.0816, 0.1141, 0.1035, 0.0617, 0.1325, 0.0881, 0.0513], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0318, 0.0340, 0.0268, 0.0248, 0.0337, 0.0292, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:46:59,544 INFO [train.py:903] (3/4) Epoch 23, batch 150, loss[loss=0.2046, simple_loss=0.2699, pruned_loss=0.06958, over 19758.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2853, pruned_loss=0.06182, over 2016873.09 frames. ], batch size: 46, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:47:59,879 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-02 22:48:01,005 INFO [train.py:903] (3/4) Epoch 23, batch 200, loss[loss=0.2589, simple_loss=0.3371, pruned_loss=0.0904, over 19671.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2878, pruned_loss=0.06379, over 2400796.42 frames. ], batch size: 55, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:48:30,850 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.764e+02 5.350e+02 6.653e+02 9.712e+02 2.771e+03, threshold=1.331e+03, percent-clipped=16.0 +2023-04-02 22:48:33,188 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=150441.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:48:38,323 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 22:49:02,669 INFO [train.py:903] (3/4) Epoch 23, batch 250, loss[loss=0.2526, simple_loss=0.3251, pruned_loss=0.09003, over 18871.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2878, pruned_loss=0.06425, over 2715303.97 frames. ], batch size: 74, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:49:20,036 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=150479.0, num_to_drop=1, layers_to_drop={1} +2023-04-02 22:49:43,865 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2310, 1.2711, 1.6010, 1.3064, 2.6411, 3.7415, 3.4543, 4.0282], + device='cuda:3'), covar=tensor([0.1741, 0.4016, 0.3759, 0.2516, 0.0661, 0.0181, 0.0217, 0.0232], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0325, 0.0355, 0.0267, 0.0246, 0.0189, 0.0217, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 22:49:48,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0615, 4.3925, 4.7492, 4.7426, 1.8071, 4.4793, 3.8963, 4.4410], + device='cuda:3'), covar=tensor([0.1577, 0.1057, 0.0615, 0.0673, 0.6124, 0.0894, 0.0671, 0.1185], + device='cuda:3'), in_proj_covar=tensor([0.0784, 0.0744, 0.0952, 0.0832, 0.0836, 0.0711, 0.0566, 0.0878], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 22:49:49,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=150504.0, num_to_drop=1, layers_to_drop={0} +2023-04-02 22:49:54,030 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5718, 1.6007, 1.8261, 1.8069, 2.6546, 2.3184, 2.7413, 1.2392], + device='cuda:3'), covar=tensor([0.2333, 0.4130, 0.2581, 0.1860, 0.1357, 0.2068, 0.1282, 0.4180], + device='cuda:3'), in_proj_covar=tensor([0.0535, 0.0642, 0.0715, 0.0484, 0.0618, 0.0530, 0.0661, 0.0548], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 22:50:05,853 INFO [train.py:903] (3/4) Epoch 23, batch 300, loss[loss=0.1771, simple_loss=0.2532, pruned_loss=0.05051, over 19778.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2864, pruned_loss=0.06338, over 2975877.00 frames. ], batch size: 48, lr: 3.61e-03, grad_scale: 4.0 +2023-04-02 22:50:34,498 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.361e+02 5.024e+02 5.928e+02 7.198e+02 2.066e+03, threshold=1.186e+03, percent-clipped=3.0 +2023-04-02 22:50:54,909 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=150555.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:50:56,043 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=150556.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:51:07,074 INFO [train.py:903] (3/4) Epoch 23, batch 350, loss[loss=0.2135, simple_loss=0.2914, pruned_loss=0.0678, over 14034.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2869, pruned_loss=0.06347, over 3160387.33 frames. ], batch size: 135, lr: 3.61e-03, grad_scale: 4.0 +2023-04-02 22:51:11,929 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-02 22:52:09,603 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.76 vs. limit=2.0 +2023-04-02 22:52:09,973 INFO [train.py:903] (3/4) Epoch 23, batch 400, loss[loss=0.1999, simple_loss=0.2736, pruned_loss=0.0631, over 19383.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2861, pruned_loss=0.0632, over 3304678.88 frames. ], batch size: 47, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:52:36,602 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=150636.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:52:39,072 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2615, 2.3352, 2.6006, 3.2460, 2.3516, 3.0068, 2.6072, 2.3661], + device='cuda:3'), covar=tensor([0.4484, 0.4460, 0.1938, 0.2600, 0.4729, 0.2270, 0.5002, 0.3392], + device='cuda:3'), in_proj_covar=tensor([0.0903, 0.0968, 0.0719, 0.0932, 0.0883, 0.0819, 0.0845, 0.0784], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 22:52:40,896 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.962e+02 5.093e+02 6.486e+02 8.008e+02 1.724e+03, threshold=1.297e+03, percent-clipped=3.0 +2023-04-02 22:53:11,936 INFO [train.py:903] (3/4) Epoch 23, batch 450, loss[loss=0.2375, simple_loss=0.3111, pruned_loss=0.08197, over 19313.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.285, pruned_loss=0.0624, over 3437222.52 frames. ], batch size: 66, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:53:46,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-02 22:53:46,089 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-02 22:54:15,722 INFO [train.py:903] (3/4) Epoch 23, batch 500, loss[loss=0.2214, simple_loss=0.2845, pruned_loss=0.07912, over 18652.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2859, pruned_loss=0.06272, over 3524266.15 frames. ], batch size: 41, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:54:29,756 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4105, 4.0348, 2.6463, 3.6052, 1.2622, 3.9412, 3.8981, 3.9610], + device='cuda:3'), covar=tensor([0.0671, 0.1078, 0.2018, 0.0881, 0.3637, 0.0653, 0.0888, 0.1098], + device='cuda:3'), in_proj_covar=tensor([0.0505, 0.0413, 0.0498, 0.0346, 0.0402, 0.0436, 0.0425, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 22:54:45,185 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.118e+02 5.191e+02 6.635e+02 8.528e+02 2.142e+03, threshold=1.327e+03, percent-clipped=5.0 +2023-04-02 22:54:57,354 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=150750.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:54:58,668 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=150751.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:55:17,409 INFO [train.py:903] (3/4) Epoch 23, batch 550, loss[loss=0.2046, simple_loss=0.2735, pruned_loss=0.06783, over 19761.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2871, pruned_loss=0.06356, over 3587797.73 frames. ], batch size: 47, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:55:40,337 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4251, 1.5261, 1.8241, 1.7061, 2.6029, 2.1986, 2.7981, 1.2864], + device='cuda:3'), covar=tensor([0.2665, 0.4414, 0.2796, 0.2069, 0.1606, 0.2344, 0.1572, 0.4467], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0642, 0.0711, 0.0483, 0.0616, 0.0531, 0.0660, 0.0547], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 22:56:14,493 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=150812.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:56:18,580 INFO [train.py:903] (3/4) Epoch 23, batch 600, loss[loss=0.1609, simple_loss=0.2369, pruned_loss=0.04244, over 19715.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2888, pruned_loss=0.06445, over 3637020.19 frames. ], batch size: 46, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:56:45,484 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=150837.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:56:48,650 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.714e+02 5.000e+02 5.859e+02 6.998e+02 1.831e+03, threshold=1.172e+03, percent-clipped=2.0 +2023-04-02 22:56:59,194 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-02 22:57:21,218 INFO [train.py:903] (3/4) Epoch 23, batch 650, loss[loss=0.2239, simple_loss=0.3087, pruned_loss=0.06955, over 19527.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2877, pruned_loss=0.06389, over 3686864.00 frames. ], batch size: 64, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:58:02,426 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=150899.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:58:23,647 INFO [train.py:903] (3/4) Epoch 23, batch 700, loss[loss=0.1705, simple_loss=0.2572, pruned_loss=0.0419, over 19484.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2867, pruned_loss=0.0633, over 3731744.40 frames. ], batch size: 49, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 22:58:27,253 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=150918.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 22:58:52,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.226e+02 5.019e+02 5.794e+02 7.164e+02 1.349e+03, threshold=1.159e+03, percent-clipped=1.0 +2023-04-02 22:58:58,122 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.74 vs. limit=5.0 +2023-04-02 22:59:26,031 INFO [train.py:903] (3/4) Epoch 23, batch 750, loss[loss=0.1854, simple_loss=0.2723, pruned_loss=0.04929, over 19564.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2866, pruned_loss=0.0633, over 3764414.05 frames. ], batch size: 61, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 23:00:17,629 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=151007.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:00:26,934 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151014.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:00:28,669 INFO [train.py:903] (3/4) Epoch 23, batch 800, loss[loss=0.2005, simple_loss=0.2893, pruned_loss=0.05589, over 19542.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2859, pruned_loss=0.06298, over 3787608.17 frames. ], batch size: 54, lr: 3.61e-03, grad_scale: 8.0 +2023-04-02 23:00:36,353 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-02 23:00:44,076 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-02 23:00:46,696 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-02 23:00:48,201 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=151032.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:00:57,135 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 5.089e+02 6.161e+02 7.478e+02 1.780e+03, threshold=1.232e+03, percent-clipped=6.0 +2023-04-02 23:01:29,773 INFO [train.py:903] (3/4) Epoch 23, batch 850, loss[loss=0.1683, simple_loss=0.2437, pruned_loss=0.04647, over 19787.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2851, pruned_loss=0.06279, over 3803136.83 frames. ], batch size: 48, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:01:38,466 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-02 23:02:00,721 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5362, 1.7393, 2.0063, 1.8015, 3.3376, 2.6968, 3.6206, 1.7080], + device='cuda:3'), covar=tensor([0.2419, 0.4099, 0.2529, 0.1873, 0.1347, 0.1985, 0.1406, 0.3977], + device='cuda:3'), in_proj_covar=tensor([0.0531, 0.0638, 0.0707, 0.0481, 0.0611, 0.0526, 0.0653, 0.0545], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:02:04,874 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=151094.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:02:25,394 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-02 23:02:31,810 INFO [train.py:903] (3/4) Epoch 23, batch 900, loss[loss=0.1842, simple_loss=0.2525, pruned_loss=0.05793, over 19731.00 frames. ], tot_loss[loss=0.205, simple_loss=0.285, pruned_loss=0.06253, over 3816937.18 frames. ], batch size: 45, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:03:02,018 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.084e+02 5.070e+02 6.543e+02 8.443e+02 1.332e+03, threshold=1.309e+03, percent-clipped=3.0 +2023-04-02 23:03:21,217 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-04-02 23:03:32,679 INFO [train.py:903] (3/4) Epoch 23, batch 950, loss[loss=0.2153, simple_loss=0.2888, pruned_loss=0.07095, over 19864.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2866, pruned_loss=0.06396, over 3823626.42 frames. ], batch size: 52, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:03:39,535 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-02 23:04:11,660 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6739, 4.2922, 2.7516, 3.7969, 1.0751, 4.1681, 4.0675, 4.1594], + device='cuda:3'), covar=tensor([0.0579, 0.0894, 0.1788, 0.0877, 0.3731, 0.0661, 0.0839, 0.0981], + device='cuda:3'), in_proj_covar=tensor([0.0504, 0.0413, 0.0497, 0.0346, 0.0400, 0.0435, 0.0426, 0.0459], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:04:16,137 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-02 23:04:17,808 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=151202.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:04:27,639 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151209.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:04:35,177 INFO [train.py:903] (3/4) Epoch 23, batch 1000, loss[loss=0.1857, simple_loss=0.2731, pruned_loss=0.04911, over 19731.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2865, pruned_loss=0.06405, over 3803965.56 frames. ], batch size: 51, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:04:47,465 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4960, 1.5076, 1.6720, 1.6683, 2.3659, 2.1656, 2.4656, 0.9360], + device='cuda:3'), covar=tensor([0.2423, 0.4259, 0.2722, 0.1977, 0.1436, 0.2141, 0.1284, 0.4552], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0642, 0.0711, 0.0484, 0.0616, 0.0528, 0.0658, 0.0548], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:05:04,890 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.333e+02 5.146e+02 6.401e+02 7.951e+02 1.702e+03, threshold=1.280e+03, percent-clipped=4.0 +2023-04-02 23:05:32,304 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-02 23:05:33,597 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=151262.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:05:37,845 INFO [train.py:903] (3/4) Epoch 23, batch 1050, loss[loss=0.1889, simple_loss=0.265, pruned_loss=0.05637, over 19610.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2871, pruned_loss=0.06413, over 3799787.83 frames. ], batch size: 50, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:05:42,845 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=151270.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:06:12,776 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-02 23:06:14,326 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=151295.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:06:21,313 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=151300.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:06:40,096 INFO [train.py:903] (3/4) Epoch 23, batch 1100, loss[loss=0.1883, simple_loss=0.2694, pruned_loss=0.0536, over 19748.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2881, pruned_loss=0.0643, over 3789647.10 frames. ], batch size: 51, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:06:40,407 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=151316.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:07:09,139 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.937e+02 5.070e+02 6.152e+02 7.617e+02 1.362e+03, threshold=1.230e+03, percent-clipped=2.0 +2023-04-02 23:07:40,855 INFO [train.py:903] (3/4) Epoch 23, batch 1150, loss[loss=0.2106, simple_loss=0.2797, pruned_loss=0.07071, over 19298.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.288, pruned_loss=0.06466, over 3792071.32 frames. ], batch size: 44, lr: 3.60e-03, grad_scale: 4.0 +2023-04-02 23:07:55,726 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151377.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:08:43,929 INFO [train.py:903] (3/4) Epoch 23, batch 1200, loss[loss=0.2029, simple_loss=0.2952, pruned_loss=0.05527, over 19768.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2877, pruned_loss=0.06437, over 3787032.02 frames. ], batch size: 54, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:09:14,770 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.106e+02 4.896e+02 6.001e+02 7.643e+02 1.247e+03, threshold=1.200e+03, percent-clipped=2.0 +2023-04-02 23:09:18,077 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-02 23:09:45,312 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=151465.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:09:46,023 INFO [train.py:903] (3/4) Epoch 23, batch 1250, loss[loss=0.1914, simple_loss=0.2708, pruned_loss=0.05596, over 19406.00 frames. ], tot_loss[loss=0.2097, simple_loss=0.2893, pruned_loss=0.06502, over 3795638.24 frames. ], batch size: 48, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:10:12,468 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3758, 2.1436, 1.6440, 1.4240, 1.9864, 1.2471, 1.2676, 1.8084], + device='cuda:3'), covar=tensor([0.1110, 0.0798, 0.1080, 0.0926, 0.0534, 0.1353, 0.0818, 0.0535], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0316, 0.0339, 0.0266, 0.0246, 0.0339, 0.0291, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:10:16,529 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=151490.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:10:17,590 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1965, 2.3076, 2.5267, 3.1308, 2.3377, 2.9777, 2.6713, 2.4063], + device='cuda:3'), covar=tensor([0.4275, 0.4088, 0.1959, 0.2517, 0.4475, 0.2134, 0.4474, 0.3296], + device='cuda:3'), in_proj_covar=tensor([0.0906, 0.0973, 0.0719, 0.0936, 0.0884, 0.0817, 0.0844, 0.0787], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 23:10:46,707 INFO [train.py:903] (3/4) Epoch 23, batch 1300, loss[loss=0.219, simple_loss=0.3047, pruned_loss=0.06665, over 19589.00 frames. ], tot_loss[loss=0.2098, simple_loss=0.2892, pruned_loss=0.06521, over 3810051.74 frames. ], batch size: 57, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:11:16,030 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.826e+02 5.118e+02 6.046e+02 8.032e+02 1.744e+03, threshold=1.209e+03, percent-clipped=5.0 +2023-04-02 23:11:22,856 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=151546.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:11:46,631 INFO [train.py:903] (3/4) Epoch 23, batch 1350, loss[loss=0.2135, simple_loss=0.2907, pruned_loss=0.06813, over 13500.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2891, pruned_loss=0.06479, over 3821619.05 frames. ], batch size: 136, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:12:48,225 INFO [train.py:903] (3/4) Epoch 23, batch 1400, loss[loss=0.1641, simple_loss=0.2453, pruned_loss=0.04145, over 19762.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2891, pruned_loss=0.06478, over 3811164.98 frames. ], batch size: 48, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:13:08,456 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=151633.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:13:13,630 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.4626, 4.9954, 3.0490, 4.3707, 1.3234, 4.9435, 4.8508, 5.0139], + device='cuda:3'), covar=tensor([0.0420, 0.0851, 0.1913, 0.0755, 0.3885, 0.0587, 0.0824, 0.1233], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0413, 0.0498, 0.0347, 0.0402, 0.0437, 0.0427, 0.0459], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:13:17,802 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.750e+02 5.043e+02 6.240e+02 8.237e+02 1.280e+03, threshold=1.248e+03, percent-clipped=3.0 +2023-04-02 23:13:21,407 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=151644.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:13:38,327 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=151658.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:13:40,485 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=151660.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:13:41,886 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151661.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:13:48,349 INFO [train.py:903] (3/4) Epoch 23, batch 1450, loss[loss=0.1983, simple_loss=0.269, pruned_loss=0.06382, over 19384.00 frames. ], tot_loss[loss=0.2086, simple_loss=0.2886, pruned_loss=0.06426, over 3809636.33 frames. ], batch size: 47, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:13:48,384 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-02 23:14:36,967 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=151706.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:14:49,055 INFO [train.py:903] (3/4) Epoch 23, batch 1500, loss[loss=0.198, simple_loss=0.2868, pruned_loss=0.05459, over 19691.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2886, pruned_loss=0.06443, over 3805390.35 frames. ], batch size: 59, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:15:18,488 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.327e+02 4.905e+02 6.054e+02 7.299e+02 2.065e+03, threshold=1.211e+03, percent-clipped=4.0 +2023-04-02 23:15:33,010 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5932, 1.8005, 2.0241, 1.8586, 3.2330, 2.6424, 3.6256, 1.6330], + device='cuda:3'), covar=tensor([0.2534, 0.4309, 0.2857, 0.1949, 0.1513, 0.2132, 0.1514, 0.4320], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0647, 0.0717, 0.0488, 0.0620, 0.0533, 0.0665, 0.0554], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 23:15:40,400 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151759.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:15:47,607 INFO [train.py:903] (3/4) Epoch 23, batch 1550, loss[loss=0.2065, simple_loss=0.2914, pruned_loss=0.06083, over 19478.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.289, pruned_loss=0.06479, over 3805939.75 frames. ], batch size: 64, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:15:59,944 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=151775.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:16:45,623 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3489, 4.0079, 3.0726, 3.5596, 1.8616, 3.8765, 3.8137, 3.8702], + device='cuda:3'), covar=tensor([0.0652, 0.0868, 0.1832, 0.0892, 0.2745, 0.0750, 0.0896, 0.1254], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0411, 0.0497, 0.0347, 0.0401, 0.0436, 0.0428, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:16:50,043 INFO [train.py:903] (3/4) Epoch 23, batch 1600, loss[loss=0.1726, simple_loss=0.2496, pruned_loss=0.04784, over 19752.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2877, pruned_loss=0.0642, over 3800836.74 frames. ], batch size: 46, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:16:53,651 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=151819.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:17:10,289 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-02 23:17:20,214 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.366e+02 4.863e+02 5.887e+02 6.951e+02 2.426e+03, threshold=1.177e+03, percent-clipped=3.0 +2023-04-02 23:17:50,192 INFO [train.py:903] (3/4) Epoch 23, batch 1650, loss[loss=0.2104, simple_loss=0.2975, pruned_loss=0.06163, over 19541.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2872, pruned_loss=0.06379, over 3801957.91 frames. ], batch size: 64, lr: 3.60e-03, grad_scale: 8.0 +2023-04-02 23:18:51,789 INFO [train.py:903] (3/4) Epoch 23, batch 1700, loss[loss=0.1905, simple_loss=0.2613, pruned_loss=0.0599, over 18616.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.287, pruned_loss=0.06393, over 3805529.77 frames. ], batch size: 41, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:18:53,411 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=151917.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:19:21,472 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.136e+02 5.018e+02 6.073e+02 7.613e+02 1.748e+03, threshold=1.215e+03, percent-clipped=5.0 +2023-04-02 23:19:23,213 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=151942.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:19:26,187 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-02 23:19:42,954 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8123, 1.8241, 1.4190, 1.8512, 1.8464, 1.4541, 1.5951, 1.6677], + device='cuda:3'), covar=tensor([0.1348, 0.1679, 0.2131, 0.1369, 0.1531, 0.1042, 0.1923, 0.1123], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0359, 0.0318, 0.0255, 0.0307, 0.0255, 0.0313, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:19:52,380 INFO [train.py:903] (3/4) Epoch 23, batch 1750, loss[loss=0.1757, simple_loss=0.2502, pruned_loss=0.05062, over 19735.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2865, pruned_loss=0.06304, over 3813851.97 frames. ], batch size: 45, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:20:41,352 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6242, 4.2041, 2.5609, 3.7070, 0.8854, 4.1514, 4.0251, 4.1447], + device='cuda:3'), covar=tensor([0.0646, 0.1079, 0.2225, 0.0919, 0.4226, 0.0714, 0.0918, 0.1218], + device='cuda:3'), in_proj_covar=tensor([0.0508, 0.0412, 0.0497, 0.0347, 0.0400, 0.0435, 0.0429, 0.0458], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:20:53,710 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=152015.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:20:54,415 INFO [train.py:903] (3/4) Epoch 23, batch 1800, loss[loss=0.2095, simple_loss=0.293, pruned_loss=0.06302, over 18269.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2865, pruned_loss=0.06296, over 3825319.13 frames. ], batch size: 83, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:21:13,149 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=152031.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:21:24,577 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=152040.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:21:26,139 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.69 vs. limit=5.0 +2023-04-02 23:21:26,672 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.151e+02 5.200e+02 6.601e+02 8.823e+02 1.720e+03, threshold=1.320e+03, percent-clipped=12.0 +2023-04-02 23:21:36,343 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152050.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:21:44,277 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=152056.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:21:48,442 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-02 23:21:55,077 INFO [train.py:903] (3/4) Epoch 23, batch 1850, loss[loss=0.1735, simple_loss=0.2467, pruned_loss=0.05015, over 19271.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2869, pruned_loss=0.06293, over 3823831.20 frames. ], batch size: 44, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:22:26,617 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-02 23:22:54,488 INFO [train.py:903] (3/4) Epoch 23, batch 1900, loss[loss=0.2072, simple_loss=0.2904, pruned_loss=0.06198, over 19608.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2875, pruned_loss=0.06345, over 3830022.48 frames. ], batch size: 52, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:23:09,868 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-02 23:23:16,386 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-02 23:23:26,803 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.505e+02 4.994e+02 5.968e+02 7.712e+02 2.482e+03, threshold=1.194e+03, percent-clipped=3.0 +2023-04-02 23:23:41,406 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-02 23:23:52,773 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152163.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:23:55,391 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=152165.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:23:56,159 INFO [train.py:903] (3/4) Epoch 23, batch 1950, loss[loss=0.2192, simple_loss=0.2964, pruned_loss=0.07099, over 19655.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2866, pruned_loss=0.06294, over 3832962.24 frames. ], batch size: 60, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:24:58,528 INFO [train.py:903] (3/4) Epoch 23, batch 2000, loss[loss=0.2393, simple_loss=0.3262, pruned_loss=0.07621, over 19404.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.286, pruned_loss=0.06225, over 3839695.33 frames. ], batch size: 70, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:25:28,732 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.131e+02 4.661e+02 5.613e+02 7.041e+02 1.127e+03, threshold=1.123e+03, percent-clipped=0.0 +2023-04-02 23:25:31,156 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152243.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:25:53,679 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-02 23:25:58,369 INFO [train.py:903] (3/4) Epoch 23, batch 2050, loss[loss=0.2581, simple_loss=0.3216, pruned_loss=0.09732, over 12924.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2856, pruned_loss=0.06239, over 3833431.43 frames. ], batch size: 137, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:26:13,093 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-02 23:26:13,415 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=152278.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:26:14,188 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-02 23:26:29,852 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152291.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:26:36,226 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-02 23:26:58,783 INFO [train.py:903] (3/4) Epoch 23, batch 2100, loss[loss=0.1909, simple_loss=0.2722, pruned_loss=0.05476, over 19769.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2852, pruned_loss=0.06215, over 3826766.64 frames. ], batch size: 56, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:27:09,415 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5349, 1.4949, 1.4791, 2.0093, 1.6679, 1.8446, 1.9358, 1.6810], + device='cuda:3'), covar=tensor([0.0864, 0.0959, 0.1047, 0.0720, 0.0808, 0.0791, 0.0784, 0.0743], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0223, 0.0239, 0.0226, 0.0212, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-02 23:27:27,818 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-02 23:27:30,769 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-02 23:27:31,218 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.686e+02 4.808e+02 5.705e+02 7.050e+02 1.568e+03, threshold=1.141e+03, percent-clipped=6.0 +2023-04-02 23:27:48,004 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-02 23:27:59,303 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.68 vs. limit=5.0 +2023-04-02 23:27:59,782 INFO [train.py:903] (3/4) Epoch 23, batch 2150, loss[loss=0.2405, simple_loss=0.3192, pruned_loss=0.08083, over 19661.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2859, pruned_loss=0.06267, over 3825995.66 frames. ], batch size: 53, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:28:37,109 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-02 23:28:45,738 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8632, 1.8221, 1.6939, 1.5476, 1.5227, 1.5935, 0.6109, 0.9875], + device='cuda:3'), covar=tensor([0.0582, 0.0621, 0.0436, 0.0639, 0.0959, 0.0706, 0.1196, 0.0943], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0352, 0.0359, 0.0381, 0.0458, 0.0387, 0.0335, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:29:00,654 INFO [train.py:903] (3/4) Epoch 23, batch 2200, loss[loss=0.2767, simple_loss=0.3385, pruned_loss=0.1075, over 18734.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2869, pruned_loss=0.06334, over 3807063.85 frames. ], batch size: 74, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:29:07,526 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=152421.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:29:17,919 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4971, 1.6399, 1.9773, 1.7753, 3.1305, 2.6139, 3.3335, 1.7317], + device='cuda:3'), covar=tensor([0.2620, 0.4348, 0.2727, 0.1980, 0.1611, 0.2122, 0.1676, 0.4165], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0643, 0.0714, 0.0486, 0.0617, 0.0532, 0.0662, 0.0549], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:29:28,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2558, 2.9733, 2.2384, 2.3217, 2.2173, 2.6361, 1.1018, 2.1044], + device='cuda:3'), covar=tensor([0.0568, 0.0597, 0.0720, 0.1076, 0.1097, 0.0974, 0.1410, 0.1095], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0353, 0.0360, 0.0382, 0.0460, 0.0388, 0.0336, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:29:31,740 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.170e+02 5.084e+02 6.029e+02 8.181e+02 1.825e+03, threshold=1.206e+03, percent-clipped=10.0 +2023-04-02 23:29:37,618 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=152446.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:29:58,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152463.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:30:01,457 INFO [train.py:903] (3/4) Epoch 23, batch 2250, loss[loss=0.2724, simple_loss=0.3418, pruned_loss=0.1015, over 19735.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2865, pruned_loss=0.06305, over 3815068.03 frames. ], batch size: 63, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:31:01,770 INFO [train.py:903] (3/4) Epoch 23, batch 2300, loss[loss=0.2177, simple_loss=0.2958, pruned_loss=0.06981, over 19698.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2867, pruned_loss=0.06343, over 3829985.30 frames. ], batch size: 59, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:31:17,225 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-02 23:31:25,345 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=152534.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:31:36,144 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.429e+02 4.927e+02 5.902e+02 7.617e+02 2.113e+03, threshold=1.180e+03, percent-clipped=5.0 +2023-04-02 23:31:41,228 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0680, 3.1397, 1.9280, 1.8897, 2.8224, 1.7173, 1.4342, 2.2020], + device='cuda:3'), covar=tensor([0.1329, 0.0689, 0.1106, 0.0916, 0.0581, 0.1287, 0.1033, 0.0718], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0336, 0.0266, 0.0248, 0.0338, 0.0290, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:31:55,675 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=152559.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:32:04,337 INFO [train.py:903] (3/4) Epoch 23, batch 2350, loss[loss=0.2531, simple_loss=0.3141, pruned_loss=0.09603, over 13535.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2861, pruned_loss=0.06294, over 3823123.42 frames. ], batch size: 135, lr: 3.59e-03, grad_scale: 4.0 +2023-04-02 23:32:30,252 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152587.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:32:43,251 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-02 23:32:45,464 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152600.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:32:54,455 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152607.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:33:03,113 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-02 23:33:05,171 INFO [train.py:903] (3/4) Epoch 23, batch 2400, loss[loss=0.2354, simple_loss=0.3137, pruned_loss=0.07859, over 19832.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2868, pruned_loss=0.06324, over 3826866.14 frames. ], batch size: 52, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:33:28,191 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152635.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:33:38,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.633e+02 5.347e+02 6.485e+02 7.646e+02 1.871e+03, threshold=1.297e+03, percent-clipped=3.0 +2023-04-02 23:34:06,659 INFO [train.py:903] (3/4) Epoch 23, batch 2450, loss[loss=0.2081, simple_loss=0.2994, pruned_loss=0.0584, over 19525.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.286, pruned_loss=0.06267, over 3824312.41 frames. ], batch size: 54, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:34:51,455 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=152702.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:34:55,912 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1028, 5.0994, 5.8624, 5.8609, 2.1937, 5.5665, 4.7022, 5.4839], + device='cuda:3'), covar=tensor([0.1523, 0.0807, 0.0537, 0.0566, 0.5706, 0.0802, 0.0607, 0.1096], + device='cuda:3'), in_proj_covar=tensor([0.0781, 0.0744, 0.0952, 0.0836, 0.0838, 0.0714, 0.0564, 0.0886], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-02 23:35:06,561 INFO [train.py:903] (3/4) Epoch 23, batch 2500, loss[loss=0.2312, simple_loss=0.3118, pruned_loss=0.07532, over 19592.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2857, pruned_loss=0.0628, over 3830631.82 frames. ], batch size: 61, lr: 3.59e-03, grad_scale: 8.0 +2023-04-02 23:35:40,606 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.829e+02 4.968e+02 5.949e+02 7.714e+02 2.745e+03, threshold=1.190e+03, percent-clipped=5.0 +2023-04-02 23:35:42,101 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=152744.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:35:48,910 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2166, 1.8125, 1.9008, 1.8563, 2.9758, 1.5838, 2.5814, 3.2093], + device='cuda:3'), covar=tensor([0.0543, 0.2232, 0.2320, 0.1628, 0.0563, 0.2101, 0.1593, 0.0356], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0368, 0.0389, 0.0349, 0.0376, 0.0353, 0.0383, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:35:48,954 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=152750.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:36:08,241 INFO [train.py:903] (3/4) Epoch 23, batch 2550, loss[loss=0.1996, simple_loss=0.2782, pruned_loss=0.06056, over 19780.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2854, pruned_loss=0.06255, over 3819520.30 frames. ], batch size: 56, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:36:11,527 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-02 23:36:57,889 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152807.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:37:01,065 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-02 23:37:08,439 INFO [train.py:903] (3/4) Epoch 23, batch 2600, loss[loss=0.1685, simple_loss=0.2644, pruned_loss=0.03629, over 19531.00 frames. ], tot_loss[loss=0.2058, simple_loss=0.2865, pruned_loss=0.0626, over 3829639.07 frames. ], batch size: 54, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:37:40,421 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.218e+02 4.724e+02 5.510e+02 7.301e+02 1.657e+03, threshold=1.102e+03, percent-clipped=4.0 +2023-04-02 23:38:08,562 INFO [train.py:903] (3/4) Epoch 23, batch 2650, loss[loss=0.1992, simple_loss=0.2742, pruned_loss=0.06207, over 19413.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2873, pruned_loss=0.06319, over 3836788.77 frames. ], batch size: 48, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:38:27,770 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-02 23:39:08,680 INFO [train.py:903] (3/4) Epoch 23, batch 2700, loss[loss=0.2516, simple_loss=0.3279, pruned_loss=0.08766, over 19751.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2865, pruned_loss=0.06266, over 3844199.48 frames. ], batch size: 63, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:39:16,557 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=152922.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:39:42,349 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.879e+02 4.895e+02 5.730e+02 7.672e+02 1.465e+03, threshold=1.146e+03, percent-clipped=5.0 +2023-04-02 23:39:43,665 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152944.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:39:51,724 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=152951.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:40:00,060 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=152958.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:40:09,443 INFO [train.py:903] (3/4) Epoch 23, batch 2750, loss[loss=0.2433, simple_loss=0.3257, pruned_loss=0.08042, over 19683.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.287, pruned_loss=0.06289, over 3825861.24 frames. ], batch size: 60, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:40:31,190 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=152983.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:40:57,885 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153006.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:41:10,184 INFO [train.py:903] (3/4) Epoch 23, batch 2800, loss[loss=0.2057, simple_loss=0.2761, pruned_loss=0.06763, over 19612.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2856, pruned_loss=0.06235, over 3840500.89 frames. ], batch size: 50, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:41:27,948 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153031.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:41:42,353 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.106e+02 4.940e+02 6.140e+02 7.865e+02 1.529e+03, threshold=1.228e+03, percent-clipped=3.0 +2023-04-02 23:42:02,553 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153059.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:42:10,892 INFO [train.py:903] (3/4) Epoch 23, batch 2850, loss[loss=0.2237, simple_loss=0.2962, pruned_loss=0.07557, over 19742.00 frames. ], tot_loss[loss=0.206, simple_loss=0.286, pruned_loss=0.06301, over 3837403.91 frames. ], batch size: 51, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:42:11,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153066.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:42:36,416 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=153088.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:43:09,824 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-02 23:43:11,008 INFO [train.py:903] (3/4) Epoch 23, batch 2900, loss[loss=0.21, simple_loss=0.292, pruned_loss=0.06397, over 19658.00 frames. ], tot_loss[loss=0.2058, simple_loss=0.286, pruned_loss=0.06283, over 3839991.60 frames. ], batch size: 53, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:43:12,485 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0120, 2.9227, 1.8667, 1.9199, 2.6524, 1.6816, 1.4751, 2.1482], + device='cuda:3'), covar=tensor([0.1285, 0.0756, 0.1028, 0.0798, 0.0607, 0.1189, 0.0945, 0.0705], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0315, 0.0337, 0.0265, 0.0248, 0.0338, 0.0290, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:43:34,524 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=153135.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:43:44,842 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.40 vs. limit=5.0 +2023-04-02 23:43:45,170 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.545e+02 5.074e+02 6.117e+02 7.609e+02 1.538e+03, threshold=1.223e+03, percent-clipped=2.0 +2023-04-02 23:44:10,298 INFO [train.py:903] (3/4) Epoch 23, batch 2950, loss[loss=0.1788, simple_loss=0.2579, pruned_loss=0.04983, over 19485.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2875, pruned_loss=0.06364, over 3818064.66 frames. ], batch size: 49, lr: 3.58e-03, grad_scale: 4.0 +2023-04-02 23:44:13,932 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2899, 1.3018, 1.2655, 1.1492, 1.0155, 1.1559, 0.4168, 0.6556], + device='cuda:3'), covar=tensor([0.0495, 0.0522, 0.0326, 0.0513, 0.0899, 0.0612, 0.1157, 0.0831], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0354, 0.0360, 0.0382, 0.0462, 0.0388, 0.0336, 0.0338], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:44:23,317 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2616, 2.0557, 1.9897, 2.1989, 2.0776, 1.9677, 1.8607, 2.2165], + device='cuda:3'), covar=tensor([0.0953, 0.1372, 0.1357, 0.1007, 0.1232, 0.0509, 0.1367, 0.0663], + device='cuda:3'), in_proj_covar=tensor([0.0270, 0.0353, 0.0310, 0.0251, 0.0301, 0.0249, 0.0307, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:44:23,395 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5754, 1.6588, 1.9541, 1.8466, 2.8731, 2.5207, 2.9974, 1.4314], + device='cuda:3'), covar=tensor([0.2436, 0.4262, 0.2707, 0.1821, 0.1440, 0.2032, 0.1457, 0.4307], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0645, 0.0716, 0.0487, 0.0619, 0.0532, 0.0663, 0.0552], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:44:25,505 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153178.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:44:26,639 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2068, 1.2661, 1.2536, 1.0577, 1.0710, 1.0636, 0.1343, 0.3650], + device='cuda:3'), covar=tensor([0.0742, 0.0719, 0.0462, 0.0609, 0.1412, 0.0698, 0.1421, 0.1197], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0354, 0.0360, 0.0382, 0.0462, 0.0388, 0.0337, 0.0339], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-02 23:44:54,245 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153203.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:44:54,287 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153203.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:45:09,887 INFO [train.py:903] (3/4) Epoch 23, batch 3000, loss[loss=0.2071, simple_loss=0.2845, pruned_loss=0.06486, over 19687.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2873, pruned_loss=0.06368, over 3824834.89 frames. ], batch size: 53, lr: 3.58e-03, grad_scale: 4.0 +2023-04-02 23:45:09,887 INFO [train.py:928] (3/4) Computing validation loss +2023-04-02 23:45:23,387 INFO [train.py:937] (3/4) Epoch 23, validation: loss=0.1686, simple_loss=0.2685, pruned_loss=0.03441, over 944034.00 frames. +2023-04-02 23:45:23,388 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-02 23:45:26,708 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-02 23:45:57,202 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.963e+02 5.132e+02 6.544e+02 7.997e+02 1.730e+03, threshold=1.309e+03, percent-clipped=4.0 +2023-04-02 23:46:24,012 INFO [train.py:903] (3/4) Epoch 23, batch 3050, loss[loss=0.1835, simple_loss=0.2789, pruned_loss=0.04406, over 19536.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2881, pruned_loss=0.06413, over 3818640.39 frames. ], batch size: 56, lr: 3.58e-03, grad_scale: 4.0 +2023-04-02 23:47:00,331 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=153296.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:47:25,789 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153315.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:47:26,484 INFO [train.py:903] (3/4) Epoch 23, batch 3100, loss[loss=0.1948, simple_loss=0.282, pruned_loss=0.05385, over 19789.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.2889, pruned_loss=0.06404, over 3796531.93 frames. ], batch size: 56, lr: 3.58e-03, grad_scale: 4.0 +2023-04-02 23:47:33,614 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153322.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:47:54,375 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153340.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:47:56,825 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-02 23:47:59,270 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.375e+02 4.897e+02 6.414e+02 9.491e+02 6.432e+03, threshold=1.283e+03, percent-clipped=11.0 +2023-04-02 23:48:03,120 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153347.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:48:06,260 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9624, 1.6575, 1.5872, 1.8670, 1.5610, 1.6717, 1.5518, 1.8185], + device='cuda:3'), covar=tensor([0.1057, 0.1478, 0.1511, 0.1178, 0.1430, 0.0559, 0.1422, 0.0771], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0358, 0.0315, 0.0254, 0.0305, 0.0252, 0.0310, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:48:25,986 INFO [train.py:903] (3/4) Epoch 23, batch 3150, loss[loss=0.1973, simple_loss=0.2743, pruned_loss=0.06019, over 19573.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2885, pruned_loss=0.06371, over 3803202.97 frames. ], batch size: 52, lr: 3.58e-03, grad_scale: 4.0 +2023-04-02 23:48:54,105 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-02 23:48:59,701 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9834, 3.6374, 2.5662, 3.2142, 0.8176, 3.6053, 3.4298, 3.5200], + device='cuda:3'), covar=tensor([0.0826, 0.1047, 0.2057, 0.0983, 0.4187, 0.0799, 0.1006, 0.1391], + device='cuda:3'), in_proj_covar=tensor([0.0506, 0.0410, 0.0494, 0.0346, 0.0397, 0.0434, 0.0426, 0.0460], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-02 23:49:26,046 INFO [train.py:903] (3/4) Epoch 23, batch 3200, loss[loss=0.2176, simple_loss=0.3022, pruned_loss=0.06646, over 19783.00 frames. ], tot_loss[loss=0.2089, simple_loss=0.2893, pruned_loss=0.06425, over 3801396.30 frames. ], batch size: 56, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:49:54,852 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=153439.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:50:00,126 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.031e+02 5.043e+02 6.093e+02 8.078e+02 1.420e+03, threshold=1.219e+03, percent-clipped=2.0 +2023-04-02 23:50:17,461 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153459.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:50:26,642 INFO [train.py:903] (3/4) Epoch 23, batch 3250, loss[loss=0.2115, simple_loss=0.2827, pruned_loss=0.07016, over 19627.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.2891, pruned_loss=0.06394, over 3811239.00 frames. ], batch size: 50, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:50:43,103 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=153479.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:50:48,914 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153484.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:51:18,580 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9417, 2.0375, 2.3254, 2.5884, 1.9197, 2.5320, 2.4322, 2.2091], + device='cuda:3'), covar=tensor([0.3931, 0.3516, 0.1802, 0.2131, 0.3738, 0.1921, 0.4417, 0.3135], + device='cuda:3'), in_proj_covar=tensor([0.0906, 0.0973, 0.0719, 0.0933, 0.0884, 0.0819, 0.0846, 0.0786], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-02 23:51:27,759 INFO [train.py:903] (3/4) Epoch 23, batch 3300, loss[loss=0.1963, simple_loss=0.278, pruned_loss=0.0573, over 19664.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2887, pruned_loss=0.06396, over 3807302.38 frames. ], batch size: 53, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:51:34,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-02 23:52:00,748 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.814e+02 5.081e+02 6.216e+02 8.009e+02 2.047e+03, threshold=1.243e+03, percent-clipped=5.0 +2023-04-02 23:52:26,358 INFO [train.py:903] (3/4) Epoch 23, batch 3350, loss[loss=0.1886, simple_loss=0.2687, pruned_loss=0.05422, over 19374.00 frames. ], tot_loss[loss=0.2092, simple_loss=0.2894, pruned_loss=0.0645, over 3803499.76 frames. ], batch size: 47, lr: 3.58e-03, grad_scale: 8.0 +2023-04-02 23:53:00,338 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153594.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:53:26,225 INFO [train.py:903] (3/4) Epoch 23, batch 3400, loss[loss=0.2261, simple_loss=0.302, pruned_loss=0.07509, over 19694.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2887, pruned_loss=0.06442, over 3814262.41 frames. ], batch size: 60, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:53:56,995 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=153640.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:54:01,364 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.785e+02 5.295e+02 6.743e+02 8.549e+02 2.424e+03, threshold=1.349e+03, percent-clipped=5.0 +2023-04-02 23:54:28,041 INFO [train.py:903] (3/4) Epoch 23, batch 3450, loss[loss=0.2118, simple_loss=0.29, pruned_loss=0.06673, over 19614.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2878, pruned_loss=0.0641, over 3816518.14 frames. ], batch size: 57, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:54:31,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-02 23:55:30,146 INFO [train.py:903] (3/4) Epoch 23, batch 3500, loss[loss=0.1892, simple_loss=0.2678, pruned_loss=0.05529, over 19717.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2863, pruned_loss=0.06327, over 3819658.86 frames. ], batch size: 51, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:55:58,566 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1768, 1.2949, 1.7411, 0.9764, 2.3600, 3.0894, 2.7434, 3.2393], + device='cuda:3'), covar=tensor([0.1603, 0.3818, 0.3212, 0.2593, 0.0602, 0.0216, 0.0257, 0.0309], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0324, 0.0353, 0.0265, 0.0245, 0.0189, 0.0218, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-02 23:56:02,444 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.496e+02 4.938e+02 5.821e+02 7.521e+02 2.332e+03, threshold=1.164e+03, percent-clipped=1.0 +2023-04-02 23:56:17,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153755.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:56:30,175 INFO [train.py:903] (3/4) Epoch 23, batch 3550, loss[loss=0.2178, simple_loss=0.3024, pruned_loss=0.06665, over 18082.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.286, pruned_loss=0.06313, over 3814812.96 frames. ], batch size: 83, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:56:50,128 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=153783.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:57:30,145 INFO [train.py:903] (3/4) Epoch 23, batch 3600, loss[loss=0.2, simple_loss=0.2844, pruned_loss=0.05782, over 19746.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2869, pruned_loss=0.06365, over 3821191.73 frames. ], batch size: 63, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:58:05,069 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.445e+02 5.155e+02 6.351e+02 8.015e+02 2.586e+03, threshold=1.270e+03, percent-clipped=6.0 +2023-04-02 23:58:12,378 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=153850.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:58:30,878 INFO [train.py:903] (3/4) Epoch 23, batch 3650, loss[loss=0.2568, simple_loss=0.3238, pruned_loss=0.09493, over 18248.00 frames. ], tot_loss[loss=0.2085, simple_loss=0.288, pruned_loss=0.0645, over 3805752.42 frames. ], batch size: 83, lr: 3.57e-03, grad_scale: 8.0 +2023-04-02 23:58:42,987 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=153875.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:59:08,894 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=153898.0, num_to_drop=0, layers_to_drop=set() +2023-04-02 23:59:31,672 INFO [train.py:903] (3/4) Epoch 23, batch 3700, loss[loss=0.2019, simple_loss=0.3016, pruned_loss=0.05108, over 19592.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2884, pruned_loss=0.06457, over 3805172.13 frames. ], batch size: 61, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:00:00,767 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=153941.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:00:04,691 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.006e+02 4.617e+02 5.510e+02 6.874e+02 2.344e+03, threshold=1.102e+03, percent-clipped=3.0 +2023-04-03 00:00:31,960 INFO [train.py:903] (3/4) Epoch 23, batch 3750, loss[loss=0.2198, simple_loss=0.3025, pruned_loss=0.06853, over 18743.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2879, pruned_loss=0.06413, over 3800037.99 frames. ], batch size: 74, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:01:27,852 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154011.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:01:33,255 INFO [train.py:903] (3/4) Epoch 23, batch 3800, loss[loss=0.2097, simple_loss=0.2939, pruned_loss=0.06274, over 19663.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2871, pruned_loss=0.06352, over 3804798.27 frames. ], batch size: 59, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:01:59,779 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154036.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:02:05,111 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 00:02:08,268 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.228e+02 4.923e+02 6.103e+02 7.526e+02 2.694e+03, threshold=1.221e+03, percent-clipped=9.0 +2023-04-03 00:02:33,012 INFO [train.py:903] (3/4) Epoch 23, batch 3850, loss[loss=0.2243, simple_loss=0.3002, pruned_loss=0.07419, over 19781.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2871, pruned_loss=0.06296, over 3815830.11 frames. ], batch size: 54, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:02:35,304 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154067.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:02:38,816 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-03 00:02:38,915 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-04-03 00:02:45,160 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8804, 1.5060, 1.9376, 1.5921, 4.4299, 1.1196, 2.6438, 4.8481], + device='cuda:3'), covar=tensor([0.0494, 0.2906, 0.2729, 0.2122, 0.0749, 0.2752, 0.1373, 0.0159], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0369, 0.0391, 0.0352, 0.0375, 0.0352, 0.0386, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:02:47,671 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1740, 2.8743, 2.2777, 2.2286, 2.1327, 2.4870, 1.0992, 2.0819], + device='cuda:3'), covar=tensor([0.0644, 0.0608, 0.0729, 0.1166, 0.1129, 0.1168, 0.1466, 0.1111], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0355, 0.0363, 0.0386, 0.0463, 0.0391, 0.0339, 0.0342], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:03:09,354 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154094.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:03:13,920 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6300, 1.5928, 1.2209, 1.6021, 1.4603, 1.2976, 1.2503, 1.4758], + device='cuda:3'), covar=tensor([0.1337, 0.1420, 0.2062, 0.1227, 0.1476, 0.1049, 0.1997, 0.1121], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0354, 0.0310, 0.0250, 0.0302, 0.0249, 0.0309, 0.0255], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:03:18,589 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154102.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:03:35,394 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.61 vs. limit=2.0 +2023-04-03 00:03:35,945 INFO [train.py:903] (3/4) Epoch 23, batch 3900, loss[loss=0.1794, simple_loss=0.2604, pruned_loss=0.04923, over 19582.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2857, pruned_loss=0.0618, over 3821618.49 frames. ], batch size: 52, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:03:57,517 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7762, 1.8443, 2.2198, 2.0453, 3.3999, 2.8791, 3.7152, 1.8529], + device='cuda:3'), covar=tensor([0.2431, 0.4345, 0.2837, 0.1830, 0.1410, 0.2024, 0.1391, 0.4042], + device='cuda:3'), in_proj_covar=tensor([0.0539, 0.0649, 0.0718, 0.0488, 0.0621, 0.0536, 0.0662, 0.0554], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:04:09,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.867e+02 4.608e+02 5.656e+02 7.392e+02 1.919e+03, threshold=1.131e+03, percent-clipped=3.0 +2023-04-03 00:04:22,808 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154154.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:04:37,504 INFO [train.py:903] (3/4) Epoch 23, batch 3950, loss[loss=0.1794, simple_loss=0.2516, pruned_loss=0.05358, over 19066.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2862, pruned_loss=0.06244, over 3825467.99 frames. ], batch size: 42, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:04:44,236 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 00:04:52,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154179.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:05:37,018 INFO [train.py:903] (3/4) Epoch 23, batch 4000, loss[loss=0.2128, simple_loss=0.2943, pruned_loss=0.06567, over 19794.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2866, pruned_loss=0.06294, over 3808768.53 frames. ], batch size: 56, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:05:50,660 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-03 00:06:02,275 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-03 00:06:06,035 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154239.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:06:12,358 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.243e+02 5.130e+02 6.145e+02 8.525e+02 2.203e+03, threshold=1.229e+03, percent-clipped=9.0 +2023-04-03 00:06:27,023 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 00:06:37,053 INFO [train.py:903] (3/4) Epoch 23, batch 4050, loss[loss=0.1847, simple_loss=0.2558, pruned_loss=0.05683, over 19350.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2873, pruned_loss=0.06324, over 3812750.55 frames. ], batch size: 44, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:07:01,676 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154285.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:07:23,111 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154304.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:07:29,657 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154309.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:07:37,599 INFO [train.py:903] (3/4) Epoch 23, batch 4100, loss[loss=0.2112, simple_loss=0.2944, pruned_loss=0.064, over 19767.00 frames. ], tot_loss[loss=0.207, simple_loss=0.287, pruned_loss=0.06349, over 3810224.97 frames. ], batch size: 54, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:08:11,156 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 5.105e+02 5.940e+02 7.682e+02 1.555e+03, threshold=1.188e+03, percent-clipped=4.0 +2023-04-03 00:08:13,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 00:08:25,634 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154355.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:08:39,411 INFO [train.py:903] (3/4) Epoch 23, batch 4150, loss[loss=0.1857, simple_loss=0.2592, pruned_loss=0.05607, over 17333.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2866, pruned_loss=0.06292, over 3810690.68 frames. ], batch size: 38, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:09:22,104 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154400.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:09:33,989 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154411.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:09:39,113 INFO [train.py:903] (3/4) Epoch 23, batch 4200, loss[loss=0.2065, simple_loss=0.2819, pruned_loss=0.06553, over 19671.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2865, pruned_loss=0.06269, over 3818794.74 frames. ], batch size: 53, lr: 3.57e-03, grad_scale: 8.0 +2023-04-03 00:09:41,423 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 00:10:07,070 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154438.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:10:14,754 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.103e+02 4.744e+02 5.863e+02 7.378e+02 1.705e+03, threshold=1.173e+03, percent-clipped=3.0 +2023-04-03 00:10:17,215 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154446.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:10:40,158 INFO [train.py:903] (3/4) Epoch 23, batch 4250, loss[loss=0.2228, simple_loss=0.3077, pruned_loss=0.06898, over 19574.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2863, pruned_loss=0.06251, over 3825954.90 frames. ], batch size: 61, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:10:54,216 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 00:11:05,307 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 00:11:12,092 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154492.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:11:40,251 INFO [train.py:903] (3/4) Epoch 23, batch 4300, loss[loss=0.2013, simple_loss=0.2778, pruned_loss=0.06239, over 19767.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2854, pruned_loss=0.06169, over 3840633.56 frames. ], batch size: 54, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:11:53,659 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154526.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 00:11:55,919 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5405, 2.2107, 1.6344, 1.4444, 2.0738, 1.3033, 1.3225, 1.9324], + device='cuda:3'), covar=tensor([0.1042, 0.0762, 0.1152, 0.0885, 0.0567, 0.1297, 0.0777, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0300, 0.0313, 0.0337, 0.0263, 0.0246, 0.0336, 0.0289, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:12:13,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.268e+02 4.652e+02 5.888e+02 7.584e+02 1.931e+03, threshold=1.178e+03, percent-clipped=3.0 +2023-04-03 00:12:24,518 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154553.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:12:33,755 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 00:12:36,076 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154561.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:12:41,454 INFO [train.py:903] (3/4) Epoch 23, batch 4350, loss[loss=0.2321, simple_loss=0.3236, pruned_loss=0.07026, over 19496.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2861, pruned_loss=0.06211, over 3846038.23 frames. ], batch size: 64, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:13:01,191 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154583.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:13:04,734 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7613, 1.5917, 1.3809, 1.6640, 1.4086, 1.3762, 1.2435, 1.6057], + device='cuda:3'), covar=tensor([0.1119, 0.1341, 0.1695, 0.1098, 0.1419, 0.0765, 0.1827, 0.0861], + device='cuda:3'), in_proj_covar=tensor([0.0272, 0.0355, 0.0313, 0.0252, 0.0302, 0.0251, 0.0311, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:13:40,266 INFO [train.py:903] (3/4) Epoch 23, batch 4400, loss[loss=0.1814, simple_loss=0.2674, pruned_loss=0.04768, over 19586.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2859, pruned_loss=0.06227, over 3844923.31 frames. ], batch size: 52, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:14:04,353 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 00:14:14,042 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.271e+02 5.197e+02 6.555e+02 7.915e+02 1.480e+03, threshold=1.311e+03, percent-clipped=6.0 +2023-04-03 00:14:15,158 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 00:14:16,486 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154646.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:14:18,497 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154648.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:14:23,887 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154653.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:14:27,578 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154656.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:14:38,569 INFO [train.py:903] (3/4) Epoch 23, batch 4450, loss[loss=0.2056, simple_loss=0.2941, pruned_loss=0.05858, over 19768.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2858, pruned_loss=0.06255, over 3841459.86 frames. ], batch size: 56, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:14:58,481 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154681.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:15:18,437 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154698.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:15:19,415 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154699.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 00:15:38,749 INFO [train.py:903] (3/4) Epoch 23, batch 4500, loss[loss=0.2477, simple_loss=0.3225, pruned_loss=0.0865, over 19335.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2866, pruned_loss=0.06313, over 3819834.89 frames. ], batch size: 66, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:16:06,367 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154738.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:16:13,918 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.080e+02 4.922e+02 6.448e+02 7.735e+02 1.395e+03, threshold=1.290e+03, percent-clipped=1.0 +2023-04-03 00:16:38,071 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154763.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:16:41,146 INFO [train.py:903] (3/4) Epoch 23, batch 4550, loss[loss=0.2341, simple_loss=0.3097, pruned_loss=0.07932, over 17316.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2867, pruned_loss=0.06303, over 3824960.00 frames. ], batch size: 101, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:16:43,820 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154768.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:16:48,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 00:17:00,102 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154782.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 00:17:08,985 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8965, 2.5986, 2.5742, 2.7723, 2.6922, 2.5184, 2.2841, 2.8642], + device='cuda:3'), covar=tensor([0.0869, 0.1508, 0.1286, 0.1090, 0.1327, 0.0485, 0.1330, 0.0630], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0356, 0.0314, 0.0253, 0.0304, 0.0253, 0.0312, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:17:11,895 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 00:17:16,374 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6370, 4.2402, 2.6988, 3.7712, 1.0807, 4.2115, 4.0807, 4.1725], + device='cuda:3'), covar=tensor([0.0592, 0.0937, 0.1872, 0.0858, 0.3646, 0.0592, 0.0842, 0.0996], + device='cuda:3'), in_proj_covar=tensor([0.0513, 0.0414, 0.0496, 0.0348, 0.0399, 0.0438, 0.0428, 0.0463], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:17:31,879 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154807.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:17:34,167 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154809.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:17:39,921 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154814.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:17:41,844 INFO [train.py:903] (3/4) Epoch 23, batch 4600, loss[loss=0.2008, simple_loss=0.2874, pruned_loss=0.05713, over 19357.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2855, pruned_loss=0.06247, over 3836002.79 frames. ], batch size: 66, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:17:43,431 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154817.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:18:02,697 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154834.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:18:05,369 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154836.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:18:13,499 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154842.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:18:17,275 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.766e+02 4.757e+02 5.456e+02 7.137e+02 2.039e+03, threshold=1.091e+03, percent-clipped=4.0 +2023-04-03 00:18:39,894 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3595, 1.4329, 1.6367, 1.4210, 2.9590, 1.0279, 2.3252, 3.4364], + device='cuda:3'), covar=tensor([0.0555, 0.2726, 0.2665, 0.1984, 0.0723, 0.2586, 0.1267, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0411, 0.0367, 0.0389, 0.0350, 0.0373, 0.0350, 0.0383, 0.0404], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:18:41,882 INFO [train.py:903] (3/4) Epoch 23, batch 4650, loss[loss=0.1736, simple_loss=0.2522, pruned_loss=0.04751, over 19475.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2854, pruned_loss=0.06257, over 3828627.93 frames. ], batch size: 49, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:18:45,721 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4921, 1.4185, 1.3708, 1.7655, 1.3728, 1.7157, 1.7370, 1.5352], + device='cuda:3'), covar=tensor([0.0894, 0.0967, 0.1084, 0.0717, 0.0862, 0.0780, 0.0771, 0.0735], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0221, 0.0226, 0.0240, 0.0226, 0.0212, 0.0187, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-03 00:18:57,547 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 00:19:09,934 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 00:19:42,549 INFO [train.py:903] (3/4) Epoch 23, batch 4700, loss[loss=0.2273, simple_loss=0.2913, pruned_loss=0.08168, over 19391.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2857, pruned_loss=0.06265, over 3833586.27 frames. ], batch size: 47, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:20:04,434 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 00:20:17,977 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.052e+02 4.550e+02 5.511e+02 7.065e+02 1.410e+03, threshold=1.102e+03, percent-clipped=2.0 +2023-04-03 00:20:25,040 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=154951.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:20:28,291 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=154954.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:20:44,148 INFO [train.py:903] (3/4) Epoch 23, batch 4750, loss[loss=0.2011, simple_loss=0.2871, pruned_loss=0.0575, over 19677.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2858, pruned_loss=0.0628, over 3839094.95 frames. ], batch size: 59, lr: 3.56e-03, grad_scale: 4.0 +2023-04-03 00:21:00,287 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=154979.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:21:12,344 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=154990.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:21:17,064 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=154994.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:21:22,627 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0187, 2.1095, 2.2642, 2.6329, 2.0380, 2.6343, 2.2106, 2.0663], + device='cuda:3'), covar=tensor([0.4353, 0.3849, 0.2022, 0.2381, 0.3966, 0.2054, 0.5361, 0.3473], + device='cuda:3'), in_proj_covar=tensor([0.0907, 0.0975, 0.0723, 0.0937, 0.0886, 0.0825, 0.0846, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 00:21:45,312 INFO [train.py:903] (3/4) Epoch 23, batch 4800, loss[loss=0.2232, simple_loss=0.2917, pruned_loss=0.07737, over 19482.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2859, pruned_loss=0.06257, over 3848854.57 frames. ], batch size: 49, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:21:49,080 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155019.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:21:54,341 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155024.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:22:18,867 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155044.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:22:19,572 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.470e+02 5.324e+02 6.216e+02 7.674e+02 2.163e+03, threshold=1.243e+03, percent-clipped=8.0 +2023-04-03 00:22:26,106 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155049.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:22:44,546 INFO [train.py:903] (3/4) Epoch 23, batch 4850, loss[loss=0.2071, simple_loss=0.29, pruned_loss=0.06207, over 19526.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2869, pruned_loss=0.06296, over 3836697.44 frames. ], batch size: 56, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:22:49,284 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0510, 1.9676, 1.7642, 1.6181, 1.2603, 1.5107, 0.5699, 0.9422], + device='cuda:3'), covar=tensor([0.0803, 0.0755, 0.0588, 0.0940, 0.1634, 0.1292, 0.1527, 0.1325], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0353, 0.0360, 0.0383, 0.0461, 0.0389, 0.0336, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:22:49,293 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155070.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:22:49,326 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6537, 1.8073, 1.7716, 2.5649, 1.8699, 2.4352, 1.8555, 1.5346], + device='cuda:3'), covar=tensor([0.5278, 0.4651, 0.3078, 0.3194, 0.4845, 0.2655, 0.6791, 0.5537], + device='cuda:3'), in_proj_covar=tensor([0.0906, 0.0975, 0.0722, 0.0937, 0.0887, 0.0824, 0.0845, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 00:23:03,365 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=155082.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:23:09,618 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 00:23:11,892 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=155087.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:23:21,450 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155095.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 00:23:29,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 00:23:32,600 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=155105.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:23:34,406 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 00:23:34,434 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 00:23:44,624 INFO [train.py:903] (3/4) Epoch 23, batch 4900, loss[loss=0.1711, simple_loss=0.2543, pruned_loss=0.04398, over 19725.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2863, pruned_loss=0.06273, over 3823414.70 frames. ], batch size: 51, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:23:44,636 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 00:24:04,411 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 00:24:20,243 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.924e+02 5.163e+02 5.938e+02 7.647e+02 1.407e+03, threshold=1.188e+03, percent-clipped=5.0 +2023-04-03 00:24:25,548 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.46 vs. limit=2.0 +2023-04-03 00:24:46,212 INFO [train.py:903] (3/4) Epoch 23, batch 4950, loss[loss=0.1954, simple_loss=0.2784, pruned_loss=0.05623, over 19769.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.286, pruned_loss=0.06221, over 3825958.64 frames. ], batch size: 54, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:25:01,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 00:25:21,567 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=155197.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:25:23,143 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 00:25:29,804 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8826, 1.9704, 2.1871, 2.4719, 1.8781, 2.3236, 2.2306, 2.0670], + device='cuda:3'), covar=tensor([0.4078, 0.3602, 0.1805, 0.2245, 0.3797, 0.2045, 0.4615, 0.3108], + device='cuda:3'), in_proj_covar=tensor([0.0910, 0.0978, 0.0725, 0.0940, 0.0890, 0.0827, 0.0847, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 00:25:34,815 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155207.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:25:44,527 INFO [train.py:903] (3/4) Epoch 23, batch 5000, loss[loss=0.1812, simple_loss=0.2733, pruned_loss=0.04451, over 19539.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2862, pruned_loss=0.06237, over 3829353.60 frames. ], batch size: 56, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:25:52,522 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 00:25:54,041 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3735, 2.3802, 2.5305, 3.1994, 2.4316, 3.0406, 2.6566, 2.5211], + device='cuda:3'), covar=tensor([0.4150, 0.4089, 0.1933, 0.2464, 0.4349, 0.2116, 0.4660, 0.3189], + device='cuda:3'), in_proj_covar=tensor([0.0911, 0.0979, 0.0726, 0.0941, 0.0891, 0.0828, 0.0849, 0.0790], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 00:26:02,955 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155232.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:26:03,632 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 00:26:19,061 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.096e+02 4.751e+02 5.889e+02 7.363e+02 1.722e+03, threshold=1.178e+03, percent-clipped=5.0 +2023-04-03 00:26:43,567 INFO [train.py:903] (3/4) Epoch 23, batch 5050, loss[loss=0.1769, simple_loss=0.2557, pruned_loss=0.04902, over 19835.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2845, pruned_loss=0.06136, over 3847420.67 frames. ], batch size: 52, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:27:17,601 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 00:27:41,260 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.79 vs. limit=5.0 +2023-04-03 00:27:42,610 INFO [train.py:903] (3/4) Epoch 23, batch 5100, loss[loss=0.2018, simple_loss=0.2693, pruned_loss=0.06713, over 19798.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2849, pruned_loss=0.06196, over 3833869.19 frames. ], batch size: 49, lr: 3.56e-03, grad_scale: 8.0 +2023-04-03 00:27:45,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7638, 1.8442, 2.1919, 2.2681, 1.6909, 2.1775, 2.1646, 1.9960], + device='cuda:3'), covar=tensor([0.4479, 0.4202, 0.2058, 0.2434, 0.4138, 0.2349, 0.5334, 0.3665], + device='cuda:3'), in_proj_covar=tensor([0.0907, 0.0976, 0.0722, 0.0937, 0.0888, 0.0824, 0.0847, 0.0787], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 00:27:53,104 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 00:27:56,480 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 00:28:01,507 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 00:28:10,587 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=155338.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:28:18,270 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.450e+02 5.084e+02 6.467e+02 7.878e+02 1.414e+03, threshold=1.293e+03, percent-clipped=6.0 +2023-04-03 00:28:37,848 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155361.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:28:43,671 INFO [train.py:903] (3/4) Epoch 23, batch 5150, loss[loss=0.2309, simple_loss=0.3059, pruned_loss=0.07795, over 19663.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2853, pruned_loss=0.06249, over 3829315.31 frames. ], batch size: 55, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:28:44,019 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7323, 1.4521, 1.6672, 1.5902, 3.3238, 1.1905, 2.5101, 3.7603], + device='cuda:3'), covar=tensor([0.0498, 0.2792, 0.2701, 0.1900, 0.0678, 0.2496, 0.1235, 0.0224], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0370, 0.0392, 0.0352, 0.0376, 0.0354, 0.0387, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:28:56,726 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 00:29:08,599 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155386.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:29:13,153 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4655, 1.5362, 1.7480, 1.6809, 2.2847, 2.1507, 2.3688, 0.9100], + device='cuda:3'), covar=tensor([0.2400, 0.4143, 0.2628, 0.1884, 0.1513, 0.2160, 0.1423, 0.4693], + device='cuda:3'), in_proj_covar=tensor([0.0538, 0.0647, 0.0719, 0.0490, 0.0622, 0.0535, 0.0659, 0.0553], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:29:30,439 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 00:29:38,190 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7349, 4.0436, 4.5369, 4.5706, 1.9838, 4.2624, 3.6072, 3.9392], + device='cuda:3'), covar=tensor([0.2109, 0.1471, 0.0980, 0.1237, 0.6880, 0.1799, 0.1283, 0.1980], + device='cuda:3'), in_proj_covar=tensor([0.0787, 0.0751, 0.0958, 0.0838, 0.0841, 0.0719, 0.0574, 0.0890], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 00:29:45,066 INFO [train.py:903] (3/4) Epoch 23, batch 5200, loss[loss=0.23, simple_loss=0.2971, pruned_loss=0.08142, over 19477.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.2846, pruned_loss=0.06203, over 3829890.54 frames. ], batch size: 49, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:29:58,653 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 00:30:02,273 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=155431.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:30:19,770 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.141e+02 5.305e+02 6.432e+02 7.969e+02 2.733e+03, threshold=1.286e+03, percent-clipped=6.0 +2023-04-03 00:30:30,716 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=155453.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:30:30,784 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155453.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:30:41,482 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 00:30:44,914 INFO [train.py:903] (3/4) Epoch 23, batch 5250, loss[loss=0.2054, simple_loss=0.2883, pruned_loss=0.06126, over 19686.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2855, pruned_loss=0.06247, over 3824179.07 frames. ], batch size: 60, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:30:55,622 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=155475.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:30:59,196 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155478.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:31:45,243 INFO [train.py:903] (3/4) Epoch 23, batch 5300, loss[loss=0.1963, simple_loss=0.2738, pruned_loss=0.05942, over 19581.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2865, pruned_loss=0.06275, over 3819163.01 frames. ], batch size: 52, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:32:03,697 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 00:32:21,416 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.105e+02 4.703e+02 5.856e+02 7.687e+02 1.612e+03, threshold=1.171e+03, percent-clipped=4.0 +2023-04-03 00:32:22,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=155546.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:32:46,442 INFO [train.py:903] (3/4) Epoch 23, batch 5350, loss[loss=0.1941, simple_loss=0.2818, pruned_loss=0.05317, over 19651.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2868, pruned_loss=0.06275, over 3809954.88 frames. ], batch size: 55, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:33:18,092 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 00:33:46,943 INFO [train.py:903] (3/4) Epoch 23, batch 5400, loss[loss=0.1876, simple_loss=0.2586, pruned_loss=0.05823, over 19778.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.287, pruned_loss=0.06304, over 3809017.37 frames. ], batch size: 47, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:33:56,225 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=155623.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:34:21,905 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.892e+02 4.747e+02 5.806e+02 7.220e+02 1.360e+03, threshold=1.161e+03, percent-clipped=1.0 +2023-04-03 00:34:36,328 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5428, 1.6005, 2.0976, 1.7847, 3.1091, 4.7658, 4.5796, 5.1940], + device='cuda:3'), covar=tensor([0.1477, 0.3569, 0.3086, 0.2119, 0.0574, 0.0170, 0.0167, 0.0179], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0325, 0.0355, 0.0265, 0.0246, 0.0190, 0.0218, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 00:34:48,073 INFO [train.py:903] (3/4) Epoch 23, batch 5450, loss[loss=0.1866, simple_loss=0.2763, pruned_loss=0.04847, over 19521.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2861, pruned_loss=0.06251, over 3805279.43 frames. ], batch size: 54, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:34:58,345 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=155675.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:35:10,198 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1103, 1.3026, 1.7290, 1.2137, 2.7458, 3.7233, 3.4042, 3.8796], + device='cuda:3'), covar=tensor([0.1746, 0.3962, 0.3541, 0.2625, 0.0629, 0.0219, 0.0220, 0.0278], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0325, 0.0354, 0.0265, 0.0246, 0.0190, 0.0218, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 00:35:39,183 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 00:35:39,988 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155709.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:35:47,581 INFO [train.py:903] (3/4) Epoch 23, batch 5500, loss[loss=0.2545, simple_loss=0.3271, pruned_loss=0.09091, over 19278.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2862, pruned_loss=0.0629, over 3801706.27 frames. ], batch size: 66, lr: 3.55e-03, grad_scale: 4.0 +2023-04-03 00:36:10,848 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155734.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:36:13,485 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 00:36:14,851 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7870, 4.3275, 2.8353, 3.8922, 0.8885, 4.3483, 4.2033, 4.3066], + device='cuda:3'), covar=tensor([0.0631, 0.0979, 0.1935, 0.0790, 0.4204, 0.0636, 0.0875, 0.1168], + device='cuda:3'), in_proj_covar=tensor([0.0518, 0.0418, 0.0502, 0.0352, 0.0404, 0.0442, 0.0433, 0.0467], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:36:21,510 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3100, 1.4228, 1.6893, 1.6147, 2.2311, 1.9463, 2.1703, 0.9122], + device='cuda:3'), covar=tensor([0.2768, 0.4663, 0.2967, 0.2207, 0.1693, 0.2650, 0.1770, 0.5099], + device='cuda:3'), in_proj_covar=tensor([0.0539, 0.0650, 0.0722, 0.0492, 0.0625, 0.0539, 0.0662, 0.0556], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:36:24,174 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.224e+02 5.057e+02 6.298e+02 8.158e+02 1.659e+03, threshold=1.260e+03, percent-clipped=6.0 +2023-04-03 00:36:46,709 INFO [train.py:903] (3/4) Epoch 23, batch 5550, loss[loss=0.1792, simple_loss=0.2628, pruned_loss=0.04781, over 19400.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2862, pruned_loss=0.06293, over 3816776.42 frames. ], batch size: 48, lr: 3.55e-03, grad_scale: 4.0 +2023-04-03 00:36:56,205 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 00:37:30,539 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=155802.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:37:30,684 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=155802.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:37:35,029 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9058, 1.3143, 1.0778, 1.0422, 1.1644, 1.0301, 1.0358, 1.2622], + device='cuda:3'), covar=tensor([0.0640, 0.0941, 0.1163, 0.0751, 0.0633, 0.1408, 0.0568, 0.0524], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0314, 0.0336, 0.0265, 0.0247, 0.0339, 0.0289, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:37:42,230 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 00:37:48,755 INFO [train.py:903] (3/4) Epoch 23, batch 5600, loss[loss=0.1678, simple_loss=0.2434, pruned_loss=0.0461, over 19742.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2878, pruned_loss=0.06384, over 3813774.74 frames. ], batch size: 46, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:37:52,350 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=155819.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:37:59,040 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3245, 3.0824, 2.4104, 2.4125, 2.1798, 2.6514, 0.9664, 2.1789], + device='cuda:3'), covar=tensor([0.0676, 0.0592, 0.0731, 0.1134, 0.1142, 0.1109, 0.1576, 0.1154], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0353, 0.0360, 0.0384, 0.0462, 0.0389, 0.0337, 0.0342], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 00:38:01,977 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-03 00:38:02,554 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=155827.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:38:23,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.767e+02 5.365e+02 7.033e+02 8.601e+02 1.530e+03, threshold=1.407e+03, percent-clipped=6.0 +2023-04-03 00:38:48,667 INFO [train.py:903] (3/4) Epoch 23, batch 5650, loss[loss=0.2401, simple_loss=0.3153, pruned_loss=0.08248, over 19469.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.287, pruned_loss=0.06401, over 3813115.39 frames. ], batch size: 64, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:38:50,814 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-03 00:39:33,331 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 00:39:47,802 INFO [train.py:903] (3/4) Epoch 23, batch 5700, loss[loss=0.2429, simple_loss=0.3186, pruned_loss=0.08362, over 18040.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2875, pruned_loss=0.0644, over 3810235.35 frames. ], batch size: 83, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:40:10,805 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=155934.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:40:17,505 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3858, 1.3079, 1.7779, 1.4288, 2.7119, 3.7455, 3.4616, 3.9625], + device='cuda:3'), covar=tensor([0.1505, 0.3901, 0.3384, 0.2397, 0.0595, 0.0176, 0.0206, 0.0243], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0324, 0.0353, 0.0264, 0.0246, 0.0190, 0.0217, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 00:40:24,773 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.238e+02 4.906e+02 6.184e+02 7.924e+02 2.131e+03, threshold=1.237e+03, percent-clipped=2.0 +2023-04-03 00:40:47,848 INFO [train.py:903] (3/4) Epoch 23, batch 5750, loss[loss=0.2027, simple_loss=0.2848, pruned_loss=0.06024, over 19569.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2871, pruned_loss=0.06387, over 3820076.14 frames. ], batch size: 61, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:40:50,070 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=155967.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:40:51,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 00:40:58,276 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.80 vs. limit=5.0 +2023-04-03 00:40:58,764 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 00:41:04,145 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 00:41:50,977 INFO [train.py:903] (3/4) Epoch 23, batch 5800, loss[loss=0.1975, simple_loss=0.2818, pruned_loss=0.05657, over 18697.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2868, pruned_loss=0.06393, over 3815883.52 frames. ], batch size: 74, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:41:54,506 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=156019.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:42:25,081 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.381e+02 4.982e+02 6.301e+02 7.857e+02 1.493e+03, threshold=1.260e+03, percent-clipped=3.0 +2023-04-03 00:42:39,979 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 00:42:50,189 INFO [train.py:903] (3/4) Epoch 23, batch 5850, loss[loss=0.257, simple_loss=0.3209, pruned_loss=0.09655, over 13921.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2869, pruned_loss=0.06398, over 3809437.69 frames. ], batch size: 136, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:42:56,254 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 00:43:08,258 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156082.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:43:15,708 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=156088.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:43:48,548 INFO [train.py:903] (3/4) Epoch 23, batch 5900, loss[loss=0.1678, simple_loss=0.2457, pruned_loss=0.04491, over 19763.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2867, pruned_loss=0.06371, over 3820490.42 frames. ], batch size: 46, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:43:52,974 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 00:44:08,213 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-03 00:44:10,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156134.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:44:14,067 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 00:44:24,575 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.271e+02 5.043e+02 6.365e+02 8.179e+02 2.050e+03, threshold=1.273e+03, percent-clipped=8.0 +2023-04-03 00:44:25,631 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=156146.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:44:48,124 INFO [train.py:903] (3/4) Epoch 23, batch 5950, loss[loss=0.1743, simple_loss=0.2508, pruned_loss=0.04891, over 19797.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2869, pruned_loss=0.06406, over 3824032.73 frames. ], batch size: 47, lr: 3.55e-03, grad_scale: 8.0 +2023-04-03 00:45:19,209 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156190.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:45:47,986 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156215.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:45:48,679 INFO [train.py:903] (3/4) Epoch 23, batch 6000, loss[loss=0.1889, simple_loss=0.2689, pruned_loss=0.05441, over 19752.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2874, pruned_loss=0.06437, over 3804428.36 frames. ], batch size: 48, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:45:48,679 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 00:46:01,141 INFO [train.py:937] (3/4) Epoch 23, validation: loss=0.1686, simple_loss=0.2684, pruned_loss=0.03439, over 944034.00 frames. +2023-04-03 00:46:01,142 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 00:46:23,292 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=156234.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:46:37,238 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.373e+02 4.873e+02 6.527e+02 8.069e+02 1.468e+03, threshold=1.305e+03, percent-clipped=4.0 +2023-04-03 00:46:55,749 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156261.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:47:01,920 INFO [train.py:903] (3/4) Epoch 23, batch 6050, loss[loss=0.2087, simple_loss=0.2914, pruned_loss=0.06304, over 19759.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2879, pruned_loss=0.06445, over 3808227.21 frames. ], batch size: 63, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:47:13,011 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-03 00:47:53,117 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=156308.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:48:02,112 INFO [train.py:903] (3/4) Epoch 23, batch 6100, loss[loss=0.2267, simple_loss=0.3034, pruned_loss=0.07496, over 19549.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2882, pruned_loss=0.0643, over 3822949.64 frames. ], batch size: 54, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:48:27,997 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156338.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:48:37,367 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.018e+02 4.768e+02 6.249e+02 8.138e+02 1.749e+03, threshold=1.250e+03, percent-clipped=2.0 +2023-04-03 00:48:58,814 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156363.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:49:01,830 INFO [train.py:903] (3/4) Epoch 23, batch 6150, loss[loss=0.2892, simple_loss=0.3523, pruned_loss=0.113, over 19764.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.287, pruned_loss=0.06397, over 3828891.82 frames. ], batch size: 63, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:49:31,069 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156390.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:49:31,828 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 00:50:00,534 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156415.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:50:01,340 INFO [train.py:903] (3/4) Epoch 23, batch 6200, loss[loss=0.2017, simple_loss=0.285, pruned_loss=0.05923, over 17702.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2878, pruned_loss=0.06396, over 3817719.21 frames. ], batch size: 101, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:50:22,393 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=156432.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:50:38,865 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.999e+02 4.815e+02 5.704e+02 6.895e+02 2.552e+03, threshold=1.141e+03, percent-clipped=3.0 +2023-04-03 00:51:02,806 INFO [train.py:903] (3/4) Epoch 23, batch 6250, loss[loss=0.2171, simple_loss=0.2939, pruned_loss=0.07015, over 19686.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2883, pruned_loss=0.06417, over 3807191.58 frames. ], batch size: 53, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:51:23,600 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-04-03 00:51:29,562 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0976, 4.4765, 4.8151, 4.8180, 1.8921, 4.5254, 3.9213, 4.5127], + device='cuda:3'), covar=tensor([0.1648, 0.0815, 0.0628, 0.0671, 0.5924, 0.0897, 0.0657, 0.1192], + device='cuda:3'), in_proj_covar=tensor([0.0780, 0.0746, 0.0948, 0.0826, 0.0833, 0.0715, 0.0569, 0.0880], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 00:51:32,680 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 00:52:02,820 INFO [train.py:903] (3/4) Epoch 23, batch 6300, loss[loss=0.2988, simple_loss=0.3608, pruned_loss=0.1184, over 17506.00 frames. ], tot_loss[loss=0.2081, simple_loss=0.2883, pruned_loss=0.06395, over 3799940.27 frames. ], batch size: 101, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:52:04,466 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156517.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:52:34,305 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156542.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:52:39,265 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.268e+02 5.353e+02 6.743e+02 8.019e+02 1.408e+03, threshold=1.349e+03, percent-clipped=4.0 +2023-04-03 00:52:40,698 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156547.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:53:03,411 INFO [train.py:903] (3/4) Epoch 23, batch 6350, loss[loss=0.2219, simple_loss=0.3066, pruned_loss=0.06859, over 17454.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2877, pruned_loss=0.06353, over 3813838.64 frames. ], batch size: 101, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:53:17,256 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=156578.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:53:22,118 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.55 vs. limit=2.0 +2023-04-03 00:54:02,664 INFO [train.py:903] (3/4) Epoch 23, batch 6400, loss[loss=0.1741, simple_loss=0.2513, pruned_loss=0.04846, over 19733.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2885, pruned_loss=0.06397, over 3815814.41 frames. ], batch size: 46, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:54:39,387 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.510e+02 4.834e+02 5.927e+02 7.987e+02 2.615e+03, threshold=1.185e+03, percent-clipped=4.0 +2023-04-03 00:54:46,364 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=156652.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:55:04,062 INFO [train.py:903] (3/4) Epoch 23, batch 6450, loss[loss=0.2511, simple_loss=0.3222, pruned_loss=0.09004, over 19666.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2893, pruned_loss=0.06469, over 3806862.08 frames. ], batch size: 60, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:55:35,962 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156693.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:55:47,917 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 00:56:04,449 INFO [train.py:903] (3/4) Epoch 23, batch 6500, loss[loss=0.2235, simple_loss=0.2823, pruned_loss=0.08234, over 19753.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2888, pruned_loss=0.06441, over 3822189.96 frames. ], batch size: 45, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:56:10,078 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 00:56:39,934 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.419e+02 5.077e+02 6.090e+02 8.057e+02 1.603e+03, threshold=1.218e+03, percent-clipped=6.0 +2023-04-03 00:57:04,749 INFO [train.py:903] (3/4) Epoch 23, batch 6550, loss[loss=0.305, simple_loss=0.3524, pruned_loss=0.1288, over 13282.00 frames. ], tot_loss[loss=0.2091, simple_loss=0.289, pruned_loss=0.06461, over 3799765.68 frames. ], batch size: 137, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:57:06,356 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=156767.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:57:06,424 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=156767.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:57:50,677 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156803.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:58:04,880 INFO [train.py:903] (3/4) Epoch 23, batch 6600, loss[loss=0.1926, simple_loss=0.2797, pruned_loss=0.05276, over 19327.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2882, pruned_loss=0.06393, over 3801995.13 frames. ], batch size: 66, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:58:16,146 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2113, 5.6749, 3.2051, 4.9183, 0.9674, 5.8414, 5.5908, 5.8755], + device='cuda:3'), covar=tensor([0.0381, 0.0915, 0.1656, 0.0741, 0.4084, 0.0440, 0.0779, 0.0832], + device='cuda:3'), in_proj_covar=tensor([0.0518, 0.0414, 0.0499, 0.0351, 0.0403, 0.0437, 0.0435, 0.0469], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 00:58:20,348 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156828.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 00:58:41,922 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.453e+02 5.163e+02 6.336e+02 8.010e+02 1.885e+03, threshold=1.267e+03, percent-clipped=5.0 +2023-04-03 00:59:05,158 INFO [train.py:903] (3/4) Epoch 23, batch 6650, loss[loss=0.2017, simple_loss=0.2868, pruned_loss=0.0583, over 19654.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06286, over 3813761.44 frames. ], batch size: 58, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 00:59:46,891 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2874, 1.5237, 1.8722, 1.4442, 2.7925, 3.4682, 3.2919, 3.5871], + device='cuda:3'), covar=tensor([0.1713, 0.3579, 0.3284, 0.2523, 0.0697, 0.0229, 0.0219, 0.0341], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0324, 0.0353, 0.0265, 0.0245, 0.0190, 0.0216, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 00:59:51,184 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=156903.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:00:07,306 INFO [train.py:903] (3/4) Epoch 23, batch 6700, loss[loss=0.2461, simple_loss=0.3205, pruned_loss=0.08587, over 19429.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2869, pruned_loss=0.0633, over 3804461.10 frames. ], batch size: 70, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 01:00:17,072 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-04-03 01:00:41,740 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.603e+02 5.252e+02 6.535e+02 7.903e+02 1.565e+03, threshold=1.307e+03, percent-clipped=2.0 +2023-04-03 01:00:44,238 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9741, 1.2445, 1.6315, 0.6121, 2.0669, 2.4440, 2.1534, 2.6041], + device='cuda:3'), covar=tensor([0.1537, 0.3816, 0.3305, 0.2831, 0.0614, 0.0292, 0.0348, 0.0370], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0324, 0.0353, 0.0264, 0.0245, 0.0190, 0.0216, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 01:00:45,432 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=156949.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:01:04,464 INFO [train.py:903] (3/4) Epoch 23, batch 6750, loss[loss=0.2671, simple_loss=0.3477, pruned_loss=0.09324, over 19665.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2862, pruned_loss=0.06355, over 3792241.18 frames. ], batch size: 60, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 01:01:13,650 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=156974.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:01:29,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0561, 2.1347, 2.4489, 2.6713, 2.0091, 2.6077, 2.3805, 2.1929], + device='cuda:3'), covar=tensor([0.4430, 0.4130, 0.1950, 0.2672, 0.4446, 0.2254, 0.5122, 0.3390], + device='cuda:3'), in_proj_covar=tensor([0.0909, 0.0978, 0.0723, 0.0935, 0.0890, 0.0824, 0.0847, 0.0787], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 01:02:00,909 INFO [train.py:903] (3/4) Epoch 23, batch 6800, loss[loss=0.1737, simple_loss=0.2526, pruned_loss=0.04744, over 19755.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.286, pruned_loss=0.06364, over 3775907.75 frames. ], batch size: 47, lr: 3.54e-03, grad_scale: 8.0 +2023-04-03 01:02:09,471 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=157023.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:02:13,976 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3588, 2.0321, 1.6171, 1.3832, 1.8555, 1.3883, 1.2672, 1.8501], + device='cuda:3'), covar=tensor([0.0986, 0.0904, 0.1118, 0.0896, 0.0622, 0.1241, 0.0724, 0.0478], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0314, 0.0337, 0.0267, 0.0247, 0.0338, 0.0289, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:02:14,983 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157028.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:02:22,951 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7741, 1.9607, 2.2664, 2.3000, 3.0251, 3.6239, 3.4764, 3.9958], + device='cuda:3'), covar=tensor([0.1550, 0.3871, 0.3513, 0.2122, 0.1283, 0.0454, 0.0265, 0.0291], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0325, 0.0354, 0.0265, 0.0246, 0.0190, 0.0217, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 01:02:44,734 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 01:02:45,839 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 01:02:48,304 INFO [train.py:903] (3/4) Epoch 24, batch 0, loss[loss=0.1896, simple_loss=0.2604, pruned_loss=0.05942, over 19307.00 frames. ], tot_loss[loss=0.1896, simple_loss=0.2604, pruned_loss=0.05942, over 19307.00 frames. ], batch size: 44, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:02:48,304 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 01:02:59,924 INFO [train.py:937] (3/4) Epoch 24, validation: loss=0.1683, simple_loss=0.2685, pruned_loss=0.03408, over 944034.00 frames. +2023-04-03 01:02:59,925 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 01:03:03,177 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.276e+02 5.212e+02 6.445e+02 8.399e+02 3.393e+03, threshold=1.289e+03, percent-clipped=7.0 +2023-04-03 01:03:05,651 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=157048.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:03:12,283 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 01:04:00,820 INFO [train.py:903] (3/4) Epoch 24, batch 50, loss[loss=0.1723, simple_loss=0.2614, pruned_loss=0.04163, over 19775.00 frames. ], tot_loss[loss=0.2089, simple_loss=0.2896, pruned_loss=0.06414, over 858159.56 frames. ], batch size: 54, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:04:01,277 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1782, 2.1119, 1.9300, 1.7561, 1.7168, 1.7522, 0.5046, 1.0692], + device='cuda:3'), covar=tensor([0.0572, 0.0617, 0.0467, 0.0747, 0.1164, 0.0865, 0.1379, 0.1143], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0357, 0.0362, 0.0386, 0.0466, 0.0391, 0.0338, 0.0342], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:04:20,669 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157111.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:04:33,371 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 01:04:47,077 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4750, 1.4833, 2.1610, 1.7870, 3.2417, 4.8053, 4.6678, 5.1728], + device='cuda:3'), covar=tensor([0.1558, 0.3824, 0.3202, 0.2227, 0.0587, 0.0170, 0.0159, 0.0173], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0326, 0.0355, 0.0266, 0.0246, 0.0190, 0.0217, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 01:05:01,215 INFO [train.py:903] (3/4) Epoch 24, batch 100, loss[loss=0.1583, simple_loss=0.2471, pruned_loss=0.03477, over 19679.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2858, pruned_loss=0.0621, over 1532504.96 frames. ], batch size: 53, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:05:03,493 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.245e+02 5.500e+02 6.534e+02 8.918e+02 1.825e+03, threshold=1.307e+03, percent-clipped=7.0 +2023-04-03 01:05:11,342 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 01:05:19,853 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157160.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:05:26,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7541, 1.5426, 1.6989, 1.4793, 4.3215, 1.2748, 2.4603, 4.5444], + device='cuda:3'), covar=tensor([0.0483, 0.2808, 0.2904, 0.2098, 0.0736, 0.2558, 0.1623, 0.0218], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0368, 0.0390, 0.0351, 0.0374, 0.0350, 0.0383, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:06:02,079 INFO [train.py:903] (3/4) Epoch 24, batch 150, loss[loss=0.1934, simple_loss=0.2744, pruned_loss=0.0562, over 19676.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2835, pruned_loss=0.06051, over 2052476.99 frames. ], batch size: 53, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:06:42,476 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157226.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:07:01,360 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 01:07:02,476 INFO [train.py:903] (3/4) Epoch 24, batch 200, loss[loss=0.2095, simple_loss=0.3005, pruned_loss=0.05922, over 18838.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2845, pruned_loss=0.06163, over 2458751.09 frames. ], batch size: 74, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:07:04,623 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.193e+02 4.992e+02 5.973e+02 7.088e+02 2.080e+03, threshold=1.195e+03, percent-clipped=2.0 +2023-04-03 01:07:05,926 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157247.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:07:12,107 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6289, 1.3456, 1.2492, 1.5354, 1.1622, 1.3788, 1.2186, 1.4574], + device='cuda:3'), covar=tensor([0.1038, 0.1113, 0.1576, 0.0989, 0.1267, 0.0606, 0.1515, 0.0825], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0356, 0.0314, 0.0252, 0.0304, 0.0253, 0.0312, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:08:03,064 INFO [train.py:903] (3/4) Epoch 24, batch 250, loss[loss=0.1968, simple_loss=0.2874, pruned_loss=0.05307, over 19722.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06281, over 2761979.74 frames. ], batch size: 59, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:08:45,261 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6623, 1.5441, 1.5781, 2.1332, 1.7925, 1.8648, 1.8510, 1.7077], + device='cuda:3'), covar=tensor([0.0808, 0.0900, 0.0989, 0.0695, 0.0777, 0.0744, 0.0842, 0.0681], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0223, 0.0227, 0.0241, 0.0228, 0.0215, 0.0190, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 01:09:03,293 INFO [train.py:903] (3/4) Epoch 24, batch 300, loss[loss=0.2069, simple_loss=0.292, pruned_loss=0.06087, over 19615.00 frames. ], tot_loss[loss=0.207, simple_loss=0.287, pruned_loss=0.06347, over 3001104.85 frames. ], batch size: 57, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:09:06,236 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.399e+02 5.405e+02 6.557e+02 9.024e+02 1.464e+03, threshold=1.311e+03, percent-clipped=9.0 +2023-04-03 01:09:25,613 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157362.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:09:34,511 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4595, 1.0718, 1.3481, 1.0866, 2.1082, 0.9527, 2.0841, 2.3613], + device='cuda:3'), covar=tensor([0.1109, 0.3341, 0.3143, 0.2119, 0.1303, 0.2439, 0.1232, 0.0622], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0369, 0.0390, 0.0352, 0.0374, 0.0350, 0.0383, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:09:37,286 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157372.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:09:39,546 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157374.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:09:41,169 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-04-03 01:09:56,099 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3958, 2.1358, 1.6700, 1.3948, 1.9985, 1.3524, 1.3943, 1.8858], + device='cuda:3'), covar=tensor([0.0945, 0.0747, 0.0973, 0.0872, 0.0531, 0.1202, 0.0643, 0.0465], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0314, 0.0336, 0.0266, 0.0246, 0.0338, 0.0288, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:10:05,068 INFO [train.py:903] (3/4) Epoch 24, batch 350, loss[loss=0.1815, simple_loss=0.2754, pruned_loss=0.04378, over 19546.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2862, pruned_loss=0.06292, over 3188307.65 frames. ], batch size: 54, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:10:10,686 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 01:10:12,265 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157400.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:10:39,435 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157422.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:11:05,175 INFO [train.py:903] (3/4) Epoch 24, batch 400, loss[loss=0.2167, simple_loss=0.297, pruned_loss=0.06817, over 19323.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2865, pruned_loss=0.06288, over 3341503.92 frames. ], batch size: 66, lr: 3.46e-03, grad_scale: 8.0 +2023-04-03 01:11:07,648 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.933e+02 4.825e+02 6.674e+02 8.153e+02 1.427e+03, threshold=1.335e+03, percent-clipped=2.0 +2023-04-03 01:11:48,277 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1789, 2.0806, 2.0035, 1.8463, 1.6772, 1.7551, 0.6203, 1.0766], + device='cuda:3'), covar=tensor([0.0595, 0.0604, 0.0404, 0.0699, 0.1126, 0.0859, 0.1315, 0.1082], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0354, 0.0361, 0.0384, 0.0463, 0.0392, 0.0337, 0.0340], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:11:52,611 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=157482.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:11:56,178 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5695, 1.7419, 1.7691, 2.2613, 1.7384, 2.1647, 1.8690, 1.5102], + device='cuda:3'), covar=tensor([0.5467, 0.4433, 0.3061, 0.2907, 0.4328, 0.2534, 0.6909, 0.5602], + device='cuda:3'), in_proj_covar=tensor([0.0908, 0.0977, 0.0723, 0.0934, 0.0888, 0.0825, 0.0847, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 01:11:58,254 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157487.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:12:01,775 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1328, 1.8387, 1.5362, 1.2874, 1.6757, 1.3161, 1.1625, 1.5878], + device='cuda:3'), covar=tensor([0.0854, 0.0812, 0.1037, 0.0844, 0.0536, 0.1282, 0.0629, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0338, 0.0267, 0.0247, 0.0339, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:12:05,707 INFO [train.py:903] (3/4) Epoch 24, batch 450, loss[loss=0.2129, simple_loss=0.286, pruned_loss=0.06985, over 19736.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2859, pruned_loss=0.06256, over 3455265.60 frames. ], batch size: 51, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:12:20,127 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157504.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:12:24,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=157507.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:12:31,978 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7193, 1.7500, 1.6297, 1.4001, 1.4195, 1.4531, 0.2526, 0.7075], + device='cuda:3'), covar=tensor([0.0740, 0.0681, 0.0448, 0.0693, 0.1342, 0.0798, 0.1318, 0.1172], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0354, 0.0362, 0.0384, 0.0464, 0.0392, 0.0338, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:12:40,680 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 01:12:40,717 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 01:12:43,480 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1424, 1.8191, 1.4887, 1.2804, 1.6093, 1.2801, 1.2075, 1.6362], + device='cuda:3'), covar=tensor([0.0870, 0.0844, 0.1186, 0.0904, 0.0619, 0.1391, 0.0652, 0.0529], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0339, 0.0268, 0.0248, 0.0340, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:12:44,468 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157524.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 01:13:08,946 INFO [train.py:903] (3/4) Epoch 24, batch 500, loss[loss=0.1996, simple_loss=0.2828, pruned_loss=0.05819, over 19667.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2857, pruned_loss=0.06268, over 3539829.90 frames. ], batch size: 60, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:13:12,128 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.617e+02 5.203e+02 6.096e+02 8.720e+02 1.456e+03, threshold=1.219e+03, percent-clipped=3.0 +2023-04-03 01:14:02,846 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157586.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:14:12,214 INFO [train.py:903] (3/4) Epoch 24, batch 550, loss[loss=0.2466, simple_loss=0.3145, pruned_loss=0.08934, over 18003.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2864, pruned_loss=0.06288, over 3603659.67 frames. ], batch size: 83, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:14:32,907 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4574, 1.5482, 1.7470, 1.6973, 2.6390, 2.2656, 2.7268, 1.0277], + device='cuda:3'), covar=tensor([0.2562, 0.4290, 0.2818, 0.1983, 0.1487, 0.2150, 0.1451, 0.4764], + device='cuda:3'), in_proj_covar=tensor([0.0538, 0.0647, 0.0721, 0.0491, 0.0620, 0.0534, 0.0661, 0.0552], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:14:41,087 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=157618.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:14:42,244 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157619.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:15:13,033 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=157643.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:15:13,772 INFO [train.py:903] (3/4) Epoch 24, batch 600, loss[loss=0.2525, simple_loss=0.3309, pruned_loss=0.08708, over 18313.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2863, pruned_loss=0.06273, over 3665289.88 frames. ], batch size: 83, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:15:15,912 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.272e+02 4.646e+02 5.574e+02 6.767e+02 1.170e+03, threshold=1.115e+03, percent-clipped=0.0 +2023-04-03 01:15:28,544 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4105, 3.9972, 2.6672, 3.5401, 0.9228, 3.9714, 3.8642, 3.9271], + device='cuda:3'), covar=tensor([0.0661, 0.1094, 0.1965, 0.0935, 0.3953, 0.0780, 0.0885, 0.1234], + device='cuda:3'), in_proj_covar=tensor([0.0515, 0.0415, 0.0498, 0.0349, 0.0402, 0.0439, 0.0432, 0.0468], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:15:29,132 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.83 vs. limit=5.0 +2023-04-03 01:15:39,900 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0069, 4.4105, 4.6949, 4.7232, 1.7663, 4.4423, 3.8793, 4.4168], + device='cuda:3'), covar=tensor([0.1639, 0.0788, 0.0586, 0.0666, 0.6092, 0.0787, 0.0655, 0.1130], + device='cuda:3'), in_proj_covar=tensor([0.0792, 0.0757, 0.0965, 0.0844, 0.0850, 0.0731, 0.0580, 0.0889], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 01:15:53,094 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 01:16:14,773 INFO [train.py:903] (3/4) Epoch 24, batch 650, loss[loss=0.2065, simple_loss=0.2939, pruned_loss=0.05952, over 19655.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06274, over 3707075.12 frames. ], batch size: 60, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:16:45,986 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157718.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:17:14,721 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=157743.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:17:15,392 INFO [train.py:903] (3/4) Epoch 24, batch 700, loss[loss=0.1912, simple_loss=0.2671, pruned_loss=0.05763, over 19493.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2859, pruned_loss=0.06277, over 3721498.15 frames. ], batch size: 49, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:17:15,554 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157744.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:17:20,750 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.032e+02 5.232e+02 6.869e+02 8.195e+02 1.483e+03, threshold=1.374e+03, percent-clipped=7.0 +2023-04-03 01:17:44,071 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157766.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:17:46,339 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=157768.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:18:19,086 INFO [train.py:903] (3/4) Epoch 24, batch 750, loss[loss=0.2476, simple_loss=0.3276, pruned_loss=0.0838, over 19320.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2877, pruned_loss=0.06359, over 3746649.65 frames. ], batch size: 70, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:18:31,207 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.28 vs. limit=5.0 +2023-04-03 01:19:06,846 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157833.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:19:19,800 INFO [train.py:903] (3/4) Epoch 24, batch 800, loss[loss=0.2005, simple_loss=0.2714, pruned_loss=0.06484, over 19396.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2872, pruned_loss=0.06312, over 3774412.23 frames. ], batch size: 48, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:19:23,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.261e+02 5.074e+02 6.358e+02 8.526e+02 1.766e+03, threshold=1.272e+03, percent-clipped=4.0 +2023-04-03 01:19:30,234 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 01:19:37,379 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157859.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:19:37,397 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6787, 1.5164, 1.5556, 2.2687, 1.7100, 1.8840, 1.8914, 1.7418], + device='cuda:3'), covar=tensor([0.0867, 0.0948, 0.1051, 0.0725, 0.0880, 0.0851, 0.0923, 0.0716], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0224, 0.0229, 0.0242, 0.0229, 0.0215, 0.0191, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 01:19:48,547 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157868.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:19:52,870 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=157871.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:19:59,192 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=157875.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:20:05,682 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157881.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:20:20,146 INFO [train.py:903] (3/4) Epoch 24, batch 850, loss[loss=0.2223, simple_loss=0.2967, pruned_loss=0.07389, over 19678.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2862, pruned_loss=0.06254, over 3788593.71 frames. ], batch size: 60, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:20:27,479 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=157900.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:20:27,510 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3418, 3.0320, 2.4581, 2.4127, 2.1761, 2.6301, 1.0075, 2.2321], + device='cuda:3'), covar=tensor([0.0702, 0.0649, 0.0692, 0.1025, 0.1164, 0.1072, 0.1460, 0.1131], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0357, 0.0363, 0.0388, 0.0467, 0.0394, 0.0339, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:20:37,623 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.56 vs. limit=2.0 +2023-04-03 01:21:05,431 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=157930.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:21:08,630 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 01:21:21,347 INFO [train.py:903] (3/4) Epoch 24, batch 900, loss[loss=0.2241, simple_loss=0.3016, pruned_loss=0.07329, over 19615.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2858, pruned_loss=0.06261, over 3785672.02 frames. ], batch size: 57, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:21:25,793 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.893e+02 4.800e+02 5.808e+02 7.910e+02 1.683e+03, threshold=1.162e+03, percent-clipped=5.0 +2023-04-03 01:22:10,200 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=157983.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 01:22:22,968 INFO [train.py:903] (3/4) Epoch 24, batch 950, loss[loss=0.2143, simple_loss=0.2927, pruned_loss=0.06793, over 19780.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2858, pruned_loss=0.06264, over 3785665.50 frames. ], batch size: 56, lr: 3.45e-03, grad_scale: 4.0 +2023-04-03 01:22:23,033 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 01:23:26,864 INFO [train.py:903] (3/4) Epoch 24, batch 1000, loss[loss=0.1748, simple_loss=0.2535, pruned_loss=0.04806, over 19360.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2851, pruned_loss=0.06221, over 3799757.23 frames. ], batch size: 47, lr: 3.45e-03, grad_scale: 4.0 +2023-04-03 01:23:28,219 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=158045.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:23:31,294 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.770e+02 5.075e+02 5.979e+02 8.043e+02 1.884e+03, threshold=1.196e+03, percent-clipped=5.0 +2023-04-03 01:24:17,498 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 01:24:22,286 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158089.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:24:22,533 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 01:24:27,542 INFO [train.py:903] (3/4) Epoch 24, batch 1050, loss[loss=0.1843, simple_loss=0.2708, pruned_loss=0.04896, over 19513.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2852, pruned_loss=0.06245, over 3806325.12 frames. ], batch size: 64, lr: 3.45e-03, grad_scale: 4.0 +2023-04-03 01:24:51,171 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158114.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:24:52,308 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158115.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:24:56,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 01:25:07,407 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9229, 1.6818, 1.8557, 1.6790, 4.4357, 1.2577, 2.7224, 4.8454], + device='cuda:3'), covar=tensor([0.0453, 0.2771, 0.2803, 0.2044, 0.0747, 0.2648, 0.1300, 0.0176], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0373, 0.0394, 0.0354, 0.0377, 0.0354, 0.0389, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:25:20,113 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158137.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 01:25:23,288 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=158140.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:25:23,375 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158140.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:25:23,660 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-03 01:25:27,551 INFO [train.py:903] (3/4) Epoch 24, batch 1100, loss[loss=0.2196, simple_loss=0.3027, pruned_loss=0.06823, over 19648.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2861, pruned_loss=0.06329, over 3805180.89 frames. ], batch size: 55, lr: 3.45e-03, grad_scale: 4.0 +2023-04-03 01:25:31,914 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.699e+02 5.122e+02 6.777e+02 7.992e+02 2.032e+03, threshold=1.355e+03, percent-clipped=5.0 +2023-04-03 01:25:51,780 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158162.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:26:28,900 INFO [train.py:903] (3/4) Epoch 24, batch 1150, loss[loss=0.2164, simple_loss=0.3104, pruned_loss=0.06115, over 19382.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2862, pruned_loss=0.0631, over 3813461.40 frames. ], batch size: 70, lr: 3.45e-03, grad_scale: 4.0 +2023-04-03 01:26:56,913 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=158215.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:26:58,667 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.70 vs. limit=5.0 +2023-04-03 01:27:25,377 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158239.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:27:31,331 INFO [train.py:903] (3/4) Epoch 24, batch 1200, loss[loss=0.1755, simple_loss=0.2565, pruned_loss=0.04718, over 19393.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2854, pruned_loss=0.06295, over 3819917.14 frames. ], batch size: 48, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:27:37,766 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.973e+02 4.926e+02 5.852e+02 7.782e+02 1.430e+03, threshold=1.170e+03, percent-clipped=2.0 +2023-04-03 01:27:56,353 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158264.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:28:02,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 01:28:34,709 INFO [train.py:903] (3/4) Epoch 24, batch 1250, loss[loss=0.196, simple_loss=0.2746, pruned_loss=0.05868, over 19487.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2865, pruned_loss=0.06332, over 3812177.66 frames. ], batch size: 49, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:28:43,367 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158301.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:29:15,088 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158326.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:29:20,361 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=158330.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:29:35,685 INFO [train.py:903] (3/4) Epoch 24, batch 1300, loss[loss=0.1988, simple_loss=0.2843, pruned_loss=0.05669, over 19529.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2872, pruned_loss=0.06357, over 3810694.08 frames. ], batch size: 56, lr: 3.45e-03, grad_scale: 8.0 +2023-04-03 01:29:40,390 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.859e+02 5.283e+02 7.241e+02 8.945e+02 2.355e+03, threshold=1.448e+03, percent-clipped=9.0 +2023-04-03 01:30:24,672 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5157, 1.5960, 1.7660, 1.9852, 1.5531, 1.9524, 1.8152, 1.7443], + device='cuda:3'), covar=tensor([0.3299, 0.2945, 0.1567, 0.1746, 0.2977, 0.1620, 0.3538, 0.2498], + device='cuda:3'), in_proj_covar=tensor([0.0909, 0.0980, 0.0724, 0.0938, 0.0889, 0.0825, 0.0843, 0.0788], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 01:30:36,952 INFO [train.py:903] (3/4) Epoch 24, batch 1350, loss[loss=0.1842, simple_loss=0.2739, pruned_loss=0.04727, over 19526.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2872, pruned_loss=0.06366, over 3799608.62 frames. ], batch size: 64, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:30:48,163 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=158402.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:31:38,843 INFO [train.py:903] (3/4) Epoch 24, batch 1400, loss[loss=0.1818, simple_loss=0.2739, pruned_loss=0.04484, over 19609.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2876, pruned_loss=0.06379, over 3808269.97 frames. ], batch size: 57, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:31:43,422 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.414e+02 5.106e+02 6.656e+02 8.188e+02 2.197e+03, threshold=1.331e+03, percent-clipped=4.0 +2023-04-03 01:32:27,057 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=158484.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:32:40,276 INFO [train.py:903] (3/4) Epoch 24, batch 1450, loss[loss=0.2068, simple_loss=0.2983, pruned_loss=0.05766, over 19701.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2882, pruned_loss=0.06389, over 3818122.81 frames. ], batch size: 59, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:32:40,706 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6893, 1.2535, 1.3228, 1.5440, 1.0964, 1.4517, 1.2777, 1.5098], + device='cuda:3'), covar=tensor([0.1113, 0.1192, 0.1637, 0.1016, 0.1393, 0.0635, 0.1592, 0.0852], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0357, 0.0315, 0.0254, 0.0306, 0.0254, 0.0314, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:32:42,618 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 01:32:52,791 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4481, 1.4917, 1.7291, 1.6817, 2.4287, 2.1430, 2.6513, 1.0526], + device='cuda:3'), covar=tensor([0.2499, 0.4292, 0.2622, 0.1896, 0.1532, 0.2146, 0.1326, 0.4593], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0652, 0.0725, 0.0495, 0.0623, 0.0538, 0.0665, 0.0558], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:33:41,525 INFO [train.py:903] (3/4) Epoch 24, batch 1500, loss[loss=0.1881, simple_loss=0.277, pruned_loss=0.04961, over 19674.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2872, pruned_loss=0.0633, over 3818423.25 frames. ], batch size: 60, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:33:46,125 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.633e+02 4.870e+02 5.968e+02 7.477e+02 1.869e+03, threshold=1.194e+03, percent-clipped=2.0 +2023-04-03 01:34:33,946 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158586.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:34:42,446 INFO [train.py:903] (3/4) Epoch 24, batch 1550, loss[loss=0.2106, simple_loss=0.2806, pruned_loss=0.0703, over 19407.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2873, pruned_loss=0.06342, over 3819203.56 frames. ], batch size: 48, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:34:48,575 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=158599.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:35:04,356 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158611.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:35:36,448 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-04-03 01:35:46,044 INFO [train.py:903] (3/4) Epoch 24, batch 1600, loss[loss=0.2094, simple_loss=0.2902, pruned_loss=0.06435, over 19660.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2883, pruned_loss=0.06358, over 3815152.71 frames. ], batch size: 55, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:35:51,809 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.433e+02 4.721e+02 6.000e+02 7.264e+02 1.836e+03, threshold=1.200e+03, percent-clipped=4.0 +2023-04-03 01:36:12,729 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 01:36:50,304 INFO [train.py:903] (3/4) Epoch 24, batch 1650, loss[loss=0.2247, simple_loss=0.3042, pruned_loss=0.0726, over 19342.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.287, pruned_loss=0.06334, over 3809269.13 frames. ], batch size: 66, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:37:02,692 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1182, 1.3233, 1.5817, 1.3647, 2.7372, 1.1138, 2.2464, 3.0894], + device='cuda:3'), covar=tensor([0.0576, 0.2691, 0.2714, 0.1864, 0.0724, 0.2382, 0.1220, 0.0322], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0374, 0.0394, 0.0354, 0.0378, 0.0356, 0.0389, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:37:52,883 INFO [train.py:903] (3/4) Epoch 24, batch 1700, loss[loss=0.2846, simple_loss=0.3524, pruned_loss=0.1085, over 18117.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2867, pruned_loss=0.06296, over 3807900.51 frames. ], batch size: 83, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:37:55,358 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=158746.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:37:57,371 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.031e+02 4.745e+02 5.793e+02 7.168e+02 1.456e+03, threshold=1.159e+03, percent-clipped=2.0 +2023-04-03 01:38:35,593 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 01:38:54,543 INFO [train.py:903] (3/4) Epoch 24, batch 1750, loss[loss=0.2229, simple_loss=0.303, pruned_loss=0.07138, over 18242.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2873, pruned_loss=0.06325, over 3808142.33 frames. ], batch size: 83, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:38:56,127 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=158795.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:39:03,155 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=158801.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:39:57,550 INFO [train.py:903] (3/4) Epoch 24, batch 1800, loss[loss=0.1918, simple_loss=0.2637, pruned_loss=0.05991, over 19463.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2873, pruned_loss=0.06305, over 3813954.07 frames. ], batch size: 49, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:40:02,407 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.024e+02 4.690e+02 5.979e+02 7.282e+02 2.087e+03, threshold=1.196e+03, percent-clipped=3.0 +2023-04-03 01:40:08,504 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=158853.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:40:11,857 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=158855.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:40:19,307 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=158861.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:40:40,526 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=158880.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:40:56,889 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 01:40:58,003 INFO [train.py:903] (3/4) Epoch 24, batch 1850, loss[loss=0.2102, simple_loss=0.2974, pruned_loss=0.06145, over 19705.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2867, pruned_loss=0.06326, over 3825149.79 frames. ], batch size: 59, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:41:33,337 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 01:42:00,400 INFO [train.py:903] (3/4) Epoch 24, batch 1900, loss[loss=0.2016, simple_loss=0.2869, pruned_loss=0.05814, over 19733.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2866, pruned_loss=0.06343, over 3809274.11 frames. ], batch size: 51, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:42:04,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.445e+02 4.936e+02 5.873e+02 7.509e+02 2.125e+03, threshold=1.175e+03, percent-clipped=8.0 +2023-04-03 01:42:18,814 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 01:42:24,109 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 01:42:41,454 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6617, 4.2512, 2.5699, 3.7820, 1.1490, 4.2233, 4.0917, 4.1871], + device='cuda:3'), covar=tensor([0.0584, 0.0999, 0.2011, 0.0826, 0.3689, 0.0625, 0.0848, 0.1170], + device='cuda:3'), in_proj_covar=tensor([0.0511, 0.0414, 0.0496, 0.0345, 0.0401, 0.0437, 0.0430, 0.0462], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:42:46,677 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 01:43:00,389 INFO [train.py:903] (3/4) Epoch 24, batch 1950, loss[loss=0.1659, simple_loss=0.2447, pruned_loss=0.04354, over 19742.00 frames. ], tot_loss[loss=0.207, simple_loss=0.287, pruned_loss=0.06351, over 3818238.38 frames. ], batch size: 45, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:44:03,844 INFO [train.py:903] (3/4) Epoch 24, batch 2000, loss[loss=0.2316, simple_loss=0.3043, pruned_loss=0.07944, over 13459.00 frames. ], tot_loss[loss=0.2074, simple_loss=0.2873, pruned_loss=0.06371, over 3806739.24 frames. ], batch size: 135, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:44:08,592 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.425e+02 5.106e+02 6.605e+02 9.481e+02 1.726e+03, threshold=1.321e+03, percent-clipped=5.0 +2023-04-03 01:44:57,706 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159087.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:44:59,847 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 01:45:03,909 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.63 vs. limit=5.0 +2023-04-03 01:45:06,688 INFO [train.py:903] (3/4) Epoch 24, batch 2050, loss[loss=0.1727, simple_loss=0.2497, pruned_loss=0.04783, over 19754.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2875, pruned_loss=0.0641, over 3794442.69 frames. ], batch size: 48, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:45:19,179 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 01:45:20,292 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 01:45:35,014 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159117.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:45:41,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 01:46:02,782 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159139.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:46:03,045 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1930, 2.0643, 1.9267, 1.8356, 1.6159, 1.7877, 0.7084, 1.2497], + device='cuda:3'), covar=tensor([0.0574, 0.0627, 0.0501, 0.0765, 0.1102, 0.0856, 0.1282, 0.0976], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0358, 0.0362, 0.0387, 0.0463, 0.0392, 0.0339, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 01:46:06,458 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159142.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:46:08,193 INFO [train.py:903] (3/4) Epoch 24, batch 2100, loss[loss=0.2245, simple_loss=0.2918, pruned_loss=0.0786, over 19760.00 frames. ], tot_loss[loss=0.2076, simple_loss=0.2872, pruned_loss=0.06403, over 3794807.63 frames. ], batch size: 51, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:46:10,463 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159145.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:46:13,634 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.527e+02 4.786e+02 5.922e+02 8.131e+02 1.881e+03, threshold=1.184e+03, percent-clipped=4.0 +2023-04-03 01:46:37,332 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 01:46:49,821 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159177.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:46:58,694 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 01:47:10,232 INFO [train.py:903] (3/4) Epoch 24, batch 2150, loss[loss=0.1943, simple_loss=0.277, pruned_loss=0.05577, over 19840.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2877, pruned_loss=0.06401, over 3816874.03 frames. ], batch size: 52, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:47:13,885 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159197.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 01:48:12,824 INFO [train.py:903] (3/4) Epoch 24, batch 2200, loss[loss=0.2471, simple_loss=0.3124, pruned_loss=0.09087, over 19611.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2881, pruned_loss=0.0641, over 3823510.62 frames. ], batch size: 50, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:48:18,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.316e+02 4.881e+02 6.157e+02 7.813e+02 2.191e+03, threshold=1.231e+03, percent-clipped=6.0 +2023-04-03 01:48:25,324 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159254.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:48:33,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159260.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:48:41,424 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8086, 1.4312, 1.6273, 1.4111, 3.2793, 0.9849, 2.3827, 3.8389], + device='cuda:3'), covar=tensor([0.0594, 0.3063, 0.2983, 0.2327, 0.0892, 0.3067, 0.1546, 0.0256], + device='cuda:3'), in_proj_covar=tensor([0.0418, 0.0374, 0.0395, 0.0354, 0.0379, 0.0357, 0.0391, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:49:14,285 INFO [train.py:903] (3/4) Epoch 24, batch 2250, loss[loss=0.1918, simple_loss=0.2809, pruned_loss=0.05138, over 19658.00 frames. ], tot_loss[loss=0.2084, simple_loss=0.2881, pruned_loss=0.06435, over 3828703.47 frames. ], batch size: 55, lr: 3.44e-03, grad_scale: 8.0 +2023-04-03 01:49:31,873 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159308.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:49:37,404 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159312.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 01:49:46,545 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8790, 1.3058, 1.5588, 1.5112, 3.4258, 0.9968, 2.5885, 3.8902], + device='cuda:3'), covar=tensor([0.0489, 0.3057, 0.3007, 0.1997, 0.0732, 0.2814, 0.1237, 0.0213], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0373, 0.0394, 0.0354, 0.0378, 0.0356, 0.0390, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:50:16,678 INFO [train.py:903] (3/4) Epoch 24, batch 2300, loss[loss=0.1874, simple_loss=0.2699, pruned_loss=0.0524, over 19620.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2873, pruned_loss=0.06402, over 3834939.29 frames. ], batch size: 57, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:50:21,055 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.893e+02 4.864e+02 6.208e+02 8.672e+02 1.812e+03, threshold=1.242e+03, percent-clipped=10.0 +2023-04-03 01:50:31,271 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 01:50:52,353 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7946, 1.3673, 1.5575, 1.6218, 3.3525, 1.0200, 2.5639, 3.8901], + device='cuda:3'), covar=tensor([0.0471, 0.2714, 0.2828, 0.1769, 0.0688, 0.2641, 0.1168, 0.0202], + device='cuda:3'), in_proj_covar=tensor([0.0415, 0.0370, 0.0391, 0.0351, 0.0375, 0.0354, 0.0387, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:51:19,170 INFO [train.py:903] (3/4) Epoch 24, batch 2350, loss[loss=0.2302, simple_loss=0.3182, pruned_loss=0.0711, over 19680.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2866, pruned_loss=0.06358, over 3815609.88 frames. ], batch size: 60, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:52:00,171 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 01:52:04,794 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159431.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:52:17,048 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 01:52:19,433 INFO [train.py:903] (3/4) Epoch 24, batch 2400, loss[loss=0.1726, simple_loss=0.2646, pruned_loss=0.04029, over 19674.00 frames. ], tot_loss[loss=0.2069, simple_loss=0.2868, pruned_loss=0.06356, over 3814945.16 frames. ], batch size: 53, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:52:25,062 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.677e+02 4.946e+02 5.943e+02 8.368e+02 2.189e+03, threshold=1.189e+03, percent-clipped=6.0 +2023-04-03 01:53:22,618 INFO [train.py:903] (3/4) Epoch 24, batch 2450, loss[loss=0.2053, simple_loss=0.2988, pruned_loss=0.05588, over 17423.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2866, pruned_loss=0.06328, over 3804258.95 frames. ], batch size: 101, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:53:42,121 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159510.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:53:49,979 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159516.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:53:55,525 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159521.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:54:14,783 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159535.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:54:21,721 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159541.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:54:24,656 INFO [train.py:903] (3/4) Epoch 24, batch 2500, loss[loss=0.2241, simple_loss=0.3034, pruned_loss=0.07244, over 19244.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2868, pruned_loss=0.06341, over 3806866.26 frames. ], batch size: 66, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:54:27,421 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159546.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:54:29,312 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.080e+02 4.941e+02 6.110e+02 7.649e+02 1.406e+03, threshold=1.222e+03, percent-clipped=1.0 +2023-04-03 01:54:55,485 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159568.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:55:25,732 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159593.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 01:55:26,480 INFO [train.py:903] (3/4) Epoch 24, batch 2550, loss[loss=0.2231, simple_loss=0.3034, pruned_loss=0.07139, over 18794.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2871, pruned_loss=0.06356, over 3805798.23 frames. ], batch size: 74, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:56:19,878 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159636.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:56:20,606 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 01:56:29,399 INFO [train.py:903] (3/4) Epoch 24, batch 2600, loss[loss=0.1924, simple_loss=0.2675, pruned_loss=0.05866, over 19758.00 frames. ], tot_loss[loss=0.2073, simple_loss=0.2869, pruned_loss=0.06382, over 3793213.92 frames. ], batch size: 45, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:56:34,434 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 01:56:34,953 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.356e+02 4.779e+02 5.928e+02 8.262e+02 1.528e+03, threshold=1.186e+03, percent-clipped=6.0 +2023-04-03 01:56:36,554 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159649.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:56:39,987 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159652.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:57:31,593 INFO [train.py:903] (3/4) Epoch 24, batch 2650, loss[loss=0.1874, simple_loss=0.2738, pruned_loss=0.05043, over 19292.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2873, pruned_loss=0.06428, over 3765018.55 frames. ], batch size: 66, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:57:43,896 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 01:58:18,682 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2253, 1.2838, 1.8651, 1.2722, 2.7214, 3.7057, 3.3846, 3.8827], + device='cuda:3'), covar=tensor([0.1629, 0.3916, 0.3225, 0.2482, 0.0583, 0.0189, 0.0228, 0.0262], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0325, 0.0356, 0.0264, 0.0245, 0.0190, 0.0216, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 01:58:20,963 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8659, 1.3468, 1.0640, 0.9267, 1.1646, 0.9684, 0.9160, 1.2577], + device='cuda:3'), covar=tensor([0.0645, 0.0842, 0.1188, 0.0907, 0.0671, 0.1418, 0.0630, 0.0518], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0315, 0.0339, 0.0268, 0.0248, 0.0339, 0.0292, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:58:34,703 INFO [train.py:903] (3/4) Epoch 24, batch 2700, loss[loss=0.2283, simple_loss=0.3096, pruned_loss=0.07348, over 19476.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2879, pruned_loss=0.0643, over 3791239.16 frames. ], batch size: 64, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:58:39,053 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.760e+02 5.237e+02 6.508e+02 8.466e+02 2.382e+03, threshold=1.302e+03, percent-clipped=8.0 +2023-04-03 01:59:02,798 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4437, 2.1827, 1.7547, 1.4712, 1.9976, 1.4363, 1.2872, 1.8757], + device='cuda:3'), covar=tensor([0.0967, 0.0746, 0.0982, 0.0848, 0.0608, 0.1218, 0.0751, 0.0488], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0315, 0.0340, 0.0269, 0.0248, 0.0340, 0.0292, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 01:59:03,870 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=159767.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 01:59:36,001 INFO [train.py:903] (3/4) Epoch 24, batch 2750, loss[loss=0.1775, simple_loss=0.2537, pruned_loss=0.05069, over 19759.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2876, pruned_loss=0.06424, over 3806587.19 frames. ], batch size: 47, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 01:59:46,851 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159802.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:00:19,561 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159827.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:00:40,383 INFO [train.py:903] (3/4) Epoch 24, batch 2800, loss[loss=0.1875, simple_loss=0.2622, pruned_loss=0.05637, over 16504.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2877, pruned_loss=0.06412, over 3800286.97 frames. ], batch size: 36, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:00:45,933 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.444e+02 4.661e+02 5.641e+02 7.181e+02 2.352e+03, threshold=1.128e+03, percent-clipped=2.0 +2023-04-03 02:01:16,166 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159873.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:01:40,888 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=159892.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:01:42,763 INFO [train.py:903] (3/4) Epoch 24, batch 2850, loss[loss=0.1781, simple_loss=0.2514, pruned_loss=0.05243, over 19785.00 frames. ], tot_loss[loss=0.208, simple_loss=0.2874, pruned_loss=0.06427, over 3812065.52 frames. ], batch size: 47, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:02:10,720 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=159917.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:02:25,870 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=159928.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:02:39,627 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 02:02:45,326 INFO [train.py:903] (3/4) Epoch 24, batch 2900, loss[loss=0.1963, simple_loss=0.283, pruned_loss=0.05481, over 19680.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2874, pruned_loss=0.06426, over 3807015.08 frames. ], batch size: 59, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:02:51,070 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.210e+02 4.874e+02 6.501e+02 8.672e+02 1.518e+03, threshold=1.300e+03, percent-clipped=5.0 +2023-04-03 02:03:45,537 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=159993.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:03:46,480 INFO [train.py:903] (3/4) Epoch 24, batch 2950, loss[loss=0.1988, simple_loss=0.277, pruned_loss=0.06035, over 19745.00 frames. ], tot_loss[loss=0.2082, simple_loss=0.2876, pruned_loss=0.06441, over 3799008.74 frames. ], batch size: 51, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:04:13,447 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1832, 2.0067, 1.8948, 2.0828, 1.9120, 1.8578, 1.7530, 2.0625], + device='cuda:3'), covar=tensor([0.1042, 0.1462, 0.1430, 0.1239, 0.1456, 0.0555, 0.1409, 0.0745], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0355, 0.0312, 0.0254, 0.0303, 0.0252, 0.0312, 0.0257], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:04:24,694 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=160023.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:04:45,842 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 02:04:49,466 INFO [train.py:903] (3/4) Epoch 24, batch 3000, loss[loss=0.1816, simple_loss=0.2665, pruned_loss=0.04837, over 19499.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2867, pruned_loss=0.06384, over 3799786.39 frames. ], batch size: 64, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:04:49,466 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 02:05:02,002 INFO [train.py:937] (3/4) Epoch 24, validation: loss=0.1679, simple_loss=0.268, pruned_loss=0.03397, over 944034.00 frames. +2023-04-03 02:05:02,004 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 02:05:08,149 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=160048.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:05:08,848 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.319e+02 4.966e+02 6.275e+02 7.790e+02 1.988e+03, threshold=1.255e+03, percent-clipped=5.0 +2023-04-03 02:05:33,081 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-04-03 02:05:47,965 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.28 vs. limit=5.0 +2023-04-03 02:06:04,586 INFO [train.py:903] (3/4) Epoch 24, batch 3050, loss[loss=0.2386, simple_loss=0.3156, pruned_loss=0.08077, over 19662.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2863, pruned_loss=0.06294, over 3816956.29 frames. ], batch size: 60, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:06:19,319 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-03 02:06:23,602 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=160108.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:07:08,733 INFO [train.py:903] (3/4) Epoch 24, batch 3100, loss[loss=0.2096, simple_loss=0.2888, pruned_loss=0.06523, over 19674.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2866, pruned_loss=0.06289, over 3817691.81 frames. ], batch size: 53, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:07:14,564 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.956e+02 4.859e+02 5.894e+02 7.109e+02 1.682e+03, threshold=1.179e+03, percent-clipped=4.0 +2023-04-03 02:07:37,222 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1549, 1.4547, 1.7598, 1.5280, 2.9798, 4.6674, 4.4637, 5.0236], + device='cuda:3'), covar=tensor([0.1762, 0.3806, 0.3692, 0.2412, 0.0709, 0.0210, 0.0180, 0.0227], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0327, 0.0357, 0.0265, 0.0246, 0.0191, 0.0217, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 02:07:52,747 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4567, 1.5591, 1.8599, 1.7420, 2.7096, 2.3313, 2.9184, 1.2563], + device='cuda:3'), covar=tensor([0.2513, 0.4194, 0.2640, 0.1920, 0.1530, 0.2132, 0.1442, 0.4448], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0656, 0.0726, 0.0495, 0.0623, 0.0537, 0.0666, 0.0559], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 02:08:10,531 INFO [train.py:903] (3/4) Epoch 24, batch 3150, loss[loss=0.271, simple_loss=0.3417, pruned_loss=0.1001, over 19743.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2866, pruned_loss=0.0627, over 3821907.83 frames. ], batch size: 63, lr: 3.43e-03, grad_scale: 8.0 +2023-04-03 02:08:36,004 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 02:08:40,602 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=160217.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:08:56,834 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2827, 2.2859, 2.5030, 2.9633, 2.2466, 2.8329, 2.5816, 2.2909], + device='cuda:3'), covar=tensor([0.4117, 0.4261, 0.1949, 0.2714, 0.4500, 0.2331, 0.4747, 0.3434], + device='cuda:3'), in_proj_covar=tensor([0.0919, 0.0990, 0.0731, 0.0945, 0.0896, 0.0831, 0.0853, 0.0794], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 02:09:14,343 INFO [train.py:903] (3/4) Epoch 24, batch 3200, loss[loss=0.1951, simple_loss=0.287, pruned_loss=0.05163, over 18204.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06278, over 3829871.96 frames. ], batch size: 83, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:09:18,035 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6958, 4.2931, 2.8306, 3.7889, 0.9610, 4.1804, 4.1563, 4.1643], + device='cuda:3'), covar=tensor([0.0589, 0.0897, 0.1790, 0.0841, 0.3964, 0.0698, 0.0843, 0.0986], + device='cuda:3'), in_proj_covar=tensor([0.0512, 0.0413, 0.0498, 0.0346, 0.0403, 0.0437, 0.0431, 0.0464], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:09:19,294 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.6407, 0.8768, 0.6785, 0.6753, 0.7955, 0.6609, 0.6629, 0.8281], + device='cuda:3'), covar=tensor([0.0486, 0.0608, 0.0774, 0.0496, 0.0423, 0.0959, 0.0431, 0.0400], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0316, 0.0339, 0.0267, 0.0247, 0.0342, 0.0292, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:09:20,052 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.297e+02 5.134e+02 6.599e+02 8.700e+02 2.161e+03, threshold=1.320e+03, percent-clipped=4.0 +2023-04-03 02:09:49,740 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=160272.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:10:17,064 INFO [train.py:903] (3/4) Epoch 24, batch 3250, loss[loss=0.1686, simple_loss=0.2397, pruned_loss=0.04876, over 19032.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2868, pruned_loss=0.0628, over 3835472.17 frames. ], batch size: 42, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:10:56,755 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=160325.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 02:11:04,804 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=160332.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:11:21,817 INFO [train.py:903] (3/4) Epoch 24, batch 3300, loss[loss=0.2398, simple_loss=0.3129, pruned_loss=0.08339, over 19765.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2874, pruned_loss=0.06291, over 3841922.85 frames. ], batch size: 54, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:11:24,391 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 02:11:27,809 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 4.973e+02 5.834e+02 7.675e+02 1.997e+03, threshold=1.167e+03, percent-clipped=3.0 +2023-04-03 02:11:46,525 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=160364.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:12:15,770 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=160387.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:12:18,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=160389.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:12:23,805 INFO [train.py:903] (3/4) Epoch 24, batch 3350, loss[loss=0.1987, simple_loss=0.2791, pruned_loss=0.05909, over 19601.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.287, pruned_loss=0.06298, over 3844416.95 frames. ], batch size: 52, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:12:47,487 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5824, 4.0644, 4.2538, 4.2620, 1.6917, 4.0564, 3.5008, 4.0203], + device='cuda:3'), covar=tensor([0.1640, 0.0925, 0.0644, 0.0749, 0.5937, 0.0909, 0.0714, 0.1139], + device='cuda:3'), in_proj_covar=tensor([0.0787, 0.0753, 0.0959, 0.0839, 0.0842, 0.0731, 0.0572, 0.0891], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 02:13:26,804 INFO [train.py:903] (3/4) Epoch 24, batch 3400, loss[loss=0.2186, simple_loss=0.3029, pruned_loss=0.0671, over 19673.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2858, pruned_loss=0.0621, over 3846653.98 frames. ], batch size: 60, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:13:32,530 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.093e+02 5.133e+02 6.647e+02 9.203e+02 1.938e+03, threshold=1.329e+03, percent-clipped=8.0 +2023-04-03 02:14:28,012 INFO [train.py:903] (3/4) Epoch 24, batch 3450, loss[loss=0.2362, simple_loss=0.3172, pruned_loss=0.0776, over 19494.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2869, pruned_loss=0.06278, over 3841150.38 frames. ], batch size: 64, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:14:31,634 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 02:15:11,039 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-04-03 02:15:20,972 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9243, 4.4523, 2.7751, 3.8984, 1.0185, 4.4364, 4.2918, 4.4155], + device='cuda:3'), covar=tensor([0.0582, 0.0981, 0.1891, 0.0817, 0.4037, 0.0579, 0.0971, 0.1113], + device='cuda:3'), in_proj_covar=tensor([0.0509, 0.0413, 0.0498, 0.0345, 0.0402, 0.0437, 0.0431, 0.0462], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:15:29,545 INFO [train.py:903] (3/4) Epoch 24, batch 3500, loss[loss=0.2127, simple_loss=0.2952, pruned_loss=0.06512, over 19688.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2863, pruned_loss=0.06257, over 3834452.36 frames. ], batch size: 60, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:15:38,053 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.068e+02 4.685e+02 5.887e+02 7.904e+02 2.662e+03, threshold=1.177e+03, percent-clipped=4.0 +2023-04-03 02:16:26,902 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=160588.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:16:33,575 INFO [train.py:903] (3/4) Epoch 24, batch 3550, loss[loss=0.2029, simple_loss=0.2916, pruned_loss=0.05714, over 18808.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2864, pruned_loss=0.06239, over 3828230.85 frames. ], batch size: 74, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:16:39,948 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5700, 1.5185, 1.5209, 2.1932, 1.6626, 1.9815, 1.9254, 1.7214], + device='cuda:3'), covar=tensor([0.0862, 0.0906, 0.1011, 0.0611, 0.0802, 0.0701, 0.0820, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0222, 0.0228, 0.0239, 0.0225, 0.0213, 0.0189, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 02:16:57,364 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=160613.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:17:35,772 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.70 vs. limit=5.0 +2023-04-03 02:17:36,578 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=160643.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:17:37,258 INFO [train.py:903] (3/4) Epoch 24, batch 3600, loss[loss=0.2388, simple_loss=0.3099, pruned_loss=0.08387, over 19543.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2874, pruned_loss=0.06295, over 3830090.65 frames. ], batch size: 56, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:17:44,410 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.188e+02 4.811e+02 5.669e+02 7.209e+02 1.690e+03, threshold=1.134e+03, percent-clipped=3.0 +2023-04-03 02:18:08,006 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=160668.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:18:09,957 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=160669.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 02:18:40,044 INFO [train.py:903] (3/4) Epoch 24, batch 3650, loss[loss=0.1738, simple_loss=0.2602, pruned_loss=0.04376, over 19713.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2877, pruned_loss=0.06256, over 3842742.61 frames. ], batch size: 45, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:19:40,899 INFO [train.py:903] (3/4) Epoch 24, batch 3700, loss[loss=0.2548, simple_loss=0.3249, pruned_loss=0.09238, over 19331.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2878, pruned_loss=0.06321, over 3850721.45 frames. ], batch size: 66, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:19:46,834 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.47 vs. limit=2.0 +2023-04-03 02:19:49,425 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.079e+02 4.616e+02 6.347e+02 7.690e+02 1.972e+03, threshold=1.269e+03, percent-clipped=6.0 +2023-04-03 02:20:26,876 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0891, 1.9996, 1.7880, 2.0956, 1.8832, 1.7748, 1.7156, 1.9932], + device='cuda:3'), covar=tensor([0.1097, 0.1413, 0.1440, 0.1199, 0.1404, 0.0571, 0.1495, 0.0744], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0358, 0.0313, 0.0256, 0.0305, 0.0254, 0.0314, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:20:30,349 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=160784.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 02:20:44,492 INFO [train.py:903] (3/4) Epoch 24, batch 3750, loss[loss=0.201, simple_loss=0.2806, pruned_loss=0.06075, over 19584.00 frames. ], tot_loss[loss=0.2075, simple_loss=0.2879, pruned_loss=0.06354, over 3856206.90 frames. ], batch size: 52, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:21:45,655 INFO [train.py:903] (3/4) Epoch 24, batch 3800, loss[loss=0.2239, simple_loss=0.3082, pruned_loss=0.06975, over 19325.00 frames. ], tot_loss[loss=0.2068, simple_loss=0.2874, pruned_loss=0.06312, over 3853660.40 frames. ], batch size: 66, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:21:53,460 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.186e+02 4.635e+02 5.250e+02 7.046e+02 1.734e+03, threshold=1.050e+03, percent-clipped=4.0 +2023-04-03 02:22:18,606 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 02:22:45,256 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4782, 2.5589, 2.1222, 2.5783, 2.4586, 2.1034, 1.9254, 2.4871], + device='cuda:3'), covar=tensor([0.1026, 0.1399, 0.1475, 0.1035, 0.1252, 0.0550, 0.1491, 0.0703], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0356, 0.0313, 0.0255, 0.0305, 0.0254, 0.0314, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:22:47,092 INFO [train.py:903] (3/4) Epoch 24, batch 3850, loss[loss=0.1465, simple_loss=0.2315, pruned_loss=0.03076, over 19763.00 frames. ], tot_loss[loss=0.207, simple_loss=0.2874, pruned_loss=0.06329, over 3827691.34 frames. ], batch size: 48, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:22:58,402 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 02:23:18,509 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=160919.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:23:48,701 INFO [train.py:903] (3/4) Epoch 24, batch 3900, loss[loss=0.2382, simple_loss=0.3142, pruned_loss=0.08105, over 19513.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2868, pruned_loss=0.06302, over 3815418.52 frames. ], batch size: 64, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:23:58,397 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.922e+02 5.134e+02 6.381e+02 7.692e+02 1.884e+03, threshold=1.276e+03, percent-clipped=12.0 +2023-04-03 02:24:38,106 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=160983.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:24:53,492 INFO [train.py:903] (3/4) Epoch 24, batch 3950, loss[loss=0.2439, simple_loss=0.3181, pruned_loss=0.08483, over 13175.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.286, pruned_loss=0.06239, over 3814193.74 frames. ], batch size: 136, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:24:57,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 02:25:41,872 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5601, 1.5663, 1.7906, 1.9059, 4.1301, 1.3644, 2.6972, 4.3965], + device='cuda:3'), covar=tensor([0.0468, 0.2868, 0.2810, 0.1827, 0.0725, 0.2538, 0.1525, 0.0208], + device='cuda:3'), in_proj_covar=tensor([0.0412, 0.0367, 0.0389, 0.0350, 0.0373, 0.0353, 0.0385, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:25:51,245 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=161040.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 02:25:55,367 INFO [train.py:903] (3/4) Epoch 24, batch 4000, loss[loss=0.1785, simple_loss=0.2519, pruned_loss=0.05258, over 19098.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.285, pruned_loss=0.06194, over 3817677.33 frames. ], batch size: 42, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:26:03,425 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.651e+02 5.074e+02 6.327e+02 7.723e+02 1.762e+03, threshold=1.265e+03, percent-clipped=6.0 +2023-04-03 02:26:21,958 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=161065.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 02:26:41,782 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 02:26:57,369 INFO [train.py:903] (3/4) Epoch 24, batch 4050, loss[loss=0.2666, simple_loss=0.3258, pruned_loss=0.1037, over 13685.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2859, pruned_loss=0.06256, over 3818382.89 frames. ], batch size: 135, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:27:57,730 INFO [train.py:903] (3/4) Epoch 24, batch 4100, loss[loss=0.1771, simple_loss=0.2523, pruned_loss=0.05098, over 19379.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06284, over 3820199.40 frames. ], batch size: 47, lr: 3.42e-03, grad_scale: 8.0 +2023-04-03 02:28:06,049 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.204e+02 4.915e+02 6.129e+02 7.795e+02 1.333e+03, threshold=1.226e+03, percent-clipped=1.0 +2023-04-03 02:28:31,037 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 02:29:00,663 INFO [train.py:903] (3/4) Epoch 24, batch 4150, loss[loss=0.2061, simple_loss=0.2829, pruned_loss=0.06461, over 19619.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2867, pruned_loss=0.06331, over 3824255.15 frames. ], batch size: 50, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:29:49,592 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8463, 1.3628, 1.0743, 0.9407, 1.1702, 0.9896, 0.8648, 1.2083], + device='cuda:3'), covar=tensor([0.0645, 0.0830, 0.1174, 0.0811, 0.0611, 0.1347, 0.0655, 0.0529], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0315, 0.0337, 0.0267, 0.0247, 0.0340, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:29:50,580 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=161234.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:30:01,368 INFO [train.py:903] (3/4) Epoch 24, batch 4200, loss[loss=0.1847, simple_loss=0.2734, pruned_loss=0.04794, over 19769.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2878, pruned_loss=0.06398, over 3817967.12 frames. ], batch size: 54, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:30:02,568 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 02:30:04,062 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3282, 1.3094, 1.2803, 1.1349, 1.0635, 1.1818, 0.3734, 0.6444], + device='cuda:3'), covar=tensor([0.0519, 0.0527, 0.0328, 0.0503, 0.0845, 0.0576, 0.1196, 0.0830], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0361, 0.0365, 0.0388, 0.0466, 0.0395, 0.0342, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 02:30:08,530 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=161249.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:30:09,255 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.885e+02 4.684e+02 6.110e+02 7.845e+02 2.290e+03, threshold=1.222e+03, percent-clipped=7.0 +2023-04-03 02:30:24,271 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161263.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:31:03,249 INFO [train.py:903] (3/4) Epoch 24, batch 4250, loss[loss=0.2068, simple_loss=0.2924, pruned_loss=0.06065, over 19606.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2867, pruned_loss=0.06311, over 3820120.58 frames. ], batch size: 57, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:31:17,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 02:31:28,269 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 02:31:44,254 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161327.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:32:04,738 INFO [train.py:903] (3/4) Epoch 24, batch 4300, loss[loss=0.232, simple_loss=0.3099, pruned_loss=0.07704, over 18658.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2863, pruned_loss=0.06254, over 3832889.04 frames. ], batch size: 74, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:32:12,543 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.040e+02 4.585e+02 5.768e+02 7.257e+02 2.214e+03, threshold=1.154e+03, percent-clipped=5.0 +2023-04-03 02:32:38,073 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3117, 1.3156, 1.7718, 1.3867, 2.7168, 3.6771, 3.3747, 3.8776], + device='cuda:3'), covar=tensor([0.1545, 0.3771, 0.3294, 0.2423, 0.0614, 0.0212, 0.0223, 0.0253], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0325, 0.0356, 0.0265, 0.0245, 0.0190, 0.0217, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 02:32:47,563 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161378.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:32:57,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 02:33:06,258 INFO [train.py:903] (3/4) Epoch 24, batch 4350, loss[loss=0.2238, simple_loss=0.2974, pruned_loss=0.07512, over 19178.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2858, pruned_loss=0.0626, over 3821737.11 frames. ], batch size: 69, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:33:37,841 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=161419.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:34:07,070 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161442.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:34:08,991 INFO [train.py:903] (3/4) Epoch 24, batch 4400, loss[loss=0.1642, simple_loss=0.2488, pruned_loss=0.03981, over 19465.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2858, pruned_loss=0.06242, over 3807703.59 frames. ], batch size: 49, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:34:15,589 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.985e+02 5.009e+02 6.093e+02 7.233e+02 1.222e+03, threshold=1.219e+03, percent-clipped=3.0 +2023-04-03 02:34:31,952 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 02:34:41,589 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 02:34:43,223 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5434, 1.5605, 1.8369, 1.7470, 2.5831, 2.2868, 2.7505, 1.3530], + device='cuda:3'), covar=tensor([0.2391, 0.4214, 0.2656, 0.1894, 0.1520, 0.2032, 0.1401, 0.4242], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0651, 0.0725, 0.0493, 0.0623, 0.0535, 0.0661, 0.0556], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 02:34:50,984 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=161478.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:35:04,478 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7445, 4.2990, 2.7982, 3.8124, 1.0121, 4.3472, 4.1896, 4.2365], + device='cuda:3'), covar=tensor([0.0588, 0.0997, 0.1966, 0.0845, 0.4137, 0.0594, 0.0863, 0.1168], + device='cuda:3'), in_proj_covar=tensor([0.0511, 0.0418, 0.0503, 0.0349, 0.0402, 0.0438, 0.0435, 0.0465], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:35:10,531 INFO [train.py:903] (3/4) Epoch 24, batch 4450, loss[loss=0.2533, simple_loss=0.3357, pruned_loss=0.08548, over 18822.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2855, pruned_loss=0.0624, over 3813661.17 frames. ], batch size: 74, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:36:10,900 INFO [train.py:903] (3/4) Epoch 24, batch 4500, loss[loss=0.2325, simple_loss=0.3078, pruned_loss=0.07859, over 18134.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2867, pruned_loss=0.06313, over 3813850.26 frames. ], batch size: 83, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:36:17,765 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.067e+02 5.206e+02 6.185e+02 8.214e+02 2.130e+03, threshold=1.237e+03, percent-clipped=6.0 +2023-04-03 02:36:53,398 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161578.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:37:11,179 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161593.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:37:12,188 INFO [train.py:903] (3/4) Epoch 24, batch 4550, loss[loss=0.1688, simple_loss=0.2503, pruned_loss=0.04362, over 19766.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2859, pruned_loss=0.06259, over 3825903.15 frames. ], batch size: 46, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:37:20,028 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 02:37:44,741 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 02:38:02,414 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=161634.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:38:15,054 INFO [train.py:903] (3/4) Epoch 24, batch 4600, loss[loss=0.1805, simple_loss=0.2582, pruned_loss=0.05139, over 19374.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2852, pruned_loss=0.06234, over 3822640.79 frames. ], batch size: 47, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:38:21,972 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.192e+02 4.894e+02 5.872e+02 7.852e+02 1.807e+03, threshold=1.174e+03, percent-clipped=3.0 +2023-04-03 02:38:33,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=161659.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:39:03,287 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1291, 3.3513, 1.9300, 2.0316, 2.9931, 1.7481, 1.4241, 2.2682], + device='cuda:3'), covar=tensor([0.1328, 0.0626, 0.1146, 0.0896, 0.0633, 0.1258, 0.1035, 0.0636], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0317, 0.0340, 0.0268, 0.0248, 0.0342, 0.0293, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:39:15,663 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161693.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:39:16,432 INFO [train.py:903] (3/4) Epoch 24, batch 4650, loss[loss=0.1983, simple_loss=0.2827, pruned_loss=0.0569, over 19608.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2856, pruned_loss=0.06204, over 3823110.99 frames. ], batch size: 52, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:39:22,580 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=161698.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:39:23,628 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6307, 4.1233, 4.2919, 4.2971, 1.6068, 4.0714, 3.5314, 4.0520], + device='cuda:3'), covar=tensor([0.1645, 0.0926, 0.0606, 0.0711, 0.6052, 0.1005, 0.0690, 0.1054], + device='cuda:3'), in_proj_covar=tensor([0.0797, 0.0762, 0.0966, 0.0846, 0.0847, 0.0732, 0.0579, 0.0897], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 02:39:33,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 02:39:33,924 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161708.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:39:44,914 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 02:39:53,524 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=161723.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:40:19,111 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-04-03 02:40:19,299 INFO [train.py:903] (3/4) Epoch 24, batch 4700, loss[loss=0.1789, simple_loss=0.2543, pruned_loss=0.05174, over 19358.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2861, pruned_loss=0.06248, over 3820055.43 frames. ], batch size: 47, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:40:26,417 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.995e+02 5.147e+02 6.134e+02 7.606e+02 1.537e+03, threshold=1.227e+03, percent-clipped=3.0 +2023-04-03 02:40:39,910 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 02:40:43,392 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161763.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:41:20,593 INFO [train.py:903] (3/4) Epoch 24, batch 4750, loss[loss=0.2144, simple_loss=0.2933, pruned_loss=0.0677, over 19268.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2869, pruned_loss=0.06244, over 3826518.40 frames. ], batch size: 66, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:41:57,360 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=161822.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:42:09,943 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=161833.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:42:23,975 INFO [train.py:903] (3/4) Epoch 24, batch 4800, loss[loss=0.2002, simple_loss=0.2852, pruned_loss=0.05757, over 19771.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2863, pruned_loss=0.06207, over 3846288.14 frames. ], batch size: 56, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:42:31,545 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.081e+02 5.225e+02 6.101e+02 7.305e+02 1.695e+03, threshold=1.220e+03, percent-clipped=2.0 +2023-04-03 02:42:45,753 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6224, 1.7591, 1.9513, 2.0378, 1.5877, 1.9255, 1.9884, 1.8480], + device='cuda:3'), covar=tensor([0.3827, 0.3179, 0.1792, 0.2229, 0.3508, 0.2003, 0.4712, 0.3079], + device='cuda:3'), in_proj_covar=tensor([0.0914, 0.0986, 0.0727, 0.0940, 0.0892, 0.0829, 0.0849, 0.0791], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 02:43:04,826 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161878.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:43:25,945 INFO [train.py:903] (3/4) Epoch 24, batch 4850, loss[loss=0.1712, simple_loss=0.2591, pruned_loss=0.0416, over 19496.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2868, pruned_loss=0.06203, over 3851892.50 frames. ], batch size: 49, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:43:50,732 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 02:44:01,375 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.59 vs. limit=5.0 +2023-04-03 02:44:04,442 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3875, 2.4688, 2.6382, 3.1895, 2.4748, 3.0254, 2.7636, 2.5190], + device='cuda:3'), covar=tensor([0.4347, 0.4315, 0.1945, 0.2526, 0.4551, 0.2267, 0.4736, 0.3353], + device='cuda:3'), in_proj_covar=tensor([0.0916, 0.0988, 0.0729, 0.0943, 0.0895, 0.0832, 0.0852, 0.0794], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 02:44:11,974 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 02:44:16,555 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 02:44:17,713 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 02:44:19,192 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=161937.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:44:27,638 INFO [train.py:903] (3/4) Epoch 24, batch 4900, loss[loss=0.2382, simple_loss=0.3195, pruned_loss=0.07848, over 19773.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2868, pruned_loss=0.06227, over 3845378.25 frames. ], batch size: 56, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:44:27,698 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 02:44:33,871 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=161949.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:44:34,696 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.672e+02 5.435e+02 6.496e+02 8.047e+02 2.666e+03, threshold=1.299e+03, percent-clipped=6.0 +2023-04-03 02:44:47,038 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 02:44:53,284 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2156, 2.0742, 1.9627, 1.7991, 1.6111, 1.7791, 0.5739, 1.2769], + device='cuda:3'), covar=tensor([0.0635, 0.0674, 0.0548, 0.0965, 0.1219, 0.1033, 0.1451, 0.1128], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0356, 0.0359, 0.0384, 0.0460, 0.0390, 0.0338, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 02:44:53,293 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=161964.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:44:54,782 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.76 vs. limit=5.0 +2023-04-03 02:45:05,679 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=161974.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:45:23,800 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=161989.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:45:24,320 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-04-03 02:45:28,868 INFO [train.py:903] (3/4) Epoch 24, batch 4950, loss[loss=0.211, simple_loss=0.2953, pruned_loss=0.06334, over 19788.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2872, pruned_loss=0.06259, over 3839509.45 frames. ], batch size: 56, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:45:38,511 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-04-03 02:45:47,823 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 02:46:13,010 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 02:46:20,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0942, 1.1943, 1.5313, 0.7014, 1.9635, 2.4448, 2.1488, 2.6305], + device='cuda:3'), covar=tensor([0.1490, 0.3934, 0.3480, 0.2775, 0.0670, 0.0294, 0.0344, 0.0397], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0326, 0.0357, 0.0266, 0.0246, 0.0191, 0.0218, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 02:46:33,465 INFO [train.py:903] (3/4) Epoch 24, batch 5000, loss[loss=0.1841, simple_loss=0.2591, pruned_loss=0.05461, over 19753.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2875, pruned_loss=0.06274, over 3836829.99 frames. ], batch size: 46, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:46:41,315 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.890e+02 4.589e+02 5.769e+02 7.322e+02 1.477e+03, threshold=1.154e+03, percent-clipped=3.0 +2023-04-03 02:46:44,803 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 02:46:56,733 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 02:47:35,233 INFO [train.py:903] (3/4) Epoch 24, batch 5050, loss[loss=0.2314, simple_loss=0.308, pruned_loss=0.07743, over 19662.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2874, pruned_loss=0.06288, over 3835396.35 frames. ], batch size: 55, lr: 3.41e-03, grad_scale: 8.0 +2023-04-03 02:47:36,676 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=162095.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:47:49,441 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4223, 4.0337, 2.5401, 3.5583, 0.9754, 3.9866, 3.9097, 3.9421], + device='cuda:3'), covar=tensor([0.0644, 0.0938, 0.2017, 0.0862, 0.3646, 0.0695, 0.0917, 0.1036], + device='cuda:3'), in_proj_covar=tensor([0.0516, 0.0421, 0.0506, 0.0353, 0.0405, 0.0444, 0.0439, 0.0470], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:48:14,945 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 02:48:25,559 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=162134.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:48:37,733 INFO [train.py:903] (3/4) Epoch 24, batch 5100, loss[loss=0.1713, simple_loss=0.2541, pruned_loss=0.04427, over 19729.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2859, pruned_loss=0.06202, over 3841428.51 frames. ], batch size: 46, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:48:44,638 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.860e+02 4.540e+02 5.777e+02 7.463e+02 1.637e+03, threshold=1.155e+03, percent-clipped=6.0 +2023-04-03 02:48:51,574 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 02:48:55,175 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 02:48:55,632 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=162159.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:48:59,602 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 02:49:19,006 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=162177.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:49:39,595 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=162193.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:49:40,374 INFO [train.py:903] (3/4) Epoch 24, batch 5150, loss[loss=0.2294, simple_loss=0.3042, pruned_loss=0.07727, over 18751.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2855, pruned_loss=0.06178, over 3848572.82 frames. ], batch size: 74, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:49:57,186 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 02:49:57,528 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9996, 1.1829, 1.6162, 1.0882, 2.2395, 2.9610, 2.7006, 3.3292], + device='cuda:3'), covar=tensor([0.1844, 0.4998, 0.4462, 0.2727, 0.0713, 0.0293, 0.0329, 0.0364], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0324, 0.0354, 0.0265, 0.0244, 0.0190, 0.0216, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 02:50:03,593 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6676, 1.7418, 2.0015, 1.9803, 1.4475, 1.8859, 2.0024, 1.9077], + device='cuda:3'), covar=tensor([0.4054, 0.3741, 0.1951, 0.2497, 0.3990, 0.2318, 0.4918, 0.3319], + device='cuda:3'), in_proj_covar=tensor([0.0917, 0.0989, 0.0730, 0.0945, 0.0896, 0.0831, 0.0852, 0.0797], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 02:50:11,313 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=162218.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:50:31,571 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 02:50:43,108 INFO [train.py:903] (3/4) Epoch 24, batch 5200, loss[loss=0.2348, simple_loss=0.3063, pruned_loss=0.08167, over 13226.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2851, pruned_loss=0.06164, over 3852109.08 frames. ], batch size: 136, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:50:50,174 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=162249.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:50:51,037 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.192e+02 4.774e+02 5.644e+02 7.681e+02 1.514e+03, threshold=1.129e+03, percent-clipped=4.0 +2023-04-03 02:50:59,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 02:51:44,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 02:51:44,378 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=162292.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:51:44,830 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-03 02:51:46,274 INFO [train.py:903] (3/4) Epoch 24, batch 5250, loss[loss=0.2103, simple_loss=0.2942, pruned_loss=0.06318, over 19660.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2851, pruned_loss=0.06159, over 3855964.17 frames. ], batch size: 58, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:52:50,091 INFO [train.py:903] (3/4) Epoch 24, batch 5300, loss[loss=0.1851, simple_loss=0.2613, pruned_loss=0.05447, over 19389.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2845, pruned_loss=0.06103, over 3859366.40 frames. ], batch size: 48, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:52:57,096 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.926e+02 4.862e+02 5.825e+02 7.901e+02 2.284e+03, threshold=1.165e+03, percent-clipped=8.0 +2023-04-03 02:53:08,421 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 02:53:51,572 INFO [train.py:903] (3/4) Epoch 24, batch 5350, loss[loss=0.2521, simple_loss=0.3185, pruned_loss=0.09287, over 13030.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2852, pruned_loss=0.06164, over 3827261.34 frames. ], batch size: 136, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:54:28,294 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 02:54:48,009 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=162439.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:54:54,720 INFO [train.py:903] (3/4) Epoch 24, batch 5400, loss[loss=0.175, simple_loss=0.248, pruned_loss=0.05096, over 19779.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2855, pruned_loss=0.06202, over 3825829.13 frames. ], batch size: 47, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:55:02,636 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.752e+02 4.703e+02 6.237e+02 7.666e+02 1.372e+03, threshold=1.247e+03, percent-clipped=3.0 +2023-04-03 02:55:56,871 INFO [train.py:903] (3/4) Epoch 24, batch 5450, loss[loss=0.1902, simple_loss=0.2773, pruned_loss=0.05156, over 19778.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2858, pruned_loss=0.06152, over 3838282.18 frames. ], batch size: 54, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:56:41,299 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=162529.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:57:00,253 INFO [train.py:903] (3/4) Epoch 24, batch 5500, loss[loss=0.2219, simple_loss=0.3013, pruned_loss=0.07123, over 19668.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2846, pruned_loss=0.06113, over 3837091.25 frames. ], batch size: 60, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:57:02,962 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4782, 1.3508, 1.5740, 1.5134, 3.0322, 1.1017, 2.4930, 3.4156], + device='cuda:3'), covar=tensor([0.0499, 0.2832, 0.2813, 0.1872, 0.0707, 0.2501, 0.1133, 0.0280], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0368, 0.0389, 0.0350, 0.0372, 0.0352, 0.0385, 0.0405], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 02:57:05,486 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=162548.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:57:07,341 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.996e+02 5.077e+02 6.464e+02 7.861e+02 1.317e+03, threshold=1.293e+03, percent-clipped=1.0 +2023-04-03 02:57:12,105 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=162554.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:57:25,898 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 02:57:36,186 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=162573.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:57:44,060 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3353, 1.2887, 1.7515, 1.3423, 2.7052, 3.6356, 3.3235, 3.8917], + device='cuda:3'), covar=tensor([0.1618, 0.3905, 0.3468, 0.2504, 0.0629, 0.0207, 0.0226, 0.0254], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0325, 0.0356, 0.0265, 0.0244, 0.0190, 0.0216, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 02:58:01,107 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=162593.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 02:58:02,007 INFO [train.py:903] (3/4) Epoch 24, batch 5550, loss[loss=0.2423, simple_loss=0.3219, pruned_loss=0.08133, over 19722.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2841, pruned_loss=0.06106, over 3838049.88 frames. ], batch size: 63, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 02:58:08,797 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 02:58:59,595 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 02:59:02,953 INFO [train.py:903] (3/4) Epoch 24, batch 5600, loss[loss=0.1859, simple_loss=0.2549, pruned_loss=0.05843, over 19784.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2847, pruned_loss=0.06149, over 3836345.73 frames. ], batch size: 48, lr: 3.40e-03, grad_scale: 16.0 +2023-04-03 02:59:03,451 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8255, 1.9388, 2.2507, 2.2985, 1.7521, 2.2093, 2.2367, 2.0382], + device='cuda:3'), covar=tensor([0.4162, 0.3845, 0.1909, 0.2403, 0.4014, 0.2209, 0.4691, 0.3411], + device='cuda:3'), in_proj_covar=tensor([0.0915, 0.0988, 0.0728, 0.0939, 0.0893, 0.0829, 0.0850, 0.0793], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 02:59:12,266 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.260e+02 5.034e+02 5.949e+02 8.933e+02 2.100e+03, threshold=1.190e+03, percent-clipped=10.0 +2023-04-03 02:59:53,181 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7951, 1.4626, 1.6623, 1.4661, 4.3169, 1.0178, 2.6446, 4.7633], + device='cuda:3'), covar=tensor([0.0543, 0.2886, 0.3154, 0.2193, 0.0816, 0.2929, 0.1564, 0.0178], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0368, 0.0389, 0.0349, 0.0373, 0.0352, 0.0387, 0.0405], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:00:07,499 INFO [train.py:903] (3/4) Epoch 24, batch 5650, loss[loss=0.2256, simple_loss=0.3005, pruned_loss=0.0754, over 19670.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2845, pruned_loss=0.0612, over 3833678.34 frames. ], batch size: 60, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:00:25,014 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=162708.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:00:41,598 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-03 03:00:55,019 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 03:01:09,541 INFO [train.py:903] (3/4) Epoch 24, batch 5700, loss[loss=0.1997, simple_loss=0.281, pruned_loss=0.05915, over 19670.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2844, pruned_loss=0.06138, over 3830107.66 frames. ], batch size: 58, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:01:17,488 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.412e+02 4.704e+02 5.740e+02 7.100e+02 1.656e+03, threshold=1.148e+03, percent-clipped=4.0 +2023-04-03 03:02:11,505 INFO [train.py:903] (3/4) Epoch 24, batch 5750, loss[loss=0.2664, simple_loss=0.3287, pruned_loss=0.102, over 13542.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2853, pruned_loss=0.06176, over 3810446.42 frames. ], batch size: 136, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:02:13,886 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 03:02:22,230 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 03:02:28,834 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 03:02:31,428 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=162810.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:03:01,627 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-03 03:03:03,522 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=162835.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:03:04,583 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4296, 1.2943, 1.4114, 1.3620, 2.1802, 1.1707, 1.8854, 2.4046], + device='cuda:3'), covar=tensor([0.0547, 0.2112, 0.2161, 0.1425, 0.0588, 0.1821, 0.1843, 0.0446], + device='cuda:3'), in_proj_covar=tensor([0.0415, 0.0369, 0.0390, 0.0350, 0.0373, 0.0354, 0.0387, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:03:13,258 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3083, 1.3565, 1.5028, 1.4712, 1.8090, 1.7580, 1.8036, 0.6370], + device='cuda:3'), covar=tensor([0.2665, 0.4516, 0.2895, 0.2152, 0.1720, 0.2585, 0.1440, 0.5156], + device='cuda:3'), in_proj_covar=tensor([0.0545, 0.0658, 0.0734, 0.0498, 0.0630, 0.0545, 0.0670, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 03:03:13,920 INFO [train.py:903] (3/4) Epoch 24, batch 5800, loss[loss=0.2464, simple_loss=0.3195, pruned_loss=0.08665, over 19600.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2854, pruned_loss=0.0617, over 3803473.48 frames. ], batch size: 61, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:03:15,413 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=162845.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:03:22,908 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.123e+02 4.857e+02 6.549e+02 8.249e+02 1.553e+03, threshold=1.310e+03, percent-clipped=3.0 +2023-04-03 03:03:51,978 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=162873.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:04:16,920 INFO [train.py:903] (3/4) Epoch 24, batch 5850, loss[loss=0.1883, simple_loss=0.2638, pruned_loss=0.05641, over 17408.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2834, pruned_loss=0.06067, over 3818686.21 frames. ], batch size: 38, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:04:49,716 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9597, 4.5688, 2.6902, 3.9594, 1.0256, 4.4917, 4.3603, 4.4074], + device='cuda:3'), covar=tensor([0.0527, 0.0823, 0.1961, 0.0797, 0.3869, 0.0587, 0.0869, 0.1083], + device='cuda:3'), in_proj_covar=tensor([0.0510, 0.0417, 0.0501, 0.0351, 0.0401, 0.0440, 0.0434, 0.0466], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:05:20,443 INFO [train.py:903] (3/4) Epoch 24, batch 5900, loss[loss=0.1569, simple_loss=0.2385, pruned_loss=0.03768, over 19316.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2836, pruned_loss=0.06083, over 3819127.12 frames. ], batch size: 44, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:05:26,347 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 03:05:28,694 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.343e+02 4.550e+02 5.410e+02 6.827e+02 1.573e+03, threshold=1.082e+03, percent-clipped=1.0 +2023-04-03 03:05:43,787 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=162964.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:05:44,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 03:05:44,660 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0613, 3.7087, 2.5659, 3.2964, 0.9484, 3.6045, 3.5517, 3.5619], + device='cuda:3'), covar=tensor([0.0824, 0.1103, 0.2056, 0.0975, 0.3807, 0.0959, 0.1060, 0.1316], + device='cuda:3'), in_proj_covar=tensor([0.0512, 0.0419, 0.0503, 0.0352, 0.0402, 0.0442, 0.0436, 0.0468], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:06:15,266 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=162988.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:06:16,492 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=162989.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:06:18,783 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8807, 2.7374, 2.3006, 2.2399, 2.0206, 2.4109, 1.2740, 2.0549], + device='cuda:3'), covar=tensor([0.0858, 0.0634, 0.0701, 0.1153, 0.1169, 0.1161, 0.1390, 0.1146], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0360, 0.0363, 0.0389, 0.0467, 0.0396, 0.0342, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 03:06:21,815 INFO [train.py:903] (3/4) Epoch 24, batch 5950, loss[loss=0.1981, simple_loss=0.281, pruned_loss=0.05764, over 19666.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2831, pruned_loss=0.06068, over 3818984.08 frames. ], batch size: 53, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:07:22,797 INFO [train.py:903] (3/4) Epoch 24, batch 6000, loss[loss=0.1989, simple_loss=0.2733, pruned_loss=0.06219, over 19760.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2848, pruned_loss=0.06169, over 3823820.53 frames. ], batch size: 46, lr: 3.40e-03, grad_scale: 8.0 +2023-04-03 03:07:22,797 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 03:07:35,177 INFO [train.py:937] (3/4) Epoch 24, validation: loss=0.1683, simple_loss=0.2679, pruned_loss=0.03436, over 944034.00 frames. +2023-04-03 03:07:35,178 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 03:07:43,467 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.958e+02 5.276e+02 6.488e+02 8.005e+02 1.643e+03, threshold=1.298e+03, percent-clipped=7.0 +2023-04-03 03:08:35,917 INFO [train.py:903] (3/4) Epoch 24, batch 6050, loss[loss=0.1796, simple_loss=0.2547, pruned_loss=0.05229, over 19152.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2852, pruned_loss=0.06196, over 3817816.46 frames. ], batch size: 42, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:08:53,228 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=163108.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:09:37,877 INFO [train.py:903] (3/4) Epoch 24, batch 6100, loss[loss=0.2226, simple_loss=0.303, pruned_loss=0.0711, over 19475.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2858, pruned_loss=0.06226, over 3804463.32 frames. ], batch size: 64, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:09:42,783 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4423, 1.5447, 1.8795, 1.6435, 2.6781, 2.2765, 2.8644, 1.3199], + device='cuda:3'), covar=tensor([0.2535, 0.4272, 0.2696, 0.1978, 0.1627, 0.2176, 0.1463, 0.4445], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0656, 0.0731, 0.0495, 0.0626, 0.0541, 0.0666, 0.0559], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 03:09:45,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.855e+02 5.048e+02 5.892e+02 7.607e+02 1.565e+03, threshold=1.178e+03, percent-clipped=5.0 +2023-04-03 03:09:52,216 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-03 03:10:34,794 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=163189.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:10:40,171 INFO [train.py:903] (3/4) Epoch 24, batch 6150, loss[loss=0.2438, simple_loss=0.3135, pruned_loss=0.08707, over 19690.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2852, pruned_loss=0.06194, over 3790971.31 frames. ], batch size: 60, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:10:52,489 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5643, 2.5561, 2.2476, 2.5156, 2.2698, 2.1662, 2.1419, 2.6310], + device='cuda:3'), covar=tensor([0.0967, 0.1434, 0.1402, 0.1120, 0.1433, 0.0520, 0.1334, 0.0638], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0356, 0.0312, 0.0254, 0.0302, 0.0254, 0.0314, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:11:10,602 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 03:11:28,554 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.68 vs. limit=5.0 +2023-04-03 03:11:29,654 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.32 vs. limit=5.0 +2023-04-03 03:11:43,674 INFO [train.py:903] (3/4) Epoch 24, batch 6200, loss[loss=0.2374, simple_loss=0.3074, pruned_loss=0.08365, over 13490.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2839, pruned_loss=0.061, over 3794132.37 frames. ], batch size: 135, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:11:44,112 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=163244.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:11:51,418 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.768e+02 4.538e+02 5.825e+02 7.342e+02 1.276e+03, threshold=1.165e+03, percent-clipped=3.0 +2023-04-03 03:12:00,556 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=163258.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:12:03,975 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4811, 2.4476, 2.2470, 2.6035, 2.4897, 2.1719, 2.0590, 2.5481], + device='cuda:3'), covar=tensor([0.1001, 0.1583, 0.1394, 0.1043, 0.1312, 0.0534, 0.1430, 0.0692], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0357, 0.0312, 0.0254, 0.0303, 0.0254, 0.0315, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:12:14,197 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=163269.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:12:38,770 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1738, 2.2818, 2.5065, 2.8078, 2.2763, 2.7449, 2.5381, 2.3409], + device='cuda:3'), covar=tensor([0.4009, 0.3759, 0.1777, 0.2252, 0.3791, 0.2075, 0.4474, 0.3137], + device='cuda:3'), in_proj_covar=tensor([0.0912, 0.0985, 0.0725, 0.0937, 0.0893, 0.0826, 0.0849, 0.0791], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 03:12:43,993 INFO [train.py:903] (3/4) Epoch 24, batch 6250, loss[loss=0.2266, simple_loss=0.3053, pruned_loss=0.07392, over 19754.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06156, over 3780816.42 frames. ], batch size: 63, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:12:56,454 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=163304.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:13:11,284 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 03:13:45,521 INFO [train.py:903] (3/4) Epoch 24, batch 6300, loss[loss=0.2289, simple_loss=0.3005, pruned_loss=0.07868, over 19523.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2866, pruned_loss=0.0628, over 3782311.97 frames. ], batch size: 54, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:13:54,807 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.313e+02 5.008e+02 6.887e+02 8.761e+02 2.377e+03, threshold=1.377e+03, percent-clipped=3.0 +2023-04-03 03:14:40,283 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-03 03:14:48,471 INFO [train.py:903] (3/4) Epoch 24, batch 6350, loss[loss=0.2168, simple_loss=0.3059, pruned_loss=0.0638, over 19515.00 frames. ], tot_loss[loss=0.2071, simple_loss=0.2876, pruned_loss=0.06332, over 3791184.00 frames. ], batch size: 56, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:15:30,614 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4332, 1.4756, 1.4363, 1.7836, 1.4263, 1.6627, 1.6435, 1.5559], + device='cuda:3'), covar=tensor([0.0908, 0.0903, 0.1032, 0.0674, 0.0832, 0.0788, 0.0870, 0.0716], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0228, 0.0240, 0.0226, 0.0214, 0.0190, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 03:15:50,639 INFO [train.py:903] (3/4) Epoch 24, batch 6400, loss[loss=0.2141, simple_loss=0.2948, pruned_loss=0.06674, over 18130.00 frames. ], tot_loss[loss=0.2059, simple_loss=0.2865, pruned_loss=0.0627, over 3793152.58 frames. ], batch size: 83, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:15:59,002 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.312e+02 4.869e+02 5.854e+02 7.378e+02 1.563e+03, threshold=1.171e+03, percent-clipped=2.0 +2023-04-03 03:16:00,351 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=163452.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:16:01,985 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.72 vs. limit=2.0 +2023-04-03 03:16:52,285 INFO [train.py:903] (3/4) Epoch 24, batch 6450, loss[loss=0.2117, simple_loss=0.2986, pruned_loss=0.06239, over 18183.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2871, pruned_loss=0.06287, over 3805289.59 frames. ], batch size: 83, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:17:01,213 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-03 03:17:26,473 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3212, 3.9736, 2.7864, 3.5518, 1.1560, 3.9031, 3.8266, 3.8617], + device='cuda:3'), covar=tensor([0.0744, 0.0968, 0.1740, 0.0811, 0.3644, 0.0729, 0.0929, 0.1236], + device='cuda:3'), in_proj_covar=tensor([0.0512, 0.0418, 0.0502, 0.0352, 0.0402, 0.0443, 0.0436, 0.0467], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:17:34,415 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 03:17:53,151 INFO [train.py:903] (3/4) Epoch 24, batch 6500, loss[loss=0.1963, simple_loss=0.2793, pruned_loss=0.0566, over 19675.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2878, pruned_loss=0.06334, over 3816776.84 frames. ], batch size: 53, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:17:56,720 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 03:18:01,366 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.063e+02 4.704e+02 6.024e+02 8.376e+02 1.457e+03, threshold=1.205e+03, percent-clipped=5.0 +2023-04-03 03:18:14,509 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=163560.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:18:22,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=163567.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:18:25,185 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.79 vs. limit=5.0 +2023-04-03 03:18:44,339 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=163585.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:18:55,200 INFO [train.py:903] (3/4) Epoch 24, batch 6550, loss[loss=0.2144, simple_loss=0.3007, pruned_loss=0.06408, over 17300.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2869, pruned_loss=0.06261, over 3819647.55 frames. ], batch size: 101, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:19:05,581 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=163602.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:19:35,121 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.45 vs. limit=2.0 +2023-04-03 03:19:57,495 INFO [train.py:903] (3/4) Epoch 24, batch 6600, loss[loss=0.2175, simple_loss=0.2785, pruned_loss=0.07824, over 14805.00 frames. ], tot_loss[loss=0.206, simple_loss=0.287, pruned_loss=0.06244, over 3824479.92 frames. ], batch size: 32, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:20:05,512 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 5.011e+02 6.018e+02 7.633e+02 1.393e+03, threshold=1.204e+03, percent-clipped=8.0 +2023-04-03 03:20:26,961 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4452, 2.4456, 2.2648, 2.6219, 2.2548, 2.1739, 2.0687, 2.4625], + device='cuda:3'), covar=tensor([0.0972, 0.1528, 0.1292, 0.1035, 0.1404, 0.0536, 0.1385, 0.0694], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0360, 0.0315, 0.0256, 0.0306, 0.0256, 0.0316, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:20:57,666 INFO [train.py:903] (3/4) Epoch 24, batch 6650, loss[loss=0.1735, simple_loss=0.256, pruned_loss=0.04548, over 19790.00 frames. ], tot_loss[loss=0.2077, simple_loss=0.2886, pruned_loss=0.06341, over 3815767.54 frames. ], batch size: 48, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:21:24,756 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8682, 1.4339, 1.4750, 1.7151, 3.4126, 1.1537, 2.2498, 3.9509], + device='cuda:3'), covar=tensor([0.0500, 0.2704, 0.3008, 0.1765, 0.0705, 0.2577, 0.1512, 0.0211], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0369, 0.0391, 0.0348, 0.0374, 0.0352, 0.0386, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:21:25,791 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=163717.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:21:41,459 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4778, 2.5208, 2.3034, 2.6186, 2.3097, 2.2029, 2.2750, 2.6638], + device='cuda:3'), covar=tensor([0.1007, 0.1484, 0.1342, 0.1095, 0.1400, 0.0519, 0.1222, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0359, 0.0314, 0.0255, 0.0306, 0.0255, 0.0316, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:21:52,067 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.58 vs. limit=5.0 +2023-04-03 03:21:58,181 INFO [train.py:903] (3/4) Epoch 24, batch 6700, loss[loss=0.2836, simple_loss=0.3487, pruned_loss=0.1093, over 13113.00 frames. ], tot_loss[loss=0.2083, simple_loss=0.2893, pruned_loss=0.06367, over 3806607.74 frames. ], batch size: 136, lr: 3.39e-03, grad_scale: 4.0 +2023-04-03 03:22:07,915 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5182, 1.4049, 1.4790, 2.1472, 1.4953, 1.8053, 1.7568, 1.6066], + device='cuda:3'), covar=tensor([0.0958, 0.1124, 0.1116, 0.0736, 0.0951, 0.0908, 0.0974, 0.0835], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0222, 0.0227, 0.0239, 0.0226, 0.0213, 0.0189, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 03:22:08,770 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.807e+02 4.744e+02 6.010e+02 8.153e+02 1.593e+03, threshold=1.202e+03, percent-clipped=6.0 +2023-04-03 03:22:14,730 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2479, 3.7956, 3.8948, 3.9039, 1.5805, 3.7309, 3.2354, 3.6504], + device='cuda:3'), covar=tensor([0.1761, 0.1115, 0.0727, 0.0819, 0.6019, 0.1004, 0.0755, 0.1266], + device='cuda:3'), in_proj_covar=tensor([0.0794, 0.0756, 0.0963, 0.0841, 0.0841, 0.0731, 0.0574, 0.0892], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 03:22:32,671 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 03:22:57,590 INFO [train.py:903] (3/4) Epoch 24, batch 6750, loss[loss=0.1964, simple_loss=0.2788, pruned_loss=0.05698, over 19746.00 frames. ], tot_loss[loss=0.2072, simple_loss=0.2881, pruned_loss=0.06313, over 3821556.49 frames. ], batch size: 51, lr: 3.39e-03, grad_scale: 4.0 +2023-04-03 03:23:30,622 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=163823.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:23:53,913 INFO [train.py:903] (3/4) Epoch 24, batch 6800, loss[loss=0.2162, simple_loss=0.2871, pruned_loss=0.07264, over 19630.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2876, pruned_loss=0.06287, over 3828710.57 frames. ], batch size: 50, lr: 3.39e-03, grad_scale: 8.0 +2023-04-03 03:23:58,807 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=163848.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:24:03,015 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.305e+02 4.890e+02 5.869e+02 7.347e+02 2.478e+03, threshold=1.174e+03, percent-clipped=2.0 +2023-04-03 03:24:40,116 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 03:24:40,602 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 03:24:42,803 INFO [train.py:903] (3/4) Epoch 25, batch 0, loss[loss=0.1805, simple_loss=0.2645, pruned_loss=0.04819, over 19673.00 frames. ], tot_loss[loss=0.1805, simple_loss=0.2645, pruned_loss=0.04819, over 19673.00 frames. ], batch size: 53, lr: 3.32e-03, grad_scale: 8.0 +2023-04-03 03:24:42,803 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 03:24:54,385 INFO [train.py:937] (3/4) Epoch 25, validation: loss=0.1672, simple_loss=0.2675, pruned_loss=0.03346, over 944034.00 frames. +2023-04-03 03:24:54,386 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 03:25:06,943 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 03:25:12,156 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4804, 1.5638, 1.7511, 1.7630, 2.5583, 2.1832, 2.6234, 1.2095], + device='cuda:3'), covar=tensor([0.2742, 0.4596, 0.3041, 0.2168, 0.1705, 0.2548, 0.1768, 0.4924], + device='cuda:3'), in_proj_covar=tensor([0.0546, 0.0662, 0.0739, 0.0498, 0.0632, 0.0545, 0.0673, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 03:25:57,057 INFO [train.py:903] (3/4) Epoch 25, batch 50, loss[loss=0.1984, simple_loss=0.2653, pruned_loss=0.06577, over 19739.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2832, pruned_loss=0.05928, over 869777.23 frames. ], batch size: 46, lr: 3.32e-03, grad_scale: 8.0 +2023-04-03 03:26:35,505 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.099e+02 4.575e+02 5.589e+02 7.102e+02 2.434e+03, threshold=1.118e+03, percent-clipped=5.0 +2023-04-03 03:26:36,734 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 03:27:00,585 INFO [train.py:903] (3/4) Epoch 25, batch 100, loss[loss=0.2114, simple_loss=0.2968, pruned_loss=0.06301, over 18197.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2825, pruned_loss=0.05911, over 1528696.02 frames. ], batch size: 84, lr: 3.32e-03, grad_scale: 8.0 +2023-04-03 03:27:03,115 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=163973.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:27:15,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 03:27:34,447 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=163998.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:27:42,634 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.62 vs. limit=5.0 +2023-04-03 03:28:05,191 INFO [train.py:903] (3/4) Epoch 25, batch 150, loss[loss=0.2042, simple_loss=0.2904, pruned_loss=0.05898, over 19306.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2828, pruned_loss=0.06019, over 2028440.54 frames. ], batch size: 66, lr: 3.32e-03, grad_scale: 8.0 +2023-04-03 03:28:42,973 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.518e+02 5.520e+02 6.265e+02 7.455e+02 1.438e+03, threshold=1.253e+03, percent-clipped=2.0 +2023-04-03 03:29:06,832 INFO [train.py:903] (3/4) Epoch 25, batch 200, loss[loss=0.2469, simple_loss=0.3275, pruned_loss=0.08311, over 18162.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2866, pruned_loss=0.0624, over 2423083.00 frames. ], batch size: 83, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:29:09,372 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 03:29:53,469 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164108.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:30:10,582 INFO [train.py:903] (3/4) Epoch 25, batch 250, loss[loss=0.2063, simple_loss=0.2965, pruned_loss=0.05803, over 19758.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.287, pruned_loss=0.06259, over 2728045.83 frames. ], batch size: 56, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:30:10,841 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7792, 1.3299, 1.6299, 1.7190, 4.0505, 1.1399, 2.8026, 4.5228], + device='cuda:3'), covar=tensor([0.0570, 0.3432, 0.3431, 0.2158, 0.1024, 0.3088, 0.1495, 0.0273], + device='cuda:3'), in_proj_covar=tensor([0.0414, 0.0369, 0.0391, 0.0348, 0.0374, 0.0352, 0.0387, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:30:32,846 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3728, 2.2421, 2.1258, 1.9901, 1.8626, 1.9777, 0.6973, 1.3497], + device='cuda:3'), covar=tensor([0.0664, 0.0633, 0.0503, 0.0786, 0.1111, 0.0907, 0.1364, 0.1101], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0362, 0.0366, 0.0388, 0.0466, 0.0397, 0.0343, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 03:30:48,736 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.799e+02 4.875e+02 6.095e+02 7.386e+02 2.001e+03, threshold=1.219e+03, percent-clipped=3.0 +2023-04-03 03:31:13,868 INFO [train.py:903] (3/4) Epoch 25, batch 300, loss[loss=0.2134, simple_loss=0.2775, pruned_loss=0.07472, over 19486.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2852, pruned_loss=0.06218, over 2961407.85 frames. ], batch size: 49, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:31:16,578 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164173.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:32:18,090 INFO [train.py:903] (3/4) Epoch 25, batch 350, loss[loss=0.2081, simple_loss=0.283, pruned_loss=0.06663, over 19752.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2843, pruned_loss=0.06187, over 3144891.05 frames. ], batch size: 51, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:32:25,200 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 03:32:42,713 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1111, 1.3223, 1.8040, 0.9298, 2.2606, 3.0614, 2.7284, 3.2328], + device='cuda:3'), covar=tensor([0.1606, 0.3768, 0.3128, 0.2743, 0.0665, 0.0235, 0.0274, 0.0320], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0328, 0.0358, 0.0266, 0.0247, 0.0190, 0.0217, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 03:32:54,601 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-03 03:32:54,999 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.912e+02 5.440e+02 6.552e+02 9.332e+02 1.789e+03, threshold=1.310e+03, percent-clipped=12.0 +2023-04-03 03:33:20,730 INFO [train.py:903] (3/4) Epoch 25, batch 400, loss[loss=0.2542, simple_loss=0.317, pruned_loss=0.0957, over 13599.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2844, pruned_loss=0.06231, over 3289307.21 frames. ], batch size: 135, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:34:24,824 INFO [train.py:903] (3/4) Epoch 25, batch 450, loss[loss=0.1992, simple_loss=0.2922, pruned_loss=0.05311, over 19660.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2842, pruned_loss=0.06196, over 3415297.23 frames. ], batch size: 55, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:34:59,251 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 03:35:00,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 03:35:02,824 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.977e+02 4.778e+02 5.577e+02 7.081e+02 1.680e+03, threshold=1.115e+03, percent-clipped=3.0 +2023-04-03 03:35:27,605 INFO [train.py:903] (3/4) Epoch 25, batch 500, loss[loss=0.1939, simple_loss=0.2854, pruned_loss=0.0512, over 19752.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2849, pruned_loss=0.06172, over 3506646.03 frames. ], batch size: 54, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:35:35,599 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164378.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:36:31,240 INFO [train.py:903] (3/4) Epoch 25, batch 550, loss[loss=0.2219, simple_loss=0.3036, pruned_loss=0.07013, over 18177.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2851, pruned_loss=0.06197, over 3578455.41 frames. ], batch size: 83, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:37:09,260 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.449e+02 5.163e+02 6.218e+02 7.641e+02 1.675e+03, threshold=1.244e+03, percent-clipped=5.0 +2023-04-03 03:37:09,454 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=164452.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:37:14,144 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7934, 4.3977, 2.6265, 3.8433, 0.6955, 4.3568, 4.1502, 4.3056], + device='cuda:3'), covar=tensor([0.0572, 0.0887, 0.2008, 0.0819, 0.4338, 0.0617, 0.0901, 0.1008], + device='cuda:3'), in_proj_covar=tensor([0.0521, 0.0424, 0.0509, 0.0356, 0.0408, 0.0447, 0.0442, 0.0473], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:37:34,089 INFO [train.py:903] (3/4) Epoch 25, batch 600, loss[loss=0.175, simple_loss=0.249, pruned_loss=0.05049, over 19048.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2848, pruned_loss=0.06168, over 3641600.97 frames. ], batch size: 42, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:38:16,678 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 03:38:30,636 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=164517.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:38:36,199 INFO [train.py:903] (3/4) Epoch 25, batch 650, loss[loss=0.2377, simple_loss=0.307, pruned_loss=0.08424, over 19473.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2853, pruned_loss=0.06221, over 3691969.67 frames. ], batch size: 64, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:39:15,261 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.327e+02 4.538e+02 5.913e+02 7.820e+02 1.600e+03, threshold=1.183e+03, percent-clipped=2.0 +2023-04-03 03:39:33,656 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=164567.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:39:39,493 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5987, 1.5369, 1.5208, 1.8522, 1.3691, 1.7297, 1.7086, 1.6052], + device='cuda:3'), covar=tensor([0.0807, 0.0887, 0.0972, 0.0620, 0.0806, 0.0768, 0.0838, 0.0701], + device='cuda:3'), in_proj_covar=tensor([0.0210, 0.0221, 0.0226, 0.0237, 0.0225, 0.0211, 0.0188, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-03 03:39:40,281 INFO [train.py:903] (3/4) Epoch 25, batch 700, loss[loss=0.1908, simple_loss=0.2626, pruned_loss=0.05948, over 19296.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2857, pruned_loss=0.06269, over 3717383.75 frames. ], batch size: 44, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:39:58,633 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164586.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:40:11,614 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164596.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:40:44,572 INFO [train.py:903] (3/4) Epoch 25, batch 750, loss[loss=0.265, simple_loss=0.3315, pruned_loss=0.09927, over 12999.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2872, pruned_loss=0.06292, over 3741331.03 frames. ], batch size: 136, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:40:57,968 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=164632.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:41:21,490 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.304e+02 5.157e+02 6.753e+02 8.219e+02 1.587e+03, threshold=1.351e+03, percent-clipped=10.0 +2023-04-03 03:41:32,788 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-03 03:41:48,138 INFO [train.py:903] (3/4) Epoch 25, batch 800, loss[loss=0.1907, simple_loss=0.2715, pruned_loss=0.05498, over 19823.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2866, pruned_loss=0.06276, over 3747559.63 frames. ], batch size: 52, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:41:52,845 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164676.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:42:03,209 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 03:42:50,200 INFO [train.py:903] (3/4) Epoch 25, batch 850, loss[loss=0.2077, simple_loss=0.2759, pruned_loss=0.06971, over 19779.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2861, pruned_loss=0.06269, over 3748688.68 frames. ], batch size: 48, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:42:50,365 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=164722.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:43:29,203 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.532e+02 4.928e+02 6.003e+02 7.929e+02 1.446e+03, threshold=1.201e+03, percent-clipped=1.0 +2023-04-03 03:43:43,456 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=164764.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:43:44,364 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 03:43:52,636 INFO [train.py:903] (3/4) Epoch 25, batch 900, loss[loss=0.2086, simple_loss=0.2968, pruned_loss=0.06023, over 19775.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2857, pruned_loss=0.06222, over 3766548.71 frames. ], batch size: 56, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:44:29,969 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-04-03 03:44:38,689 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8994, 2.0076, 2.2439, 2.5309, 1.8543, 2.3921, 2.2733, 2.0666], + device='cuda:3'), covar=tensor([0.4719, 0.4316, 0.2116, 0.2487, 0.4482, 0.2391, 0.5372, 0.3733], + device='cuda:3'), in_proj_covar=tensor([0.0915, 0.0986, 0.0726, 0.0936, 0.0894, 0.0826, 0.0852, 0.0791], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 03:44:56,071 INFO [train.py:903] (3/4) Epoch 25, batch 950, loss[loss=0.2464, simple_loss=0.3162, pruned_loss=0.0883, over 19676.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06278, over 3767679.03 frames. ], batch size: 60, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:44:57,278 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 03:44:57,674 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=164823.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:45:15,711 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=164837.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:45:28,720 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=164848.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:45:32,821 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.080e+02 4.681e+02 5.489e+02 6.922e+02 1.500e+03, threshold=1.098e+03, percent-clipped=4.0 +2023-04-03 03:45:43,192 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 03:46:00,733 INFO [train.py:903] (3/4) Epoch 25, batch 1000, loss[loss=0.186, simple_loss=0.2783, pruned_loss=0.04688, over 19683.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2857, pruned_loss=0.06218, over 3783003.50 frames. ], batch size: 60, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:46:19,837 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=164888.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:46:53,199 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=164913.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:46:55,232 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 03:47:03,440 INFO [train.py:903] (3/4) Epoch 25, batch 1050, loss[loss=0.2001, simple_loss=0.2836, pruned_loss=0.05833, over 19589.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2864, pruned_loss=0.06241, over 3797759.30 frames. ], batch size: 61, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:47:13,161 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=164930.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:47:27,201 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=164940.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:47:36,149 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 03:47:41,880 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.266e+02 4.678e+02 5.688e+02 6.996e+02 1.189e+03, threshold=1.138e+03, percent-clipped=3.0 +2023-04-03 03:47:45,613 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6034, 1.2266, 1.4869, 1.5822, 3.1936, 1.2370, 2.3834, 3.6511], + device='cuda:3'), covar=tensor([0.0534, 0.2883, 0.2989, 0.1843, 0.0730, 0.2451, 0.1333, 0.0238], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0369, 0.0394, 0.0349, 0.0377, 0.0353, 0.0388, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:48:06,293 INFO [train.py:903] (3/4) Epoch 25, batch 1100, loss[loss=0.1915, simple_loss=0.2846, pruned_loss=0.04923, over 19684.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2853, pruned_loss=0.06171, over 3813644.77 frames. ], batch size: 58, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:49:07,178 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=165020.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:49:09,318 INFO [train.py:903] (3/4) Epoch 25, batch 1150, loss[loss=0.2435, simple_loss=0.3159, pruned_loss=0.08558, over 19514.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.285, pruned_loss=0.06201, over 3822998.60 frames. ], batch size: 56, lr: 3.31e-03, grad_scale: 8.0 +2023-04-03 03:49:40,417 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=165045.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:49:48,271 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.006e+02 5.101e+02 6.369e+02 8.453e+02 1.568e+03, threshold=1.274e+03, percent-clipped=10.0 +2023-04-03 03:49:52,205 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=165055.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:50:14,733 INFO [train.py:903] (3/4) Epoch 25, batch 1200, loss[loss=0.1757, simple_loss=0.2597, pruned_loss=0.04586, over 19748.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2842, pruned_loss=0.06154, over 3818063.97 frames. ], batch size: 47, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 03:50:41,038 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165093.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:50:46,493 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 03:51:01,325 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=165108.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:51:14,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165118.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:51:18,446 INFO [train.py:903] (3/4) Epoch 25, batch 1250, loss[loss=0.2339, simple_loss=0.3225, pruned_loss=0.07263, over 17593.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2844, pruned_loss=0.06149, over 3825949.05 frames. ], batch size: 101, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:51:34,119 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=165135.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:51:41,190 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 03:51:43,133 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3334, 3.8626, 3.9639, 3.9660, 1.5561, 3.7870, 3.3132, 3.7140], + device='cuda:3'), covar=tensor([0.1638, 0.0873, 0.0657, 0.0764, 0.5788, 0.0966, 0.0686, 0.1197], + device='cuda:3'), in_proj_covar=tensor([0.0802, 0.0766, 0.0974, 0.0850, 0.0856, 0.0736, 0.0579, 0.0903], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 03:51:43,152 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=165141.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:51:49,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3540, 1.3719, 1.8170, 1.6098, 3.0827, 4.7071, 4.5115, 5.1245], + device='cuda:3'), covar=tensor([0.1636, 0.3980, 0.3565, 0.2341, 0.0607, 0.0193, 0.0182, 0.0230], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0329, 0.0359, 0.0267, 0.0248, 0.0191, 0.0218, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 03:51:57,595 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.318e+02 4.844e+02 6.322e+02 8.270e+02 1.695e+03, threshold=1.264e+03, percent-clipped=2.0 +2023-04-03 03:52:21,099 INFO [train.py:903] (3/4) Epoch 25, batch 1300, loss[loss=0.1777, simple_loss=0.2644, pruned_loss=0.04552, over 19763.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2846, pruned_loss=0.06147, over 3835137.47 frames. ], batch size: 54, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:53:23,838 INFO [train.py:903] (3/4) Epoch 25, batch 1350, loss[loss=0.2121, simple_loss=0.2898, pruned_loss=0.06724, over 19391.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2848, pruned_loss=0.06153, over 3846102.35 frames. ], batch size: 48, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:53:27,597 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=165223.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:54:04,599 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.257e+02 4.919e+02 6.162e+02 7.502e+02 1.875e+03, threshold=1.232e+03, percent-clipped=3.0 +2023-04-03 03:54:29,444 INFO [train.py:903] (3/4) Epoch 25, batch 1400, loss[loss=0.223, simple_loss=0.3003, pruned_loss=0.07287, over 19755.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2842, pruned_loss=0.06133, over 3828453.63 frames. ], batch size: 54, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:54:51,016 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1338, 1.6841, 1.9654, 1.6521, 4.6551, 1.1630, 2.7446, 5.1386], + device='cuda:3'), covar=tensor([0.0422, 0.2707, 0.2712, 0.2071, 0.0700, 0.2752, 0.1352, 0.0147], + device='cuda:3'), in_proj_covar=tensor([0.0413, 0.0367, 0.0391, 0.0348, 0.0373, 0.0350, 0.0386, 0.0407], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:54:53,455 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4177, 2.4049, 2.1856, 2.0695, 1.9148, 2.2631, 1.5986, 1.8972], + device='cuda:3'), covar=tensor([0.0552, 0.0584, 0.0467, 0.0764, 0.0868, 0.0899, 0.1026, 0.0865], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0359, 0.0363, 0.0388, 0.0465, 0.0394, 0.0341, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 03:55:05,171 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165301.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:55:18,718 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165311.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:55:19,117 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 03:55:32,128 INFO [train.py:903] (3/4) Epoch 25, batch 1450, loss[loss=0.2544, simple_loss=0.3251, pruned_loss=0.09186, over 19606.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2837, pruned_loss=0.06136, over 3834302.72 frames. ], batch size: 57, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:55:33,175 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 03:55:37,017 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165326.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:55:48,893 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165336.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:56:10,662 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.320e+02 4.584e+02 6.100e+02 8.063e+02 1.924e+03, threshold=1.220e+03, percent-clipped=4.0 +2023-04-03 03:56:33,843 INFO [train.py:903] (3/4) Epoch 25, batch 1500, loss[loss=0.1807, simple_loss=0.2614, pruned_loss=0.05001, over 19791.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2851, pruned_loss=0.06182, over 3842125.08 frames. ], batch size: 47, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:56:59,436 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165391.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:57:30,062 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165416.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:57:36,929 INFO [train.py:903] (3/4) Epoch 25, batch 1550, loss[loss=0.2114, simple_loss=0.2917, pruned_loss=0.06558, over 18301.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2859, pruned_loss=0.06185, over 3838119.59 frames. ], batch size: 83, lr: 3.30e-03, grad_scale: 4.0 +2023-04-03 03:58:06,810 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5832, 2.3477, 1.8014, 1.5921, 2.1935, 1.4914, 1.3410, 1.9957], + device='cuda:3'), covar=tensor([0.1194, 0.0837, 0.1071, 0.0920, 0.0576, 0.1284, 0.0822, 0.0552], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0318, 0.0340, 0.0268, 0.0249, 0.0343, 0.0293, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 03:58:17,713 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.229e+02 4.957e+02 6.075e+02 6.924e+02 1.069e+03, threshold=1.215e+03, percent-clipped=0.0 +2023-04-03 03:58:37,746 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 03:58:42,876 INFO [train.py:903] (3/4) Epoch 25, batch 1600, loss[loss=0.1763, simple_loss=0.2595, pruned_loss=0.04653, over 19716.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2863, pruned_loss=0.06228, over 3837689.12 frames. ], batch size: 51, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 03:58:51,631 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165479.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:58:59,592 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=165485.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:59:06,509 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 03:59:21,464 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165504.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 03:59:44,266 INFO [train.py:903] (3/4) Epoch 25, batch 1650, loss[loss=0.2009, simple_loss=0.2846, pruned_loss=0.05862, over 18384.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2861, pruned_loss=0.06183, over 3841622.68 frames. ], batch size: 83, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:00:23,740 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.160e+02 4.584e+02 6.289e+02 7.621e+02 1.672e+03, threshold=1.258e+03, percent-clipped=5.0 +2023-04-03 04:00:44,111 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=165569.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:00:47,354 INFO [train.py:903] (3/4) Epoch 25, batch 1700, loss[loss=0.2414, simple_loss=0.3152, pruned_loss=0.08374, over 13111.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2856, pruned_loss=0.06176, over 3841783.26 frames. ], batch size: 135, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:01:22,694 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=165600.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:01:30,700 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 04:01:49,446 INFO [train.py:903] (3/4) Epoch 25, batch 1750, loss[loss=0.1946, simple_loss=0.273, pruned_loss=0.05807, over 19600.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2853, pruned_loss=0.06155, over 3833310.23 frames. ], batch size: 52, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:01:53,781 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.45 vs. limit=5.0 +2023-04-03 04:02:29,218 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 4.689e+02 5.610e+02 7.383e+02 2.270e+03, threshold=1.122e+03, percent-clipped=4.0 +2023-04-03 04:02:53,379 INFO [train.py:903] (3/4) Epoch 25, batch 1800, loss[loss=0.1966, simple_loss=0.2828, pruned_loss=0.05523, over 19399.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2855, pruned_loss=0.06229, over 3830920.68 frames. ], batch size: 48, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:03:51,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 04:03:56,452 INFO [train.py:903] (3/4) Epoch 25, batch 1850, loss[loss=0.2244, simple_loss=0.3015, pruned_loss=0.07363, over 13727.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2855, pruned_loss=0.06224, over 3825569.50 frames. ], batch size: 135, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:04:28,957 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 04:04:37,061 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.780e+02 4.826e+02 5.776e+02 6.973e+02 2.376e+03, threshold=1.155e+03, percent-clipped=4.0 +2023-04-03 04:04:56,259 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=165768.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:05:00,660 INFO [train.py:903] (3/4) Epoch 25, batch 1900, loss[loss=0.1876, simple_loss=0.2743, pruned_loss=0.05045, over 19596.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2854, pruned_loss=0.06216, over 3815495.22 frames. ], batch size: 52, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:05:16,869 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 04:05:21,710 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 04:05:24,267 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2762, 3.8213, 3.9462, 3.9550, 1.5923, 3.8002, 3.2420, 3.6858], + device='cuda:3'), covar=tensor([0.1764, 0.0914, 0.0683, 0.0777, 0.5829, 0.0861, 0.0724, 0.1241], + device='cuda:3'), in_proj_covar=tensor([0.0809, 0.0770, 0.0979, 0.0852, 0.0855, 0.0740, 0.0581, 0.0907], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 04:05:48,088 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 04:06:02,741 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=165821.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:06:03,531 INFO [train.py:903] (3/4) Epoch 25, batch 1950, loss[loss=0.2145, simple_loss=0.3047, pruned_loss=0.0622, over 19783.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2856, pruned_loss=0.06285, over 3810058.34 frames. ], batch size: 56, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:06:20,378 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4317, 1.5020, 1.7359, 1.6853, 2.5708, 2.2499, 2.7327, 1.2325], + device='cuda:3'), covar=tensor([0.2554, 0.4340, 0.2869, 0.1985, 0.1564, 0.2269, 0.1443, 0.4698], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0655, 0.0730, 0.0493, 0.0625, 0.0538, 0.0661, 0.0560], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:06:44,209 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.055e+02 4.671e+02 6.009e+02 7.545e+02 1.239e+03, threshold=1.202e+03, percent-clipped=2.0 +2023-04-03 04:06:46,991 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5601, 1.1909, 1.4088, 1.1745, 2.2027, 0.9464, 2.1660, 2.4875], + device='cuda:3'), covar=tensor([0.0758, 0.2941, 0.2885, 0.1844, 0.0884, 0.2399, 0.1060, 0.0470], + device='cuda:3'), in_proj_covar=tensor([0.0415, 0.0370, 0.0393, 0.0350, 0.0375, 0.0354, 0.0388, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:06:48,301 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=165856.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:07:08,057 INFO [train.py:903] (3/4) Epoch 25, batch 2000, loss[loss=0.2164, simple_loss=0.3012, pruned_loss=0.06576, over 19574.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2858, pruned_loss=0.06251, over 3821200.58 frames. ], batch size: 61, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:07:20,319 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=165881.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:08:00,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=165913.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:08:07,746 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 04:08:12,473 INFO [train.py:903] (3/4) Epoch 25, batch 2050, loss[loss=0.2176, simple_loss=0.3054, pruned_loss=0.06487, over 19502.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.286, pruned_loss=0.06235, over 3817641.50 frames. ], batch size: 64, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:08:27,907 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 04:08:27,941 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 04:08:38,772 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6193, 4.2035, 2.6713, 3.7081, 1.0850, 4.1924, 4.0166, 4.1166], + device='cuda:3'), covar=tensor([0.0635, 0.0959, 0.1987, 0.0886, 0.3807, 0.0640, 0.0882, 0.1147], + device='cuda:3'), in_proj_covar=tensor([0.0514, 0.0421, 0.0503, 0.0351, 0.0402, 0.0444, 0.0437, 0.0468], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:08:48,982 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 04:08:51,225 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.115e+02 5.221e+02 6.289e+02 7.653e+02 1.251e+03, threshold=1.258e+03, percent-clipped=2.0 +2023-04-03 04:09:15,824 INFO [train.py:903] (3/4) Epoch 25, batch 2100, loss[loss=0.1894, simple_loss=0.2679, pruned_loss=0.05552, over 19863.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2852, pruned_loss=0.06202, over 3830456.60 frames. ], batch size: 52, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:09:44,599 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 04:10:08,501 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 04:10:19,021 INFO [train.py:903] (3/4) Epoch 25, batch 2150, loss[loss=0.1834, simple_loss=0.2638, pruned_loss=0.05152, over 19858.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2848, pruned_loss=0.06198, over 3826215.28 frames. ], batch size: 52, lr: 3.30e-03, grad_scale: 8.0 +2023-04-03 04:10:26,455 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1139, 2.1988, 2.4618, 2.7709, 2.1814, 2.6755, 2.4983, 2.2458], + device='cuda:3'), covar=tensor([0.4284, 0.4116, 0.2018, 0.2449, 0.4337, 0.2244, 0.4740, 0.3414], + device='cuda:3'), in_proj_covar=tensor([0.0917, 0.0992, 0.0729, 0.0937, 0.0895, 0.0829, 0.0855, 0.0793], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 04:10:27,620 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166028.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:10:48,714 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2232, 2.1505, 1.8764, 1.7490, 1.4779, 1.7633, 0.5415, 1.3019], + device='cuda:3'), covar=tensor([0.0661, 0.0673, 0.0605, 0.1046, 0.1410, 0.1119, 0.1576, 0.1175], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0359, 0.0363, 0.0387, 0.0465, 0.0392, 0.0341, 0.0346], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:10:58,799 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.324e+02 5.055e+02 6.614e+02 9.413e+02 1.694e+03, threshold=1.323e+03, percent-clipped=9.0 +2023-04-03 04:11:21,716 INFO [train.py:903] (3/4) Epoch 25, batch 2200, loss[loss=0.1742, simple_loss=0.2516, pruned_loss=0.04838, over 19479.00 frames. ], tot_loss[loss=0.2058, simple_loss=0.2857, pruned_loss=0.06299, over 3809950.65 frames. ], batch size: 49, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:11:22,010 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166072.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 04:11:38,557 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0874, 2.0081, 1.7895, 2.1607, 1.9970, 1.7779, 1.7329, 2.0115], + device='cuda:3'), covar=tensor([0.1044, 0.1462, 0.1444, 0.1008, 0.1307, 0.0561, 0.1443, 0.0717], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0355, 0.0311, 0.0254, 0.0301, 0.0252, 0.0313, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:12:02,237 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4838, 2.2557, 1.6713, 1.5868, 2.1166, 1.4497, 1.3182, 1.9966], + device='cuda:3'), covar=tensor([0.1152, 0.0794, 0.1155, 0.0891, 0.0564, 0.1278, 0.0841, 0.0512], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0318, 0.0340, 0.0267, 0.0249, 0.0341, 0.0293, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:12:14,463 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166112.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:12:26,148 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4441, 3.7545, 2.4112, 2.1118, 3.6255, 2.0287, 1.8483, 2.9548], + device='cuda:3'), covar=tensor([0.1066, 0.0619, 0.0883, 0.0970, 0.0411, 0.1192, 0.0871, 0.0426], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0317, 0.0339, 0.0266, 0.0248, 0.0340, 0.0292, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:12:26,901 INFO [train.py:903] (3/4) Epoch 25, batch 2250, loss[loss=0.1798, simple_loss=0.2528, pruned_loss=0.05346, over 17011.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2856, pruned_loss=0.06252, over 3814662.85 frames. ], batch size: 37, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:12:34,230 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9595, 1.9188, 1.8120, 1.5971, 1.4746, 1.6320, 0.4047, 0.9127], + device='cuda:3'), covar=tensor([0.0662, 0.0666, 0.0415, 0.0719, 0.1188, 0.0797, 0.1391, 0.1120], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0360, 0.0364, 0.0389, 0.0466, 0.0393, 0.0343, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:13:04,902 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.096e+02 4.777e+02 5.716e+02 7.657e+02 1.924e+03, threshold=1.143e+03, percent-clipped=2.0 +2023-04-03 04:13:17,411 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0624, 3.0873, 1.8166, 2.0118, 2.8471, 1.6578, 1.4729, 2.3714], + device='cuda:3'), covar=tensor([0.1285, 0.0793, 0.1200, 0.0808, 0.0562, 0.1314, 0.1032, 0.0637], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0317, 0.0340, 0.0267, 0.0248, 0.0341, 0.0292, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:13:21,685 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166165.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:13:21,827 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166165.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:13:31,168 INFO [train.py:903] (3/4) Epoch 25, batch 2300, loss[loss=0.1889, simple_loss=0.2748, pruned_loss=0.05154, over 19754.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2866, pruned_loss=0.0628, over 3824907.77 frames. ], batch size: 54, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:13:42,665 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 04:14:00,647 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5717, 1.4198, 1.4600, 2.1173, 1.5995, 1.7649, 1.7104, 1.6000], + device='cuda:3'), covar=tensor([0.0926, 0.1098, 0.1107, 0.0751, 0.0965, 0.0937, 0.1041, 0.0842], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0222, 0.0226, 0.0239, 0.0226, 0.0213, 0.0189, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 04:14:34,779 INFO [train.py:903] (3/4) Epoch 25, batch 2350, loss[loss=0.1774, simple_loss=0.2721, pruned_loss=0.04139, over 19676.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2864, pruned_loss=0.06251, over 3830496.37 frames. ], batch size: 58, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:14:42,129 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166227.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:15:14,911 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.796e+02 5.104e+02 6.317e+02 8.219e+02 1.547e+03, threshold=1.263e+03, percent-clipped=3.0 +2023-04-03 04:15:17,158 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 04:15:24,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9736, 1.9258, 1.7412, 2.1055, 1.8994, 1.7978, 1.6747, 1.9713], + device='cuda:3'), covar=tensor([0.1113, 0.1524, 0.1464, 0.1067, 0.1340, 0.0583, 0.1476, 0.0747], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0354, 0.0311, 0.0253, 0.0301, 0.0252, 0.0312, 0.0258], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:15:34,627 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 04:15:38,210 INFO [train.py:903] (3/4) Epoch 25, batch 2400, loss[loss=0.1725, simple_loss=0.2583, pruned_loss=0.04335, over 19750.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2863, pruned_loss=0.06256, over 3840988.83 frames. ], batch size: 54, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:15:48,864 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166280.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:15:55,274 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=166284.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:16:25,818 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=166309.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:16:29,069 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.0396, 5.4444, 3.2746, 4.7104, 1.2619, 5.6416, 5.4610, 5.6725], + device='cuda:3'), covar=tensor([0.0391, 0.0815, 0.1691, 0.0786, 0.3978, 0.0548, 0.0784, 0.0788], + device='cuda:3'), in_proj_covar=tensor([0.0517, 0.0421, 0.0504, 0.0353, 0.0404, 0.0446, 0.0438, 0.0469], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:16:41,764 INFO [train.py:903] (3/4) Epoch 25, batch 2450, loss[loss=0.2198, simple_loss=0.2996, pruned_loss=0.06995, over 19533.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2857, pruned_loss=0.06195, over 3841799.73 frames. ], batch size: 54, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:16:46,748 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166325.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:17:20,232 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.023e+02 4.903e+02 5.916e+02 7.490e+02 1.353e+03, threshold=1.183e+03, percent-clipped=1.0 +2023-04-03 04:17:44,933 INFO [train.py:903] (3/4) Epoch 25, batch 2500, loss[loss=0.1851, simple_loss=0.2749, pruned_loss=0.0476, over 19651.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2851, pruned_loss=0.06184, over 3844955.45 frames. ], batch size: 55, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:17:52,048 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5242, 1.6100, 1.7500, 1.7078, 2.4501, 2.1727, 2.5885, 1.1946], + device='cuda:3'), covar=tensor([0.2480, 0.4208, 0.2667, 0.1939, 0.1547, 0.2266, 0.1407, 0.4499], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0656, 0.0731, 0.0494, 0.0625, 0.0539, 0.0664, 0.0561], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:18:41,295 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166416.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 04:18:48,043 INFO [train.py:903] (3/4) Epoch 25, batch 2550, loss[loss=0.2281, simple_loss=0.2993, pruned_loss=0.07851, over 19668.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2854, pruned_loss=0.06224, over 3829651.32 frames. ], batch size: 53, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:19:21,986 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.12 vs. limit=2.0 +2023-04-03 04:19:28,551 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.584e+02 5.112e+02 6.516e+02 8.109e+02 2.174e+03, threshold=1.303e+03, percent-clipped=8.0 +2023-04-03 04:19:40,847 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2296, 2.9237, 2.3203, 2.3245, 2.3107, 2.5513, 0.7661, 2.0809], + device='cuda:3'), covar=tensor([0.0694, 0.0598, 0.0703, 0.1198, 0.1035, 0.1125, 0.1618, 0.1094], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0360, 0.0363, 0.0387, 0.0467, 0.0393, 0.0340, 0.0345], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:19:46,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 04:19:53,287 INFO [train.py:903] (3/4) Epoch 25, batch 2600, loss[loss=0.1704, simple_loss=0.2481, pruned_loss=0.04637, over 19375.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2854, pruned_loss=0.06214, over 3829889.27 frames. ], batch size: 48, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:20:08,599 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=166483.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:20:29,546 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166499.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:20:34,775 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.42 vs. limit=2.0 +2023-04-03 04:20:40,256 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=166508.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:20:41,204 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166509.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:20:58,302 INFO [train.py:903] (3/4) Epoch 25, batch 2650, loss[loss=0.191, simple_loss=0.2745, pruned_loss=0.05373, over 17213.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2853, pruned_loss=0.06206, over 3831911.30 frames. ], batch size: 101, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:21:05,517 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166527.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:21:10,215 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166531.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 04:21:12,677 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5113, 1.4562, 1.7758, 1.4850, 2.4254, 3.3143, 3.0543, 3.5303], + device='cuda:3'), covar=tensor([0.1174, 0.3095, 0.2695, 0.2023, 0.0576, 0.0211, 0.0200, 0.0239], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0328, 0.0359, 0.0267, 0.0248, 0.0192, 0.0217, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 04:21:17,599 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=166536.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:21:19,774 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 04:21:32,642 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166549.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:21:38,171 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.868e+02 4.612e+02 5.770e+02 6.756e+02 1.610e+03, threshold=1.154e+03, percent-clipped=1.0 +2023-04-03 04:21:49,313 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=166561.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:22:03,059 INFO [train.py:903] (3/4) Epoch 25, batch 2700, loss[loss=0.1905, simple_loss=0.2639, pruned_loss=0.05859, over 19779.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.284, pruned_loss=0.06153, over 3800229.69 frames. ], batch size: 47, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:22:21,047 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-03 04:23:06,181 INFO [train.py:903] (3/4) Epoch 25, batch 2750, loss[loss=0.2375, simple_loss=0.3233, pruned_loss=0.07582, over 19596.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2831, pruned_loss=0.06107, over 3808676.45 frames. ], batch size: 61, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:23:08,979 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166624.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:23:45,197 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.014e+02 4.918e+02 6.109e+02 7.546e+02 1.552e+03, threshold=1.222e+03, percent-clipped=5.0 +2023-04-03 04:24:04,581 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166669.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:24:07,645 INFO [train.py:903] (3/4) Epoch 25, batch 2800, loss[loss=0.223, simple_loss=0.3067, pruned_loss=0.06959, over 19577.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2844, pruned_loss=0.06192, over 3817822.41 frames. ], batch size: 61, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:24:43,344 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3244, 3.8389, 3.9380, 3.9219, 1.5887, 3.7235, 3.2595, 3.6968], + device='cuda:3'), covar=tensor([0.1707, 0.0835, 0.0666, 0.0787, 0.5903, 0.1128, 0.0709, 0.1146], + device='cuda:3'), in_proj_covar=tensor([0.0803, 0.0766, 0.0975, 0.0853, 0.0851, 0.0740, 0.0580, 0.0905], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 04:25:11,679 INFO [train.py:903] (3/4) Epoch 25, batch 2850, loss[loss=0.2345, simple_loss=0.3075, pruned_loss=0.08078, over 13509.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2857, pruned_loss=0.06217, over 3816188.24 frames. ], batch size: 136, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:25:50,394 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.797e+02 4.563e+02 5.950e+02 8.060e+02 1.987e+03, threshold=1.190e+03, percent-clipped=10.0 +2023-04-03 04:25:50,711 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166753.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:26:11,850 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 04:26:14,215 INFO [train.py:903] (3/4) Epoch 25, batch 2900, loss[loss=0.1917, simple_loss=0.2627, pruned_loss=0.06033, over 19765.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2866, pruned_loss=0.06296, over 3798086.35 frames. ], batch size: 47, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:26:27,366 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166782.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:26:29,642 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166784.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:26:33,237 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=166787.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 04:27:05,025 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=166812.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 04:27:07,372 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3352, 2.3367, 2.6326, 3.1106, 2.3786, 2.9781, 2.6562, 2.4946], + device='cuda:3'), covar=tensor([0.4311, 0.4415, 0.2046, 0.2623, 0.4591, 0.2288, 0.4907, 0.3271], + device='cuda:3'), in_proj_covar=tensor([0.0918, 0.0990, 0.0729, 0.0938, 0.0894, 0.0829, 0.0853, 0.0793], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 04:27:17,225 INFO [train.py:903] (3/4) Epoch 25, batch 2950, loss[loss=0.2142, simple_loss=0.2992, pruned_loss=0.06463, over 19716.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2851, pruned_loss=0.0623, over 3794879.95 frames. ], batch size: 63, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:27:37,592 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8342, 1.6130, 2.1002, 2.0321, 3.3246, 4.8926, 4.7640, 5.3114], + device='cuda:3'), covar=tensor([0.1362, 0.3556, 0.3159, 0.1937, 0.0524, 0.0190, 0.0154, 0.0179], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0329, 0.0359, 0.0268, 0.0249, 0.0192, 0.0218, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 04:27:44,175 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166843.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:27:56,482 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.385e+02 4.674e+02 5.957e+02 7.713e+02 2.101e+03, threshold=1.191e+03, percent-clipped=6.0 +2023-04-03 04:28:20,024 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166871.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:28:21,107 INFO [train.py:903] (3/4) Epoch 25, batch 3000, loss[loss=0.1596, simple_loss=0.2478, pruned_loss=0.03573, over 19857.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2834, pruned_loss=0.06119, over 3812665.73 frames. ], batch size: 52, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:28:21,107 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 04:28:33,780 INFO [train.py:937] (3/4) Epoch 25, validation: loss=0.1677, simple_loss=0.2674, pruned_loss=0.034, over 944034.00 frames. +2023-04-03 04:28:33,781 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 04:28:35,111 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 04:28:44,963 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=166880.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:29:01,376 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=166893.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:29:17,226 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=166905.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:29:18,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9024, 1.9805, 2.2855, 2.5007, 1.8522, 2.4107, 2.2900, 2.0477], + device='cuda:3'), covar=tensor([0.4296, 0.4095, 0.1973, 0.2491, 0.4263, 0.2297, 0.4952, 0.3543], + device='cuda:3'), in_proj_covar=tensor([0.0919, 0.0992, 0.0730, 0.0939, 0.0894, 0.0829, 0.0853, 0.0795], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 04:29:33,259 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9015, 1.4877, 1.8229, 1.8114, 4.4120, 1.0677, 2.7560, 4.8643], + device='cuda:3'), covar=tensor([0.0433, 0.2953, 0.2880, 0.1920, 0.0735, 0.2781, 0.1345, 0.0150], + device='cuda:3'), in_proj_covar=tensor([0.0415, 0.0371, 0.0393, 0.0348, 0.0375, 0.0353, 0.0390, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:29:38,478 INFO [train.py:903] (3/4) Epoch 25, batch 3050, loss[loss=0.1922, simple_loss=0.2751, pruned_loss=0.05463, over 19660.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2833, pruned_loss=0.0608, over 3810481.60 frames. ], batch size: 55, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:29:53,819 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0251, 2.0675, 2.3441, 2.6684, 2.0433, 2.6136, 2.3904, 2.0948], + device='cuda:3'), covar=tensor([0.4279, 0.4075, 0.1995, 0.2493, 0.4263, 0.2127, 0.4912, 0.3452], + device='cuda:3'), in_proj_covar=tensor([0.0918, 0.0992, 0.0730, 0.0939, 0.0894, 0.0829, 0.0854, 0.0795], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 04:30:09,956 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=166947.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:30:17,431 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.386e+02 4.942e+02 6.233e+02 8.295e+02 1.859e+03, threshold=1.247e+03, percent-clipped=9.0 +2023-04-03 04:30:23,351 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166958.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:30:41,891 INFO [train.py:903] (3/4) Epoch 25, batch 3100, loss[loss=0.2142, simple_loss=0.2962, pruned_loss=0.06608, over 19649.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2842, pruned_loss=0.06133, over 3801839.56 frames. ], batch size: 55, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:30:58,955 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=166986.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:31:27,868 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167008.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:31:44,391 INFO [train.py:903] (3/4) Epoch 25, batch 3150, loss[loss=0.2525, simple_loss=0.3188, pruned_loss=0.09309, over 12833.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2847, pruned_loss=0.06189, over 3780337.82 frames. ], batch size: 136, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:32:08,044 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167040.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:32:08,790 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 04:32:24,685 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.516e+02 5.039e+02 6.011e+02 8.459e+02 2.094e+03, threshold=1.202e+03, percent-clipped=4.0 +2023-04-03 04:32:39,303 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167065.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:32:48,324 INFO [train.py:903] (3/4) Epoch 25, batch 3200, loss[loss=0.2037, simple_loss=0.2978, pruned_loss=0.05477, over 19601.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.284, pruned_loss=0.06132, over 3794345.73 frames. ], batch size: 57, lr: 3.29e-03, grad_scale: 8.0 +2023-04-03 04:33:20,651 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167097.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:33:47,236 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=167118.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:33:52,593 INFO [train.py:903] (3/4) Epoch 25, batch 3250, loss[loss=0.205, simple_loss=0.2857, pruned_loss=0.06222, over 17327.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2845, pruned_loss=0.06136, over 3799392.65 frames. ], batch size: 101, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:33:52,913 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=167122.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:33:57,478 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167126.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:34:33,315 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.514e+02 4.897e+02 6.248e+02 7.764e+02 1.427e+03, threshold=1.250e+03, percent-clipped=4.0 +2023-04-03 04:34:56,647 INFO [train.py:903] (3/4) Epoch 25, batch 3300, loss[loss=0.1683, simple_loss=0.253, pruned_loss=0.04185, over 19394.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2839, pruned_loss=0.06106, over 3799522.75 frames. ], batch size: 48, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:34:56,699 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 04:35:15,541 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8179, 1.5535, 1.8682, 2.0917, 4.3759, 1.2566, 2.6182, 4.8439], + device='cuda:3'), covar=tensor([0.0508, 0.2922, 0.2798, 0.1765, 0.0756, 0.2705, 0.1438, 0.0168], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0372, 0.0393, 0.0349, 0.0377, 0.0353, 0.0391, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:35:47,873 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167212.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:35:50,116 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167214.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:35:59,020 INFO [train.py:903] (3/4) Epoch 25, batch 3350, loss[loss=0.2698, simple_loss=0.3469, pruned_loss=0.09628, over 19781.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2853, pruned_loss=0.06217, over 3791609.17 frames. ], batch size: 56, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:36:21,826 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167239.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:36:24,302 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167241.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:36:25,490 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167242.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:36:41,239 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.112e+02 4.821e+02 5.770e+02 7.565e+02 1.496e+03, threshold=1.154e+03, percent-clipped=2.0 +2023-04-03 04:36:53,552 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167264.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:36:56,851 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167267.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:37:02,320 INFO [train.py:903] (3/4) Epoch 25, batch 3400, loss[loss=0.192, simple_loss=0.2655, pruned_loss=0.05926, over 19304.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2839, pruned_loss=0.06096, over 3809691.23 frames. ], batch size: 44, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:37:26,142 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167289.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:37:28,368 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167291.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:38:08,136 INFO [train.py:903] (3/4) Epoch 25, batch 3450, loss[loss=0.1967, simple_loss=0.2756, pruned_loss=0.05884, over 19479.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2836, pruned_loss=0.06054, over 3820654.35 frames. ], batch size: 64, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:38:10,566 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 04:38:36,023 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2370, 1.2958, 1.3083, 1.1087, 1.1253, 1.1929, 0.0695, 0.3995], + device='cuda:3'), covar=tensor([0.0784, 0.0714, 0.0487, 0.0639, 0.1398, 0.0718, 0.1422, 0.1208], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0360, 0.0362, 0.0386, 0.0465, 0.0393, 0.0341, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:38:49,778 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.112e+02 4.608e+02 5.631e+02 6.989e+02 1.333e+03, threshold=1.126e+03, percent-clipped=1.0 +2023-04-03 04:39:07,013 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2648, 2.9889, 2.4138, 2.3490, 2.1940, 2.6653, 1.0929, 2.1503], + device='cuda:3'), covar=tensor([0.0714, 0.0566, 0.0680, 0.1091, 0.1140, 0.0988, 0.1470, 0.1081], + device='cuda:3'), in_proj_covar=tensor([0.0361, 0.0358, 0.0361, 0.0385, 0.0464, 0.0392, 0.0340, 0.0343], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 04:39:12,381 INFO [train.py:903] (3/4) Epoch 25, batch 3500, loss[loss=0.1638, simple_loss=0.2445, pruned_loss=0.04151, over 19713.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2841, pruned_loss=0.06102, over 3830277.85 frames. ], batch size: 46, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:39:42,171 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3552, 2.0794, 1.6264, 1.4373, 1.9502, 1.2721, 1.3384, 1.7922], + device='cuda:3'), covar=tensor([0.1094, 0.0860, 0.1109, 0.0880, 0.0590, 0.1380, 0.0761, 0.0534], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0318, 0.0336, 0.0266, 0.0248, 0.0342, 0.0291, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:39:56,149 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167406.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:40:15,422 INFO [train.py:903] (3/4) Epoch 25, batch 3550, loss[loss=0.1649, simple_loss=0.2498, pruned_loss=0.04001, over 19471.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2833, pruned_loss=0.06077, over 3829108.67 frames. ], batch size: 49, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:40:51,453 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=167450.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:40:55,637 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.122e+02 5.129e+02 6.252e+02 7.981e+02 1.792e+03, threshold=1.250e+03, percent-clipped=6.0 +2023-04-03 04:41:06,612 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167462.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:41:11,119 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167466.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:41:13,868 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167468.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:41:17,992 INFO [train.py:903] (3/4) Epoch 25, batch 3600, loss[loss=0.2065, simple_loss=0.2877, pruned_loss=0.06259, over 19589.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2822, pruned_loss=0.06046, over 3819631.86 frames. ], batch size: 52, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:41:44,885 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167493.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:41:50,869 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167497.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:42:01,230 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4231, 1.2290, 1.5140, 1.6872, 2.9918, 1.3052, 2.2618, 3.4709], + device='cuda:3'), covar=tensor([0.0525, 0.3157, 0.3043, 0.1740, 0.0743, 0.2399, 0.1285, 0.0260], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0373, 0.0394, 0.0349, 0.0377, 0.0353, 0.0391, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:42:20,840 INFO [train.py:903] (3/4) Epoch 25, batch 3650, loss[loss=0.161, simple_loss=0.2363, pruned_loss=0.04291, over 19734.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2828, pruned_loss=0.06057, over 3831034.03 frames. ], batch size: 45, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:42:21,260 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167522.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:42:36,707 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 04:43:00,369 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.203e+02 5.264e+02 6.466e+02 8.043e+02 1.528e+03, threshold=1.293e+03, percent-clipped=3.0 +2023-04-03 04:43:24,179 INFO [train.py:903] (3/4) Epoch 25, batch 3700, loss[loss=0.2084, simple_loss=0.2701, pruned_loss=0.07334, over 19046.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2842, pruned_loss=0.06166, over 3819049.42 frames. ], batch size: 42, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:43:31,637 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167577.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:43:36,335 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167581.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:44:28,009 INFO [train.py:903] (3/4) Epoch 25, batch 3750, loss[loss=0.2509, simple_loss=0.3214, pruned_loss=0.09015, over 17201.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2847, pruned_loss=0.06188, over 3813522.99 frames. ], batch size: 101, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:44:57,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6161, 1.7330, 1.9525, 1.9992, 1.5624, 1.9205, 1.9402, 1.7886], + device='cuda:3'), covar=tensor([0.4400, 0.3663, 0.2029, 0.2417, 0.3836, 0.2215, 0.5441, 0.3568], + device='cuda:3'), in_proj_covar=tensor([0.0923, 0.0995, 0.0731, 0.0943, 0.0898, 0.0832, 0.0857, 0.0797], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 04:45:08,827 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.966e+02 5.285e+02 6.593e+02 8.150e+02 1.742e+03, threshold=1.319e+03, percent-clipped=4.0 +2023-04-03 04:45:19,734 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3829, 3.1366, 2.2217, 2.8481, 0.7676, 3.0584, 2.9610, 3.0083], + device='cuda:3'), covar=tensor([0.1131, 0.1385, 0.2160, 0.0995, 0.3869, 0.1045, 0.1159, 0.1477], + device='cuda:3'), in_proj_covar=tensor([0.0521, 0.0422, 0.0509, 0.0356, 0.0407, 0.0450, 0.0441, 0.0473], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:45:19,962 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167662.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:45:31,297 INFO [train.py:903] (3/4) Epoch 25, batch 3800, loss[loss=0.2032, simple_loss=0.2874, pruned_loss=0.05947, over 19662.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.284, pruned_loss=0.06157, over 3816245.27 frames. ], batch size: 53, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:45:50,870 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167687.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:46:01,963 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 04:46:32,955 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.70 vs. limit=2.0 +2023-04-03 04:46:35,334 INFO [train.py:903] (3/4) Epoch 25, batch 3850, loss[loss=0.2616, simple_loss=0.3391, pruned_loss=0.09205, over 19335.00 frames. ], tot_loss[loss=0.206, simple_loss=0.2864, pruned_loss=0.06277, over 3776450.86 frames. ], batch size: 66, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:46:54,897 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=167737.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:47:16,489 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.257e+02 4.991e+02 5.959e+02 7.597e+02 1.650e+03, threshold=1.192e+03, percent-clipped=2.0 +2023-04-03 04:47:39,474 INFO [train.py:903] (3/4) Epoch 25, batch 3900, loss[loss=0.2064, simple_loss=0.2935, pruned_loss=0.05962, over 17611.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2861, pruned_loss=0.06244, over 3780746.46 frames. ], batch size: 101, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:47:55,782 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=167785.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:48:06,238 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=167794.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:48:43,214 INFO [train.py:903] (3/4) Epoch 25, batch 3950, loss[loss=0.1644, simple_loss=0.2429, pruned_loss=0.04289, over 19392.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2856, pruned_loss=0.06217, over 3799399.64 frames. ], batch size: 47, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:48:45,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 04:48:56,960 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167833.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:49:02,584 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=167837.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:49:13,779 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-03 04:49:25,130 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.090e+02 4.848e+02 5.890e+02 7.638e+02 1.655e+03, threshold=1.178e+03, percent-clipped=7.0 +2023-04-03 04:49:29,085 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167858.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:49:34,616 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=167862.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:49:46,052 INFO [train.py:903] (3/4) Epoch 25, batch 4000, loss[loss=0.2213, simple_loss=0.3081, pruned_loss=0.06729, over 19111.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2861, pruned_loss=0.06235, over 3806909.50 frames. ], batch size: 69, lr: 3.28e-03, grad_scale: 8.0 +2023-04-03 04:50:32,137 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 04:50:33,687 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=167909.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:50:50,811 INFO [train.py:903] (3/4) Epoch 25, batch 4050, loss[loss=0.1683, simple_loss=0.2431, pruned_loss=0.04672, over 19750.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2864, pruned_loss=0.06243, over 3817419.68 frames. ], batch size: 48, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:51:32,273 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.281e+02 5.057e+02 6.529e+02 8.113e+02 1.821e+03, threshold=1.306e+03, percent-clipped=7.0 +2023-04-03 04:51:52,926 INFO [train.py:903] (3/4) Epoch 25, batch 4100, loss[loss=0.2079, simple_loss=0.2894, pruned_loss=0.06323, over 19840.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2861, pruned_loss=0.06253, over 3834776.59 frames. ], batch size: 52, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:52:24,786 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.43 vs. limit=2.0 +2023-04-03 04:52:28,533 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 04:52:56,836 INFO [train.py:903] (3/4) Epoch 25, batch 4150, loss[loss=0.1855, simple_loss=0.2674, pruned_loss=0.05177, over 19849.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2852, pruned_loss=0.06226, over 3822364.22 frames. ], batch size: 52, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:53:36,881 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=168053.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:53:39,909 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.124e+02 5.393e+02 6.582e+02 8.080e+02 1.683e+03, threshold=1.316e+03, percent-clipped=2.0 +2023-04-03 04:53:59,615 INFO [train.py:903] (3/4) Epoch 25, batch 4200, loss[loss=0.2303, simple_loss=0.3151, pruned_loss=0.07281, over 19666.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2845, pruned_loss=0.06183, over 3810048.86 frames. ], batch size: 58, lr: 3.28e-03, grad_scale: 4.0 +2023-04-03 04:54:01,978 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 04:54:10,933 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=168081.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:54:18,319 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-03 04:55:03,255 INFO [train.py:903] (3/4) Epoch 25, batch 4250, loss[loss=0.1816, simple_loss=0.2628, pruned_loss=0.05019, over 19729.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2852, pruned_loss=0.06225, over 3803554.33 frames. ], batch size: 51, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 04:55:13,289 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=168129.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:55:18,000 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 04:55:20,783 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.36 vs. limit=5.0 +2023-04-03 04:55:29,602 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 04:55:46,756 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.141e+02 4.961e+02 6.439e+02 7.740e+02 2.119e+03, threshold=1.288e+03, percent-clipped=3.0 +2023-04-03 04:55:59,670 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=168165.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:56:07,290 INFO [train.py:903] (3/4) Epoch 25, batch 4300, loss[loss=0.2167, simple_loss=0.2955, pruned_loss=0.06893, over 19546.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2845, pruned_loss=0.0616, over 3817799.38 frames. ], batch size: 56, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 04:56:15,589 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.33 vs. limit=5.0 +2023-04-03 04:56:29,213 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5060, 1.1010, 1.3735, 1.0262, 2.1516, 0.8834, 2.0621, 2.4105], + device='cuda:3'), covar=tensor([0.0940, 0.3260, 0.3003, 0.2042, 0.1142, 0.2410, 0.1194, 0.0550], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0375, 0.0395, 0.0351, 0.0379, 0.0354, 0.0393, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 04:56:30,458 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=168190.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:56:37,478 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=168196.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:56:42,775 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.97 vs. limit=5.0 +2023-04-03 04:57:00,322 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 04:57:10,760 INFO [train.py:903] (3/4) Epoch 25, batch 4350, loss[loss=0.1918, simple_loss=0.2649, pruned_loss=0.0593, over 19479.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.283, pruned_loss=0.06092, over 3831291.19 frames. ], batch size: 49, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 04:57:38,657 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=168244.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:57:47,840 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=168251.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 04:57:49,665 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-03 04:57:53,364 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.429e+02 4.774e+02 5.735e+02 7.211e+02 1.236e+03, threshold=1.147e+03, percent-clipped=0.0 +2023-04-03 04:58:13,426 INFO [train.py:903] (3/4) Epoch 25, batch 4400, loss[loss=0.1908, simple_loss=0.2827, pruned_loss=0.0495, over 19676.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2828, pruned_loss=0.06085, over 3837384.71 frames. ], batch size: 59, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 04:58:40,147 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 04:58:50,365 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 04:59:16,256 INFO [train.py:903] (3/4) Epoch 25, batch 4450, loss[loss=0.1843, simple_loss=0.2577, pruned_loss=0.05548, over 19759.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2819, pruned_loss=0.06035, over 3831631.30 frames. ], batch size: 46, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 04:59:59,625 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.939e+02 4.609e+02 5.804e+02 7.720e+02 1.927e+03, threshold=1.161e+03, percent-clipped=7.0 +2023-04-03 05:00:17,551 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-04-03 05:00:20,129 INFO [train.py:903] (3/4) Epoch 25, batch 4500, loss[loss=0.1758, simple_loss=0.2592, pruned_loss=0.0462, over 16118.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2829, pruned_loss=0.0608, over 3818312.33 frames. ], batch size: 35, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:00:44,698 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 05:00:52,068 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=168397.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:01:23,386 INFO [train.py:903] (3/4) Epoch 25, batch 4550, loss[loss=0.2012, simple_loss=0.2869, pruned_loss=0.0577, over 19747.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2833, pruned_loss=0.06083, over 3823557.33 frames. ], batch size: 63, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:01:34,659 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 05:01:59,985 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 05:02:02,785 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=168452.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:02:08,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.350e+02 4.803e+02 5.692e+02 6.651e+02 1.392e+03, threshold=1.138e+03, percent-clipped=3.0 +2023-04-03 05:02:15,309 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2846, 1.3570, 1.8243, 1.4283, 2.8283, 3.8893, 3.5826, 4.0656], + device='cuda:3'), covar=tensor([0.1532, 0.3781, 0.3175, 0.2372, 0.0595, 0.0181, 0.0200, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0327, 0.0357, 0.0267, 0.0248, 0.0191, 0.0217, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:02:27,736 INFO [train.py:903] (3/4) Epoch 25, batch 4600, loss[loss=0.1905, simple_loss=0.2846, pruned_loss=0.04817, over 19713.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2829, pruned_loss=0.06052, over 3821183.65 frames. ], batch size: 63, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 05:02:35,453 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=168477.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:02:59,875 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7121, 4.2897, 2.8111, 3.7357, 0.9563, 4.2654, 4.1533, 4.2558], + device='cuda:3'), covar=tensor([0.0639, 0.0959, 0.1994, 0.0925, 0.4194, 0.0658, 0.0918, 0.1142], + device='cuda:3'), in_proj_covar=tensor([0.0517, 0.0420, 0.0504, 0.0353, 0.0405, 0.0447, 0.0441, 0.0471], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:03:04,783 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=168500.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:03:20,628 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=168512.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:03:27,493 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 05:03:31,417 INFO [train.py:903] (3/4) Epoch 25, batch 4650, loss[loss=0.2083, simple_loss=0.2893, pruned_loss=0.06367, over 19592.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2831, pruned_loss=0.06037, over 3822034.19 frames. ], batch size: 52, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 05:03:36,201 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=168525.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:03:50,731 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 05:04:02,086 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 05:04:02,433 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2083, 1.1643, 1.5389, 1.3306, 2.3807, 3.4151, 3.1566, 3.7042], + device='cuda:3'), covar=tensor([0.1862, 0.5213, 0.4664, 0.2764, 0.0891, 0.0260, 0.0293, 0.0333], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0326, 0.0356, 0.0267, 0.0248, 0.0191, 0.0217, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:04:16,036 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.438e+02 5.060e+02 6.015e+02 7.632e+02 1.453e+03, threshold=1.203e+03, percent-clipped=3.0 +2023-04-03 05:04:34,758 INFO [train.py:903] (3/4) Epoch 25, batch 4700, loss[loss=0.2433, simple_loss=0.3197, pruned_loss=0.08348, over 19507.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2835, pruned_loss=0.06046, over 3835380.81 frames. ], batch size: 64, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 05:04:57,660 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 05:05:04,395 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=168595.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:05:38,975 INFO [train.py:903] (3/4) Epoch 25, batch 4750, loss[loss=0.1948, simple_loss=0.2839, pruned_loss=0.0528, over 19372.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2848, pruned_loss=0.06131, over 3824310.44 frames. ], batch size: 70, lr: 3.27e-03, grad_scale: 4.0 +2023-04-03 05:06:06,418 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.09 vs. limit=2.0 +2023-04-03 05:06:09,934 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 05:06:22,546 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.376e+02 5.029e+02 6.051e+02 7.124e+02 1.309e+03, threshold=1.210e+03, percent-clipped=1.0 +2023-04-03 05:06:41,004 INFO [train.py:903] (3/4) Epoch 25, batch 4800, loss[loss=0.2006, simple_loss=0.2701, pruned_loss=0.06552, over 19766.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2841, pruned_loss=0.06108, over 3835783.37 frames. ], batch size: 48, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:06:59,405 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8889, 2.6262, 2.4114, 2.7720, 2.5525, 2.3874, 2.1093, 2.6683], + device='cuda:3'), covar=tensor([0.0793, 0.1321, 0.1233, 0.1000, 0.1282, 0.0478, 0.1327, 0.0628], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0356, 0.0315, 0.0253, 0.0303, 0.0254, 0.0314, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:07:29,217 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=168710.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:07:43,821 INFO [train.py:903] (3/4) Epoch 25, batch 4850, loss[loss=0.2072, simple_loss=0.2854, pruned_loss=0.06446, over 19312.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2843, pruned_loss=0.06122, over 3822686.57 frames. ], batch size: 66, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:08:09,409 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 05:08:29,303 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.525e+02 5.536e+02 6.758e+02 9.275e+02 1.787e+03, threshold=1.352e+03, percent-clipped=12.0 +2023-04-03 05:08:30,530 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 05:08:36,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 05:08:36,451 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 05:08:43,837 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=168768.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:08:47,080 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 05:08:48,234 INFO [train.py:903] (3/4) Epoch 25, batch 4900, loss[loss=0.1895, simple_loss=0.276, pruned_loss=0.05146, over 19506.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2845, pruned_loss=0.06102, over 3805206.05 frames. ], batch size: 64, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:09:06,552 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 05:09:10,277 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3128, 0.9897, 1.1586, 1.9365, 1.4611, 1.2289, 1.4733, 1.2003], + device='cuda:3'), covar=tensor([0.1156, 0.1774, 0.1424, 0.0854, 0.1106, 0.1452, 0.1282, 0.1147], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0226, 0.0238, 0.0226, 0.0213, 0.0188, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 05:09:16,001 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=168793.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:09:52,880 INFO [train.py:903] (3/4) Epoch 25, batch 4950, loss[loss=0.1856, simple_loss=0.2708, pruned_loss=0.0502, over 19766.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2842, pruned_loss=0.06101, over 3815637.32 frames. ], batch size: 54, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:10:04,467 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 05:10:30,151 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 05:10:36,919 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.098e+02 4.693e+02 5.645e+02 7.417e+02 1.662e+03, threshold=1.129e+03, percent-clipped=1.0 +2023-04-03 05:10:48,988 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9779, 1.9737, 2.1839, 2.0705, 2.8380, 2.5709, 2.8467, 1.9861], + device='cuda:3'), covar=tensor([0.1936, 0.3450, 0.2262, 0.1720, 0.1315, 0.1809, 0.1288, 0.3812], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0666, 0.0739, 0.0501, 0.0630, 0.0544, 0.0670, 0.0567], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 05:10:50,514 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 05:10:55,827 INFO [train.py:903] (3/4) Epoch 25, batch 5000, loss[loss=0.2148, simple_loss=0.2998, pruned_loss=0.06486, over 19109.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2841, pruned_loss=0.06088, over 3795283.67 frames. ], batch size: 69, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:11:02,595 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 05:11:13,720 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 05:11:58,388 INFO [train.py:903] (3/4) Epoch 25, batch 5050, loss[loss=0.2033, simple_loss=0.2789, pruned_loss=0.06383, over 19464.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2864, pruned_loss=0.06206, over 3813215.79 frames. ], batch size: 49, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:12:06,816 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9622, 2.0564, 2.3378, 2.6329, 2.0165, 2.5322, 2.3307, 2.0534], + device='cuda:3'), covar=tensor([0.4399, 0.4216, 0.1933, 0.2605, 0.4305, 0.2313, 0.4911, 0.3518], + device='cuda:3'), in_proj_covar=tensor([0.0920, 0.0994, 0.0731, 0.0941, 0.0896, 0.0833, 0.0851, 0.0796], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 05:12:18,646 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5901, 1.6856, 1.9756, 1.9216, 1.4299, 1.8678, 1.9502, 1.7683], + device='cuda:3'), covar=tensor([0.4387, 0.4008, 0.2007, 0.2580, 0.4075, 0.2337, 0.5403, 0.3662], + device='cuda:3'), in_proj_covar=tensor([0.0920, 0.0994, 0.0731, 0.0942, 0.0897, 0.0833, 0.0851, 0.0796], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 05:12:33,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 05:12:41,880 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.044e+02 4.713e+02 5.534e+02 7.099e+02 1.364e+03, threshold=1.107e+03, percent-clipped=2.0 +2023-04-03 05:12:55,415 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=168966.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:13:02,137 INFO [train.py:903] (3/4) Epoch 25, batch 5100, loss[loss=0.1702, simple_loss=0.2465, pruned_loss=0.04695, over 19838.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2863, pruned_loss=0.06199, over 3809644.08 frames. ], batch size: 49, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:13:11,321 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 05:13:14,822 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 05:13:19,303 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=168985.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:13:20,126 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 05:13:21,522 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=168987.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 05:13:26,030 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=168991.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:13:32,755 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7660, 1.6110, 1.6168, 2.2117, 1.6786, 2.1335, 2.1092, 1.8364], + device='cuda:3'), covar=tensor([0.0839, 0.0946, 0.1006, 0.0770, 0.0868, 0.0736, 0.0873, 0.0685], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0224, 0.0226, 0.0239, 0.0226, 0.0214, 0.0189, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 05:14:05,065 INFO [train.py:903] (3/4) Epoch 25, batch 5150, loss[loss=0.2162, simple_loss=0.3019, pruned_loss=0.06532, over 19500.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2855, pruned_loss=0.06171, over 3810333.62 frames. ], batch size: 64, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:14:16,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 05:14:31,986 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3159, 2.0086, 1.5074, 1.3732, 1.8354, 1.2312, 1.2741, 1.7770], + device='cuda:3'), covar=tensor([0.0984, 0.0799, 0.1047, 0.0851, 0.0486, 0.1366, 0.0715, 0.0476], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0317, 0.0338, 0.0268, 0.0249, 0.0342, 0.0294, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:14:48,111 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.896e+02 5.295e+02 6.704e+02 8.101e+02 2.101e+03, threshold=1.341e+03, percent-clipped=6.0 +2023-04-03 05:14:52,537 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 05:15:08,241 INFO [train.py:903] (3/4) Epoch 25, batch 5200, loss[loss=0.1707, simple_loss=0.2552, pruned_loss=0.04304, over 19495.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2856, pruned_loss=0.06161, over 3807251.33 frames. ], batch size: 49, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:15:23,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 05:15:44,208 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169100.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:16:08,649 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 05:16:13,028 INFO [train.py:903] (3/4) Epoch 25, batch 5250, loss[loss=0.1891, simple_loss=0.2795, pruned_loss=0.04936, over 19540.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2848, pruned_loss=0.06105, over 3815494.87 frames. ], batch size: 56, lr: 3.27e-03, grad_scale: 8.0 +2023-04-03 05:16:27,871 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169134.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:16:56,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 5.211e+02 5.791e+02 7.204e+02 1.532e+03, threshold=1.158e+03, percent-clipped=2.0 +2023-04-03 05:17:16,195 INFO [train.py:903] (3/4) Epoch 25, batch 5300, loss[loss=0.1777, simple_loss=0.2581, pruned_loss=0.0487, over 19605.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2842, pruned_loss=0.06107, over 3820992.07 frames. ], batch size: 50, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:17:22,413 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169177.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:17:34,462 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 05:18:18,936 INFO [train.py:903] (3/4) Epoch 25, batch 5350, loss[loss=0.2541, simple_loss=0.3278, pruned_loss=0.09018, over 19791.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06149, over 3829336.99 frames. ], batch size: 56, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:18:57,326 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 05:19:04,351 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.964e+02 4.992e+02 6.506e+02 8.048e+02 1.510e+03, threshold=1.301e+03, percent-clipped=6.0 +2023-04-03 05:19:24,334 INFO [train.py:903] (3/4) Epoch 25, batch 5400, loss[loss=0.2194, simple_loss=0.3, pruned_loss=0.06944, over 19678.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2848, pruned_loss=0.06173, over 3833439.36 frames. ], batch size: 60, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:19:38,735 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1163, 2.1014, 1.7137, 2.1315, 2.0056, 1.8120, 1.7073, 1.9570], + device='cuda:3'), covar=tensor([0.1114, 0.1446, 0.1592, 0.1133, 0.1391, 0.0583, 0.1616, 0.0813], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0355, 0.0313, 0.0251, 0.0302, 0.0254, 0.0314, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:20:27,248 INFO [train.py:903] (3/4) Epoch 25, batch 5450, loss[loss=0.2186, simple_loss=0.2872, pruned_loss=0.07505, over 19579.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.285, pruned_loss=0.06201, over 3812255.38 frames. ], batch size: 52, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:20:33,628 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.34 vs. limit=5.0 +2023-04-03 05:20:36,090 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-04-03 05:20:36,598 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169329.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:20:39,942 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169331.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 05:20:46,814 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7384, 1.7297, 1.6123, 1.4102, 1.3396, 1.4733, 0.2545, 0.6729], + device='cuda:3'), covar=tensor([0.0638, 0.0599, 0.0395, 0.0632, 0.1146, 0.0724, 0.1332, 0.1061], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0363, 0.0364, 0.0391, 0.0467, 0.0399, 0.0344, 0.0346], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 05:20:50,641 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 05:21:11,465 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.767e+02 4.467e+02 5.289e+02 7.172e+02 1.661e+03, threshold=1.058e+03, percent-clipped=2.0 +2023-04-03 05:21:29,438 INFO [train.py:903] (3/4) Epoch 25, batch 5500, loss[loss=0.2189, simple_loss=0.2961, pruned_loss=0.07085, over 18850.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.285, pruned_loss=0.06162, over 3825831.54 frames. ], batch size: 74, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:21:57,710 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 05:22:24,188 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-03 05:22:33,004 INFO [train.py:903] (3/4) Epoch 25, batch 5550, loss[loss=0.2068, simple_loss=0.2932, pruned_loss=0.06018, over 19461.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2849, pruned_loss=0.06167, over 3837841.32 frames. ], batch size: 64, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:22:43,769 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 05:23:01,515 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169444.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:23:01,725 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=169444.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:23:04,029 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=169446.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 05:23:17,176 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.302e+02 5.192e+02 6.082e+02 7.576e+02 1.216e+03, threshold=1.216e+03, percent-clipped=3.0 +2023-04-03 05:23:33,281 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 05:23:36,949 INFO [train.py:903] (3/4) Epoch 25, batch 5600, loss[loss=0.1793, simple_loss=0.2604, pruned_loss=0.04911, over 19804.00 frames. ], tot_loss[loss=0.2056, simple_loss=0.2863, pruned_loss=0.06249, over 3832261.11 frames. ], batch size: 49, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:23:44,277 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169478.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:24:38,815 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169521.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:24:39,771 INFO [train.py:903] (3/4) Epoch 25, batch 5650, loss[loss=0.2206, simple_loss=0.3005, pruned_loss=0.07037, over 19481.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2861, pruned_loss=0.06235, over 3837468.40 frames. ], batch size: 64, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:25:24,921 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.766e+02 4.961e+02 6.158e+02 8.288e+02 1.627e+03, threshold=1.232e+03, percent-clipped=5.0 +2023-04-03 05:25:27,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=169559.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:25:30,419 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 05:25:37,670 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4877, 1.3660, 1.4169, 1.8688, 1.4035, 1.7245, 1.7524, 1.5171], + device='cuda:3'), covar=tensor([0.0896, 0.0973, 0.1043, 0.0650, 0.0814, 0.0754, 0.0783, 0.0737], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0222, 0.0225, 0.0238, 0.0225, 0.0212, 0.0188, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0004], + device='cuda:3') +2023-04-03 05:25:43,212 INFO [train.py:903] (3/4) Epoch 25, batch 5700, loss[loss=0.195, simple_loss=0.2634, pruned_loss=0.06329, over 19034.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2867, pruned_loss=0.06235, over 3832517.26 frames. ], batch size: 42, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:26:11,336 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=169593.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:26:11,692 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.73 vs. limit=2.0 +2023-04-03 05:26:12,795 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.00 vs. limit=5.0 +2023-04-03 05:26:22,651 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-04-03 05:26:29,603 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3820, 2.4276, 2.5833, 2.9213, 2.4652, 2.8278, 2.6137, 2.4747], + device='cuda:3'), covar=tensor([0.3323, 0.2883, 0.1447, 0.1915, 0.3128, 0.1631, 0.3407, 0.2446], + device='cuda:3'), in_proj_covar=tensor([0.0920, 0.0995, 0.0730, 0.0942, 0.0898, 0.0833, 0.0853, 0.0795], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 05:26:47,655 INFO [train.py:903] (3/4) Epoch 25, batch 5750, loss[loss=0.1763, simple_loss=0.2544, pruned_loss=0.04908, over 19735.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2873, pruned_loss=0.06281, over 3816687.43 frames. ], batch size: 46, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:26:48,849 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 05:26:59,218 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 05:27:04,021 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 05:27:06,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=169636.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:27:09,111 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2884, 1.1954, 1.7297, 1.2863, 2.6237, 3.5133, 3.1463, 3.6325], + device='cuda:3'), covar=tensor([0.1667, 0.4162, 0.3582, 0.2711, 0.0648, 0.0213, 0.0248, 0.0304], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0326, 0.0356, 0.0267, 0.0248, 0.0192, 0.0217, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:27:10,238 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169639.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:27:32,987 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.159e+02 4.765e+02 5.968e+02 8.362e+02 1.575e+03, threshold=1.194e+03, percent-clipped=5.0 +2023-04-03 05:27:52,468 INFO [train.py:903] (3/4) Epoch 25, batch 5800, loss[loss=0.2027, simple_loss=0.2987, pruned_loss=0.05334, over 19587.00 frames. ], tot_loss[loss=0.2055, simple_loss=0.2867, pruned_loss=0.06214, over 3824608.80 frames. ], batch size: 57, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:28:20,543 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2643, 1.3492, 1.8628, 1.1876, 2.5749, 3.4059, 3.0492, 3.5823], + device='cuda:3'), covar=tensor([0.1570, 0.3748, 0.3167, 0.2484, 0.0575, 0.0219, 0.0251, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0326, 0.0357, 0.0267, 0.0248, 0.0192, 0.0217, 0.0267], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:28:27,916 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=169700.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:28:30,491 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=169702.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 05:28:55,622 INFO [train.py:903] (3/4) Epoch 25, batch 5850, loss[loss=0.2248, simple_loss=0.3136, pruned_loss=0.06803, over 19668.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2858, pruned_loss=0.06179, over 3829542.61 frames. ], batch size: 58, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:29:00,330 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=169725.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:29:02,806 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=169727.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 05:29:09,611 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3900, 1.4280, 1.6512, 1.6212, 2.3082, 2.1159, 2.2902, 0.9918], + device='cuda:3'), covar=tensor([0.2615, 0.4527, 0.2851, 0.2026, 0.1509, 0.2192, 0.1467, 0.4822], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0663, 0.0736, 0.0501, 0.0630, 0.0542, 0.0668, 0.0565], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 05:29:28,448 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169747.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:29:41,207 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.382e+02 5.003e+02 6.037e+02 8.413e+02 1.989e+03, threshold=1.207e+03, percent-clipped=6.0 +2023-04-03 05:29:51,800 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169765.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:29:59,817 INFO [train.py:903] (3/4) Epoch 25, batch 5900, loss[loss=0.1791, simple_loss=0.256, pruned_loss=0.0511, over 19394.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2848, pruned_loss=0.06123, over 3841940.32 frames. ], batch size: 48, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:30:04,507 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 05:30:27,861 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 05:30:42,243 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=169804.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:30:56,654 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=169815.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:31:04,213 INFO [train.py:903] (3/4) Epoch 25, batch 5950, loss[loss=0.2184, simple_loss=0.3011, pruned_loss=0.06784, over 18150.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2851, pruned_loss=0.06143, over 3844933.00 frames. ], batch size: 83, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:31:28,362 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=169840.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:31:39,968 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=169849.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:31:49,601 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.314e+02 4.837e+02 6.128e+02 7.395e+02 1.765e+03, threshold=1.226e+03, percent-clipped=3.0 +2023-04-03 05:32:09,286 INFO [train.py:903] (3/4) Epoch 25, batch 6000, loss[loss=0.2005, simple_loss=0.2715, pruned_loss=0.0647, over 19377.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2846, pruned_loss=0.06137, over 3839345.84 frames. ], batch size: 47, lr: 3.26e-03, grad_scale: 8.0 +2023-04-03 05:32:09,286 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 05:32:21,932 INFO [train.py:937] (3/4) Epoch 25, validation: loss=0.1675, simple_loss=0.2674, pruned_loss=0.03383, over 944034.00 frames. +2023-04-03 05:32:21,933 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 05:32:25,726 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=169874.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:32:27,349 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.69 vs. limit=5.0 +2023-04-03 05:32:48,264 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=169892.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:32:57,515 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0559, 1.8579, 1.6843, 2.0451, 1.8761, 1.7422, 1.5903, 1.9416], + device='cuda:3'), covar=tensor([0.1124, 0.1646, 0.1644, 0.1120, 0.1492, 0.0643, 0.1683, 0.0870], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0354, 0.0312, 0.0252, 0.0302, 0.0254, 0.0314, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:33:03,279 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9930, 1.6571, 1.5692, 1.8508, 1.6148, 1.6496, 1.4546, 1.8472], + device='cuda:3'), covar=tensor([0.0994, 0.1238, 0.1483, 0.0978, 0.1287, 0.0575, 0.1594, 0.0749], + device='cuda:3'), in_proj_covar=tensor([0.0274, 0.0355, 0.0313, 0.0252, 0.0302, 0.0254, 0.0314, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:33:20,573 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=169917.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:33:26,959 INFO [train.py:903] (3/4) Epoch 25, batch 6050, loss[loss=0.2015, simple_loss=0.2928, pruned_loss=0.05509, over 19495.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2842, pruned_loss=0.06125, over 3829383.52 frames. ], batch size: 64, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:34:12,787 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.356e+02 4.971e+02 6.200e+02 7.955e+02 1.563e+03, threshold=1.240e+03, percent-clipped=4.0 +2023-04-03 05:34:30,170 INFO [train.py:903] (3/4) Epoch 25, batch 6100, loss[loss=0.1632, simple_loss=0.2503, pruned_loss=0.03807, over 19498.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2835, pruned_loss=0.06084, over 3830950.26 frames. ], batch size: 49, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:34:44,098 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=169983.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:35:19,629 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-03 05:35:35,274 INFO [train.py:903] (3/4) Epoch 25, batch 6150, loss[loss=0.1777, simple_loss=0.2522, pruned_loss=0.0516, over 18625.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2832, pruned_loss=0.06075, over 3818179.92 frames. ], batch size: 41, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:35:40,504 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7576, 1.8602, 2.1064, 1.9563, 2.9822, 2.5694, 3.2409, 1.7526], + device='cuda:3'), covar=tensor([0.2366, 0.4053, 0.2649, 0.1883, 0.1678, 0.2167, 0.1527, 0.4275], + device='cuda:3'), in_proj_covar=tensor([0.0546, 0.0663, 0.0737, 0.0500, 0.0630, 0.0541, 0.0666, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 05:36:07,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 05:36:22,372 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.332e+02 4.967e+02 5.743e+02 7.363e+02 2.013e+03, threshold=1.149e+03, percent-clipped=2.0 +2023-04-03 05:36:28,775 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=170063.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:36:40,353 INFO [train.py:903] (3/4) Epoch 25, batch 6200, loss[loss=0.202, simple_loss=0.2689, pruned_loss=0.06753, over 16821.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2834, pruned_loss=0.0608, over 3812101.68 frames. ], batch size: 37, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:37:04,433 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170091.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:37:13,850 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170098.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:37:26,517 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170109.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:37:39,279 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8522, 1.2826, 1.6365, 2.8417, 1.7549, 1.6035, 2.0411, 1.5132], + device='cuda:3'), covar=tensor([0.1094, 0.1820, 0.1382, 0.0944, 0.1251, 0.1520, 0.1435, 0.1173], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0225, 0.0238, 0.0225, 0.0214, 0.0189, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 05:37:43,648 INFO [train.py:903] (3/4) Epoch 25, batch 6250, loss[loss=0.1842, simple_loss=0.2598, pruned_loss=0.05427, over 19383.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2829, pruned_loss=0.06062, over 3821124.90 frames. ], batch size: 47, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:38:16,018 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 05:38:16,221 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170148.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:38:29,525 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.163e+02 5.199e+02 6.170e+02 7.836e+02 1.706e+03, threshold=1.234e+03, percent-clipped=7.0 +2023-04-03 05:38:47,598 INFO [train.py:903] (3/4) Epoch 25, batch 6300, loss[loss=0.1807, simple_loss=0.2637, pruned_loss=0.04884, over 19623.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2837, pruned_loss=0.06073, over 3826290.73 frames. ], batch size: 50, lr: 3.26e-03, grad_scale: 4.0 +2023-04-03 05:39:08,055 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4502, 2.4250, 2.1685, 2.0662, 1.9637, 2.2220, 1.4455, 1.9206], + device='cuda:3'), covar=tensor([0.0628, 0.0601, 0.0469, 0.0801, 0.0796, 0.0998, 0.1145, 0.0806], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0359, 0.0362, 0.0387, 0.0463, 0.0396, 0.0341, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 05:39:32,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170206.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:39:51,012 INFO [train.py:903] (3/4) Epoch 25, batch 6350, loss[loss=0.2433, simple_loss=0.3157, pruned_loss=0.08551, over 19611.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2846, pruned_loss=0.06122, over 3819943.15 frames. ], batch size: 57, lr: 3.25e-03, grad_scale: 4.0 +2023-04-03 05:39:53,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170224.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:40:36,695 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.666e+02 4.417e+02 5.475e+02 7.140e+02 1.571e+03, threshold=1.095e+03, percent-clipped=3.0 +2023-04-03 05:40:42,944 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170263.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:40:54,617 INFO [train.py:903] (3/4) Epoch 25, batch 6400, loss[loss=0.1964, simple_loss=0.2765, pruned_loss=0.05815, over 19739.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2845, pruned_loss=0.06092, over 3819486.58 frames. ], batch size: 51, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:41:04,248 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8473, 1.6500, 1.4562, 1.7924, 1.5333, 1.6152, 1.4616, 1.6929], + device='cuda:3'), covar=tensor([0.1141, 0.1356, 0.1617, 0.1024, 0.1338, 0.0600, 0.1607, 0.0841], + device='cuda:3'), in_proj_covar=tensor([0.0273, 0.0354, 0.0312, 0.0252, 0.0301, 0.0253, 0.0313, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:41:59,011 INFO [train.py:903] (3/4) Epoch 25, batch 6450, loss[loss=0.1842, simple_loss=0.2744, pruned_loss=0.047, over 19761.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2853, pruned_loss=0.06139, over 3808890.97 frames. ], batch size: 54, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:42:32,625 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2438, 1.2868, 1.6231, 1.0502, 2.4415, 3.3515, 3.0516, 3.5896], + device='cuda:3'), covar=tensor([0.1549, 0.4138, 0.3785, 0.2801, 0.0657, 0.0210, 0.0239, 0.0296], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0330, 0.0361, 0.0269, 0.0251, 0.0193, 0.0219, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:42:40,350 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=170354.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:42:44,566 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.061e+02 4.688e+02 6.468e+02 8.318e+02 2.178e+03, threshold=1.294e+03, percent-clipped=13.0 +2023-04-03 05:42:45,706 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 05:43:03,216 INFO [train.py:903] (3/4) Epoch 25, batch 6500, loss[loss=0.1635, simple_loss=0.2432, pruned_loss=0.04187, over 19738.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2864, pruned_loss=0.06194, over 3789877.22 frames. ], batch size: 45, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:43:07,921 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 05:43:12,752 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=170379.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:43:48,145 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170407.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:44:06,260 INFO [train.py:903] (3/4) Epoch 25, batch 6550, loss[loss=0.1712, simple_loss=0.2527, pruned_loss=0.04488, over 19809.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2874, pruned_loss=0.0625, over 3799448.53 frames. ], batch size: 49, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:44:52,748 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.983e+02 4.881e+02 6.158e+02 7.799e+02 1.457e+03, threshold=1.232e+03, percent-clipped=1.0 +2023-04-03 05:44:57,811 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=170462.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:45:09,731 INFO [train.py:903] (3/4) Epoch 25, batch 6600, loss[loss=0.2023, simple_loss=0.2941, pruned_loss=0.05532, over 19579.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2867, pruned_loss=0.06209, over 3806591.72 frames. ], batch size: 61, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:45:19,115 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=170479.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:45:20,455 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=170480.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:45:27,515 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4786, 1.4893, 1.8066, 1.7228, 2.6072, 2.2856, 2.8083, 1.1988], + device='cuda:3'), covar=tensor([0.2464, 0.4488, 0.2800, 0.1917, 0.1617, 0.2149, 0.1497, 0.4618], + device='cuda:3'), in_proj_covar=tensor([0.0547, 0.0664, 0.0738, 0.0498, 0.0628, 0.0542, 0.0667, 0.0564], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 05:45:28,736 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=170487.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:45:51,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=170505.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:46:09,043 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=170519.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:46:13,017 INFO [train.py:903] (3/4) Epoch 25, batch 6650, loss[loss=0.191, simple_loss=0.2733, pruned_loss=0.05436, over 19692.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2865, pruned_loss=0.06196, over 3822248.27 frames. ], batch size: 53, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:46:13,409 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170522.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:46:41,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=170544.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:46:58,412 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.718e+02 4.827e+02 5.978e+02 7.974e+02 2.215e+03, threshold=1.196e+03, percent-clipped=6.0 +2023-04-03 05:47:16,979 INFO [train.py:903] (3/4) Epoch 25, batch 6700, loss[loss=0.1953, simple_loss=0.2751, pruned_loss=0.05773, over 19383.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2866, pruned_loss=0.06202, over 3819036.85 frames. ], batch size: 70, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:47:36,701 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=170587.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:48:17,851 INFO [train.py:903] (3/4) Epoch 25, batch 6750, loss[loss=0.1992, simple_loss=0.2907, pruned_loss=0.05384, over 19710.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.2855, pruned_loss=0.06152, over 3810413.01 frames. ], batch size: 59, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:48:59,810 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.047e+02 4.526e+02 5.801e+02 6.899e+02 1.670e+03, threshold=1.160e+03, percent-clipped=2.0 +2023-04-03 05:49:15,798 INFO [train.py:903] (3/4) Epoch 25, batch 6800, loss[loss=0.181, simple_loss=0.2595, pruned_loss=0.05126, over 19361.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2845, pruned_loss=0.0612, over 3817934.76 frames. ], batch size: 47, lr: 3.25e-03, grad_scale: 8.0 +2023-04-03 05:50:02,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 05:50:04,140 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 05:50:07,357 INFO [train.py:903] (3/4) Epoch 26, batch 0, loss[loss=0.1832, simple_loss=0.2607, pruned_loss=0.05279, over 19752.00 frames. ], tot_loss[loss=0.1832, simple_loss=0.2607, pruned_loss=0.05279, over 19752.00 frames. ], batch size: 47, lr: 3.19e-03, grad_scale: 8.0 +2023-04-03 05:50:07,357 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 05:50:19,303 INFO [train.py:937] (3/4) Epoch 26, validation: loss=0.1673, simple_loss=0.2675, pruned_loss=0.03355, over 944034.00 frames. +2023-04-03 05:50:19,305 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 05:50:32,180 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 05:50:44,728 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6746, 4.2346, 2.7595, 3.7330, 0.8655, 4.2225, 4.1134, 4.1277], + device='cuda:3'), covar=tensor([0.0607, 0.1124, 0.2028, 0.0899, 0.4237, 0.0712, 0.0967, 0.1182], + device='cuda:3'), in_proj_covar=tensor([0.0520, 0.0424, 0.0510, 0.0357, 0.0409, 0.0450, 0.0445, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:51:06,034 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.07 vs. limit=5.0 +2023-04-03 05:51:20,628 INFO [train.py:903] (3/4) Epoch 26, batch 50, loss[loss=0.2041, simple_loss=0.2814, pruned_loss=0.06337, over 19586.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2863, pruned_loss=0.06214, over 859335.05 frames. ], batch size: 52, lr: 3.19e-03, grad_scale: 8.0 +2023-04-03 05:51:28,226 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9075, 1.1944, 1.4581, 0.6126, 2.0237, 2.4606, 2.2110, 2.6306], + device='cuda:3'), covar=tensor([0.1780, 0.4265, 0.3765, 0.3049, 0.0708, 0.0301, 0.0358, 0.0429], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0328, 0.0358, 0.0268, 0.0250, 0.0191, 0.0218, 0.0269], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 05:51:30,184 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.374e+02 5.268e+02 6.216e+02 7.845e+02 1.668e+03, threshold=1.243e+03, percent-clipped=9.0 +2023-04-03 05:51:54,979 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=170778.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:51:55,730 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 05:52:04,556 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-03 05:52:21,931 INFO [train.py:903] (3/4) Epoch 26, batch 100, loss[loss=0.1908, simple_loss=0.2706, pruned_loss=0.05545, over 19576.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2871, pruned_loss=0.06288, over 1515004.02 frames. ], batch size: 52, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:52:26,029 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=170803.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:52:32,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 05:52:50,745 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170823.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:53:24,741 INFO [train.py:903] (3/4) Epoch 26, batch 150, loss[loss=0.207, simple_loss=0.2942, pruned_loss=0.05989, over 19667.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2854, pruned_loss=0.06221, over 2019548.27 frames. ], batch size: 55, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:53:36,324 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.111e+02 4.908e+02 6.470e+02 7.917e+02 1.560e+03, threshold=1.294e+03, percent-clipped=6.0 +2023-04-03 05:54:25,519 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 05:54:26,689 INFO [train.py:903] (3/4) Epoch 26, batch 200, loss[loss=0.1987, simple_loss=0.2744, pruned_loss=0.06147, over 19587.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2845, pruned_loss=0.06086, over 2431743.09 frames. ], batch size: 52, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:55:05,066 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=170931.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:55:14,504 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=170938.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:55:24,808 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=170946.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:55:30,413 INFO [train.py:903] (3/4) Epoch 26, batch 250, loss[loss=0.1762, simple_loss=0.2581, pruned_loss=0.04712, over 19724.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2849, pruned_loss=0.06132, over 2733755.80 frames. ], batch size: 51, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:55:31,162 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.70 vs. limit=2.0 +2023-04-03 05:55:39,639 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4256, 2.1661, 1.6878, 1.5125, 2.0207, 1.4202, 1.4077, 1.8883], + device='cuda:3'), covar=tensor([0.1109, 0.0876, 0.1080, 0.0907, 0.0601, 0.1329, 0.0744, 0.0497], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0318, 0.0336, 0.0270, 0.0249, 0.0342, 0.0294, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 05:55:42,602 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.950e+02 4.848e+02 5.950e+02 8.014e+02 1.769e+03, threshold=1.190e+03, percent-clipped=1.0 +2023-04-03 05:56:34,953 INFO [train.py:903] (3/4) Epoch 26, batch 300, loss[loss=0.1971, simple_loss=0.2878, pruned_loss=0.05317, over 19618.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2833, pruned_loss=0.06049, over 2990761.81 frames. ], batch size: 57, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:57:34,254 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171046.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:57:34,585 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.59 vs. limit=5.0 +2023-04-03 05:57:38,583 INFO [train.py:903] (3/4) Epoch 26, batch 350, loss[loss=0.1951, simple_loss=0.2736, pruned_loss=0.05828, over 19571.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2831, pruned_loss=0.06028, over 3178883.33 frames. ], batch size: 52, lr: 3.18e-03, grad_scale: 4.0 +2023-04-03 05:57:45,651 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 05:57:49,063 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.001e+02 4.764e+02 6.020e+02 7.720e+02 1.602e+03, threshold=1.204e+03, percent-clipped=4.0 +2023-04-03 05:58:42,176 INFO [train.py:903] (3/4) Epoch 26, batch 400, loss[loss=0.2304, simple_loss=0.3142, pruned_loss=0.07329, over 19357.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06028, over 3327392.67 frames. ], batch size: 70, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 05:58:55,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=171110.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 05:59:43,996 INFO [train.py:903] (3/4) Epoch 26, batch 450, loss[loss=0.1734, simple_loss=0.2574, pruned_loss=0.04471, over 19675.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2838, pruned_loss=0.06056, over 3439884.62 frames. ], batch size: 55, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 05:59:56,343 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.853e+02 5.217e+02 6.577e+02 9.059e+02 2.566e+03, threshold=1.315e+03, percent-clipped=7.0 +2023-04-03 06:00:19,585 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 06:00:20,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 06:00:30,528 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5359, 1.1082, 1.1773, 1.4357, 0.9983, 1.2953, 1.1608, 1.3532], + device='cuda:3'), covar=tensor([0.1189, 0.1404, 0.1621, 0.1017, 0.1429, 0.0672, 0.1682, 0.0897], + device='cuda:3'), in_proj_covar=tensor([0.0275, 0.0354, 0.0314, 0.0254, 0.0304, 0.0255, 0.0315, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:00:38,900 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=171194.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:00:47,725 INFO [train.py:903] (3/4) Epoch 26, batch 500, loss[loss=0.1802, simple_loss=0.2535, pruned_loss=0.05348, over 19774.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2837, pruned_loss=0.06072, over 3535727.24 frames. ], batch size: 49, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:01:05,590 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=171214.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:01:11,786 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=171219.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:01:52,223 INFO [train.py:903] (3/4) Epoch 26, batch 550, loss[loss=0.2919, simple_loss=0.3466, pruned_loss=0.1186, over 19675.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2842, pruned_loss=0.06108, over 3608918.03 frames. ], batch size: 60, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:02:03,094 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.195e+02 4.798e+02 6.471e+02 7.842e+02 1.459e+03, threshold=1.294e+03, percent-clipped=3.0 +2023-04-03 06:02:12,064 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 06:02:44,064 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=171290.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:02:55,675 INFO [train.py:903] (3/4) Epoch 26, batch 600, loss[loss=0.2259, simple_loss=0.3032, pruned_loss=0.07428, over 19324.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2841, pruned_loss=0.0608, over 3657623.63 frames. ], batch size: 66, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:02:58,422 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=171302.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:03:31,106 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=171327.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:03:36,407 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 06:03:57,549 INFO [train.py:903] (3/4) Epoch 26, batch 650, loss[loss=0.2053, simple_loss=0.2982, pruned_loss=0.05626, over 19714.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2838, pruned_loss=0.06097, over 3689902.91 frames. ], batch size: 63, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:04:09,335 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.689e+02 4.759e+02 5.860e+02 7.918e+02 1.260e+03, threshold=1.172e+03, percent-clipped=0.0 +2023-04-03 06:04:16,450 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=171363.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:05:01,712 INFO [train.py:903] (3/4) Epoch 26, batch 700, loss[loss=0.2074, simple_loss=0.2815, pruned_loss=0.06665, over 19482.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2825, pruned_loss=0.06043, over 3720610.03 frames. ], batch size: 49, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:05:10,683 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171405.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:06:08,107 INFO [train.py:903] (3/4) Epoch 26, batch 750, loss[loss=0.2683, simple_loss=0.3275, pruned_loss=0.1045, over 13549.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2842, pruned_loss=0.06172, over 3733174.22 frames. ], batch size: 136, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:06:11,870 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7910, 1.4737, 1.5713, 1.5973, 3.3704, 1.1544, 2.4787, 3.8012], + device='cuda:3'), covar=tensor([0.0494, 0.2776, 0.2972, 0.1876, 0.0677, 0.2580, 0.1311, 0.0230], + device='cuda:3'), in_proj_covar=tensor([0.0420, 0.0374, 0.0394, 0.0354, 0.0379, 0.0354, 0.0392, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:06:12,968 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=171454.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:06:18,671 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.178e+02 5.123e+02 6.550e+02 8.686e+02 2.549e+03, threshold=1.310e+03, percent-clipped=11.0 +2023-04-03 06:07:12,528 INFO [train.py:903] (3/4) Epoch 26, batch 800, loss[loss=0.1981, simple_loss=0.2902, pruned_loss=0.05302, over 19669.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2846, pruned_loss=0.06149, over 3748254.08 frames. ], batch size: 58, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:07:25,367 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 06:07:32,365 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=171516.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:08:14,608 INFO [train.py:903] (3/4) Epoch 26, batch 850, loss[loss=0.2007, simple_loss=0.286, pruned_loss=0.05769, over 19777.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2857, pruned_loss=0.06222, over 3769618.82 frames. ], batch size: 56, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:08:24,876 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=171558.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:08:25,743 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.196e+02 4.689e+02 5.699e+02 7.261e+02 1.636e+03, threshold=1.140e+03, percent-clipped=2.0 +2023-04-03 06:08:40,934 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171569.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:09:05,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 06:09:18,245 INFO [train.py:903] (3/4) Epoch 26, batch 900, loss[loss=0.2213, simple_loss=0.2997, pruned_loss=0.07148, over 19788.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2858, pruned_loss=0.06194, over 3790805.29 frames. ], batch size: 56, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:10:17,110 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.82 vs. limit=5.0 +2023-04-03 06:10:22,182 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 06:10:23,384 INFO [train.py:903] (3/4) Epoch 26, batch 950, loss[loss=0.1975, simple_loss=0.2888, pruned_loss=0.05311, over 19611.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.286, pruned_loss=0.06179, over 3810322.38 frames. ], batch size: 57, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:10:34,908 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.376e+02 5.213e+02 6.211e+02 8.451e+02 1.981e+03, threshold=1.242e+03, percent-clipped=10.0 +2023-04-03 06:10:37,599 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=171661.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:10:45,898 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9401, 1.3491, 1.0799, 0.8472, 1.1874, 0.9149, 0.9535, 1.2876], + device='cuda:3'), covar=tensor([0.0630, 0.0777, 0.1069, 0.0874, 0.0543, 0.1313, 0.0574, 0.0426], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0318, 0.0337, 0.0270, 0.0250, 0.0343, 0.0295, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:10:51,825 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171673.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:11:00,208 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2600, 2.3038, 2.0313, 2.3303, 2.2379, 1.8753, 1.9644, 2.2307], + device='cuda:3'), covar=tensor([0.0958, 0.1309, 0.1431, 0.1014, 0.1213, 0.0835, 0.1450, 0.0792], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0354, 0.0314, 0.0254, 0.0305, 0.0254, 0.0315, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:11:09,460 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=171686.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:11:27,631 INFO [train.py:903] (3/4) Epoch 26, batch 1000, loss[loss=0.1976, simple_loss=0.2818, pruned_loss=0.05668, over 19736.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2859, pruned_loss=0.06195, over 3808465.21 frames. ], batch size: 63, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:11:36,120 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=171707.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:12:19,826 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 06:12:31,841 INFO [train.py:903] (3/4) Epoch 26, batch 1050, loss[loss=0.2671, simple_loss=0.3327, pruned_loss=0.1007, over 13761.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2845, pruned_loss=0.06105, over 3816641.26 frames. ], batch size: 135, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:12:42,470 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.548e+02 5.465e+02 6.383e+02 7.807e+02 1.569e+03, threshold=1.277e+03, percent-clipped=7.0 +2023-04-03 06:12:50,010 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.62 vs. limit=2.0 +2023-04-03 06:13:01,890 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 06:13:34,460 INFO [train.py:903] (3/4) Epoch 26, batch 1100, loss[loss=0.2159, simple_loss=0.2986, pruned_loss=0.06665, over 19590.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2851, pruned_loss=0.0614, over 3824836.13 frames. ], batch size: 57, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:14:05,456 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171822.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:14:09,068 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=171825.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:14:40,968 INFO [train.py:903] (3/4) Epoch 26, batch 1150, loss[loss=0.2209, simple_loss=0.2985, pruned_loss=0.07159, over 19617.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2835, pruned_loss=0.06021, over 3835249.03 frames. ], batch size: 57, lr: 3.18e-03, grad_scale: 8.0 +2023-04-03 06:14:41,487 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=171850.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:14:54,273 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.309e+02 5.298e+02 6.784e+02 8.276e+02 1.649e+03, threshold=1.357e+03, percent-clipped=5.0 +2023-04-03 06:14:55,527 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=171860.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:15:46,139 INFO [train.py:903] (3/4) Epoch 26, batch 1200, loss[loss=0.1783, simple_loss=0.2626, pruned_loss=0.04701, over 19853.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2834, pruned_loss=0.06035, over 3833402.43 frames. ], batch size: 52, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:16:12,790 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 06:16:22,268 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=171929.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:16:29,742 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-03 06:16:48,465 INFO [train.py:903] (3/4) Epoch 26, batch 1250, loss[loss=0.2095, simple_loss=0.2881, pruned_loss=0.0654, over 19753.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.284, pruned_loss=0.06082, over 3824238.16 frames. ], batch size: 51, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:16:53,750 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=171954.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:16:54,742 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=171955.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:16:58,916 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.249e+02 5.182e+02 6.196e+02 7.754e+02 1.405e+03, threshold=1.239e+03, percent-clipped=1.0 +2023-04-03 06:17:21,629 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=171975.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:17:51,744 INFO [train.py:903] (3/4) Epoch 26, batch 1300, loss[loss=0.1952, simple_loss=0.2864, pruned_loss=0.05202, over 19681.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2834, pruned_loss=0.06035, over 3826027.43 frames. ], batch size: 59, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:18:56,743 INFO [train.py:903] (3/4) Epoch 26, batch 1350, loss[loss=0.2338, simple_loss=0.3117, pruned_loss=0.07794, over 19579.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2831, pruned_loss=0.06017, over 3829821.09 frames. ], batch size: 61, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:18:59,494 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1122, 5.6009, 2.7868, 4.7759, 1.1796, 5.7163, 5.4907, 5.6684], + device='cuda:3'), covar=tensor([0.0354, 0.0719, 0.1983, 0.0708, 0.3662, 0.0461, 0.0790, 0.0851], + device='cuda:3'), in_proj_covar=tensor([0.0521, 0.0423, 0.0510, 0.0357, 0.0409, 0.0449, 0.0444, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:19:09,279 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-03 06:19:09,471 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.423e+02 5.487e+02 6.725e+02 8.152e+02 1.378e+03, threshold=1.345e+03, percent-clipped=4.0 +2023-04-03 06:19:16,130 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.57 vs. limit=2.0 +2023-04-03 06:19:25,635 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0402, 1.9793, 1.7141, 2.1040, 1.9285, 1.7939, 1.6601, 1.9466], + device='cuda:3'), covar=tensor([0.1139, 0.1509, 0.1582, 0.1136, 0.1409, 0.0581, 0.1549, 0.0791], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0357, 0.0318, 0.0256, 0.0308, 0.0256, 0.0318, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:19:34,023 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=172078.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:20:01,391 INFO [train.py:903] (3/4) Epoch 26, batch 1400, loss[loss=0.2327, simple_loss=0.3096, pruned_loss=0.07786, over 19600.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2826, pruned_loss=0.05968, over 3844393.79 frames. ], batch size: 61, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:20:07,127 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=172103.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:20:47,887 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=172136.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:21:05,089 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 06:21:06,112 INFO [train.py:903] (3/4) Epoch 26, batch 1450, loss[loss=0.228, simple_loss=0.3022, pruned_loss=0.07696, over 19529.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2845, pruned_loss=0.06072, over 3836156.01 frames. ], batch size: 56, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:21:16,564 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.245e+02 4.778e+02 6.066e+02 7.310e+02 2.231e+03, threshold=1.213e+03, percent-clipped=2.0 +2023-04-03 06:21:22,865 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4354, 1.4627, 1.6060, 1.5943, 1.7674, 1.9332, 1.7913, 0.5370], + device='cuda:3'), covar=tensor([0.2464, 0.4378, 0.2787, 0.2042, 0.1758, 0.2363, 0.1570, 0.5162], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0662, 0.0740, 0.0500, 0.0630, 0.0541, 0.0667, 0.0565], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 06:22:09,972 INFO [train.py:903] (3/4) Epoch 26, batch 1500, loss[loss=0.2091, simple_loss=0.294, pruned_loss=0.06209, over 19502.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2845, pruned_loss=0.06112, over 3834277.92 frames. ], batch size: 64, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:22:33,633 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0188, 3.6357, 2.4427, 3.2966, 1.0682, 3.6618, 3.5526, 3.5955], + device='cuda:3'), covar=tensor([0.0814, 0.1220, 0.2142, 0.1002, 0.3734, 0.0812, 0.1115, 0.1352], + device='cuda:3'), in_proj_covar=tensor([0.0522, 0.0425, 0.0513, 0.0358, 0.0410, 0.0452, 0.0446, 0.0476], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:22:51,466 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=172231.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:23:14,778 INFO [train.py:903] (3/4) Epoch 26, batch 1550, loss[loss=0.2047, simple_loss=0.2937, pruned_loss=0.05787, over 19697.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2848, pruned_loss=0.06104, over 3845906.40 frames. ], batch size: 59, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:23:23,380 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=172256.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:23:26,539 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.787e+02 4.585e+02 5.725e+02 7.044e+02 1.122e+03, threshold=1.145e+03, percent-clipped=0.0 +2023-04-03 06:23:59,126 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6676, 1.5544, 1.7292, 1.7406, 3.2330, 1.4997, 2.7769, 3.6631], + device='cuda:3'), covar=tensor([0.0638, 0.3002, 0.2972, 0.2127, 0.0811, 0.2671, 0.1471, 0.0319], + device='cuda:3'), in_proj_covar=tensor([0.0420, 0.0374, 0.0395, 0.0354, 0.0380, 0.0354, 0.0393, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:24:17,373 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=172299.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:24:18,407 INFO [train.py:903] (3/4) Epoch 26, batch 1600, loss[loss=0.2043, simple_loss=0.2854, pruned_loss=0.06158, over 19766.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2845, pruned_loss=0.0612, over 3846994.08 frames. ], batch size: 54, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:24:30,664 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 06:24:45,012 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 06:24:49,285 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1381, 3.4052, 1.9673, 2.0680, 3.1057, 1.7323, 1.4964, 2.3597], + device='cuda:3'), covar=tensor([0.1490, 0.0719, 0.1174, 0.0963, 0.0550, 0.1349, 0.1088, 0.0715], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0320, 0.0340, 0.0272, 0.0251, 0.0345, 0.0294, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:25:22,886 INFO [train.py:903] (3/4) Epoch 26, batch 1650, loss[loss=0.2643, simple_loss=0.3396, pruned_loss=0.09456, over 19674.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2848, pruned_loss=0.06128, over 3848082.53 frames. ], batch size: 58, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:25:32,975 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.237e+02 4.806e+02 6.116e+02 7.815e+02 1.931e+03, threshold=1.223e+03, percent-clipped=6.0 +2023-04-03 06:26:24,985 INFO [train.py:903] (3/4) Epoch 26, batch 1700, loss[loss=0.2592, simple_loss=0.3257, pruned_loss=0.09637, over 12871.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2844, pruned_loss=0.06161, over 3837993.97 frames. ], batch size: 136, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:26:27,915 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-04-03 06:26:43,124 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=172414.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:27:07,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 06:27:27,220 INFO [train.py:903] (3/4) Epoch 26, batch 1750, loss[loss=0.1816, simple_loss=0.2645, pruned_loss=0.04942, over 19728.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.284, pruned_loss=0.06127, over 3832523.74 frames. ], batch size: 51, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:27:39,757 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.268e+02 4.883e+02 5.717e+02 7.034e+02 1.807e+03, threshold=1.143e+03, percent-clipped=3.0 +2023-04-03 06:27:52,745 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9015, 1.8713, 1.4835, 1.8671, 1.7793, 1.5266, 1.4775, 1.7721], + device='cuda:3'), covar=tensor([0.1215, 0.1511, 0.1838, 0.1221, 0.1473, 0.0941, 0.1859, 0.1001], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0356, 0.0315, 0.0254, 0.0307, 0.0254, 0.0316, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:28:06,465 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=172480.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:28:09,158 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8736, 1.2048, 1.6187, 0.5714, 2.0267, 2.4714, 2.1927, 2.6537], + device='cuda:3'), covar=tensor([0.1682, 0.3996, 0.3434, 0.2918, 0.0671, 0.0305, 0.0358, 0.0412], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0330, 0.0361, 0.0270, 0.0252, 0.0193, 0.0219, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 06:28:26,586 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8517, 4.4259, 2.7661, 3.8923, 0.9562, 4.3643, 4.2554, 4.3161], + device='cuda:3'), covar=tensor([0.0537, 0.0865, 0.1913, 0.0816, 0.3821, 0.0625, 0.0903, 0.1034], + device='cuda:3'), in_proj_covar=tensor([0.0518, 0.0420, 0.0507, 0.0355, 0.0406, 0.0447, 0.0441, 0.0472], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:28:32,020 INFO [train.py:903] (3/4) Epoch 26, batch 1800, loss[loss=0.217, simple_loss=0.2954, pruned_loss=0.06932, over 19774.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2833, pruned_loss=0.06078, over 3845614.50 frames. ], batch size: 54, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:29:31,754 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 06:29:36,206 INFO [train.py:903] (3/4) Epoch 26, batch 1850, loss[loss=0.174, simple_loss=0.2515, pruned_loss=0.04824, over 19154.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2835, pruned_loss=0.06083, over 3838233.97 frames. ], batch size: 42, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:29:46,975 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.860e+02 4.880e+02 5.794e+02 7.257e+02 1.575e+03, threshold=1.159e+03, percent-clipped=2.0 +2023-04-03 06:30:11,359 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 06:30:20,758 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0264, 4.6123, 2.7863, 4.0693, 1.1376, 4.5351, 4.4630, 4.5223], + device='cuda:3'), covar=tensor([0.0497, 0.0847, 0.1888, 0.0799, 0.3626, 0.0566, 0.0848, 0.0885], + device='cuda:3'), in_proj_covar=tensor([0.0521, 0.0423, 0.0511, 0.0357, 0.0407, 0.0450, 0.0443, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:30:34,192 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=172595.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:30:39,839 INFO [train.py:903] (3/4) Epoch 26, batch 1900, loss[loss=0.2186, simple_loss=0.3074, pruned_loss=0.06487, over 19369.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2824, pruned_loss=0.06027, over 3836712.72 frames. ], batch size: 66, lr: 3.17e-03, grad_scale: 4.0 +2023-04-03 06:30:59,239 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 06:31:01,809 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=172616.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:31:05,048 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 06:31:23,069 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9053, 4.3726, 4.6684, 4.6607, 1.6713, 4.3371, 3.7741, 4.3688], + device='cuda:3'), covar=tensor([0.1906, 0.0873, 0.0639, 0.0723, 0.6559, 0.0991, 0.0732, 0.1320], + device='cuda:3'), in_proj_covar=tensor([0.0806, 0.0774, 0.0978, 0.0859, 0.0852, 0.0745, 0.0584, 0.0906], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 06:31:29,995 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 06:31:36,115 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=172643.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:31:44,211 INFO [train.py:903] (3/4) Epoch 26, batch 1950, loss[loss=0.2231, simple_loss=0.3033, pruned_loss=0.07144, over 19660.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2832, pruned_loss=0.06059, over 3834127.95 frames. ], batch size: 60, lr: 3.17e-03, grad_scale: 4.0 +2023-04-03 06:31:57,688 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.017e+02 5.105e+02 6.247e+02 7.487e+02 1.872e+03, threshold=1.249e+03, percent-clipped=7.0 +2023-04-03 06:32:12,276 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=172670.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:32:43,425 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=172695.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:32:48,408 INFO [train.py:903] (3/4) Epoch 26, batch 2000, loss[loss=0.2246, simple_loss=0.3032, pruned_loss=0.07301, over 19789.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2834, pruned_loss=0.0606, over 3843421.92 frames. ], batch size: 56, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:33:14,449 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 06:33:39,914 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9031, 4.2856, 4.6176, 4.6116, 2.1387, 4.3186, 3.7966, 4.3609], + device='cuda:3'), covar=tensor([0.1701, 0.1424, 0.0587, 0.0691, 0.5547, 0.1127, 0.0672, 0.1003], + device='cuda:3'), in_proj_covar=tensor([0.0803, 0.0774, 0.0978, 0.0859, 0.0852, 0.0744, 0.0584, 0.0905], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 06:33:46,801 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 06:33:51,657 INFO [train.py:903] (3/4) Epoch 26, batch 2050, loss[loss=0.2107, simple_loss=0.274, pruned_loss=0.07369, over 19711.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.284, pruned_loss=0.06126, over 3830580.17 frames. ], batch size: 46, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:33:55,301 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=172752.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:34:03,947 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.396e+02 5.217e+02 6.186e+02 8.512e+02 2.102e+03, threshold=1.237e+03, percent-clipped=6.0 +2023-04-03 06:34:06,456 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 06:34:07,797 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 06:34:27,709 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 06:34:54,268 INFO [train.py:903] (3/4) Epoch 26, batch 2100, loss[loss=0.1779, simple_loss=0.2712, pruned_loss=0.04229, over 19570.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2836, pruned_loss=0.06111, over 3833256.79 frames. ], batch size: 52, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:35:25,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 06:35:47,309 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 06:35:56,679 INFO [train.py:903] (3/4) Epoch 26, batch 2150, loss[loss=0.153, simple_loss=0.2375, pruned_loss=0.03422, over 19742.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.283, pruned_loss=0.06091, over 3834916.56 frames. ], batch size: 48, lr: 3.17e-03, grad_scale: 8.0 +2023-04-03 06:35:58,327 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=172851.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:36:02,861 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6141, 1.3334, 1.5281, 1.5603, 3.1730, 1.0522, 2.4560, 3.6453], + device='cuda:3'), covar=tensor([0.0496, 0.2916, 0.2944, 0.1864, 0.0717, 0.2543, 0.1288, 0.0233], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0371, 0.0392, 0.0351, 0.0378, 0.0352, 0.0389, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:36:10,068 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.604e+02 4.749e+02 6.121e+02 8.086e+02 1.727e+03, threshold=1.224e+03, percent-clipped=6.0 +2023-04-03 06:36:14,750 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=172863.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:36:21,385 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-04-03 06:36:31,089 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=172876.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:36:59,989 INFO [train.py:903] (3/4) Epoch 26, batch 2200, loss[loss=0.1851, simple_loss=0.2749, pruned_loss=0.04761, over 19467.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.284, pruned_loss=0.06127, over 3828316.21 frames. ], batch size: 49, lr: 3.17e-03, grad_scale: 4.0 +2023-04-03 06:38:04,037 INFO [train.py:903] (3/4) Epoch 26, batch 2250, loss[loss=0.1894, simple_loss=0.2796, pruned_loss=0.04963, over 19764.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2845, pruned_loss=0.06131, over 3829052.51 frames. ], batch size: 54, lr: 3.16e-03, grad_scale: 4.0 +2023-04-03 06:38:16,977 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=172960.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:38:17,990 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 4.806e+02 5.924e+02 7.729e+02 1.368e+03, threshold=1.185e+03, percent-clipped=2.0 +2023-04-03 06:38:51,894 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=172987.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:39:07,853 INFO [train.py:903] (3/4) Epoch 26, batch 2300, loss[loss=0.1722, simple_loss=0.2583, pruned_loss=0.04306, over 19767.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2846, pruned_loss=0.06167, over 3822945.67 frames. ], batch size: 54, lr: 3.16e-03, grad_scale: 4.0 +2023-04-03 06:39:13,393 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-03 06:39:19,742 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 06:40:11,067 INFO [train.py:903] (3/4) Epoch 26, batch 2350, loss[loss=0.1838, simple_loss=0.2793, pruned_loss=0.04411, over 19781.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2832, pruned_loss=0.06076, over 3835238.77 frames. ], batch size: 56, lr: 3.16e-03, grad_scale: 4.0 +2023-04-03 06:40:25,945 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.095e+02 5.175e+02 6.518e+02 7.971e+02 2.695e+03, threshold=1.304e+03, percent-clipped=8.0 +2023-04-03 06:40:43,713 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173075.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:40:54,414 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 06:41:08,290 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173096.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:41:10,530 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 06:41:14,056 INFO [train.py:903] (3/4) Epoch 26, batch 2400, loss[loss=0.2054, simple_loss=0.2798, pruned_loss=0.06551, over 19611.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2836, pruned_loss=0.06106, over 3832090.82 frames. ], batch size: 50, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:41:18,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173102.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:41:48,405 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3391, 1.5890, 2.1076, 1.6111, 3.0881, 4.7584, 4.5893, 5.1685], + device='cuda:3'), covar=tensor([0.1620, 0.3681, 0.3165, 0.2407, 0.0635, 0.0176, 0.0169, 0.0191], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0330, 0.0362, 0.0271, 0.0252, 0.0193, 0.0220, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 06:42:19,069 INFO [train.py:903] (3/4) Epoch 26, batch 2450, loss[loss=0.1783, simple_loss=0.2701, pruned_loss=0.04326, over 18278.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2837, pruned_loss=0.06097, over 3836267.02 frames. ], batch size: 40, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:42:32,913 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.093e+02 4.972e+02 5.973e+02 7.732e+02 1.743e+03, threshold=1.195e+03, percent-clipped=3.0 +2023-04-03 06:42:41,761 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=173168.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:43:22,682 INFO [train.py:903] (3/4) Epoch 26, batch 2500, loss[loss=0.1825, simple_loss=0.2708, pruned_loss=0.04708, over 19777.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2833, pruned_loss=0.0607, over 3822413.55 frames. ], batch size: 54, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:43:31,272 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173207.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:43:36,111 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173211.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:44:25,912 INFO [train.py:903] (3/4) Epoch 26, batch 2550, loss[loss=0.2117, simple_loss=0.2897, pruned_loss=0.06687, over 17415.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2831, pruned_loss=0.06055, over 3834947.28 frames. ], batch size: 101, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:44:40,272 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3687, 2.0250, 1.5932, 1.3762, 1.8238, 1.2823, 1.3935, 1.7946], + device='cuda:3'), covar=tensor([0.0962, 0.0793, 0.1073, 0.0884, 0.0606, 0.1293, 0.0658, 0.0498], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0317, 0.0337, 0.0269, 0.0248, 0.0341, 0.0291, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:44:40,983 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.733e+02 4.982e+02 5.984e+02 8.594e+02 2.255e+03, threshold=1.197e+03, percent-clipped=6.0 +2023-04-03 06:45:23,060 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 06:45:30,064 INFO [train.py:903] (3/4) Epoch 26, batch 2600, loss[loss=0.1863, simple_loss=0.2688, pruned_loss=0.0519, over 19682.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2832, pruned_loss=0.06046, over 3838911.09 frames. ], batch size: 53, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:45:59,771 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173322.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:46:05,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=173327.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:46:05,522 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4305, 2.0335, 2.1291, 2.9939, 1.9463, 2.4519, 2.5130, 2.2623], + device='cuda:3'), covar=tensor([0.0737, 0.0922, 0.0914, 0.0760, 0.0887, 0.0780, 0.0900, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0226, 0.0228, 0.0242, 0.0228, 0.0214, 0.0190, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 06:46:10,053 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=173331.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:46:35,582 INFO [train.py:903] (3/4) Epoch 26, batch 2650, loss[loss=0.2221, simple_loss=0.2986, pruned_loss=0.07277, over 18785.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2828, pruned_loss=0.06035, over 3828860.15 frames. ], batch size: 74, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:46:42,202 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 06:46:43,071 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=173356.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:46:45,486 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=173358.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:46:49,593 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.269e+02 5.033e+02 5.962e+02 7.363e+02 1.395e+03, threshold=1.192e+03, percent-clipped=2.0 +2023-04-03 06:46:55,536 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 06:47:17,711 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=173383.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:47:39,142 INFO [train.py:903] (3/4) Epoch 26, batch 2700, loss[loss=0.1982, simple_loss=0.2868, pruned_loss=0.05486, over 19569.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.283, pruned_loss=0.06019, over 3834176.38 frames. ], batch size: 61, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:48:00,262 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0249, 4.4604, 4.8036, 4.8034, 1.7251, 4.4965, 3.8766, 4.5145], + device='cuda:3'), covar=tensor([0.1764, 0.0883, 0.0602, 0.0702, 0.6366, 0.0868, 0.0684, 0.1155], + device='cuda:3'), in_proj_covar=tensor([0.0806, 0.0773, 0.0978, 0.0861, 0.0856, 0.0742, 0.0584, 0.0909], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 06:48:32,625 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7951, 1.8621, 2.1335, 2.2162, 1.6853, 2.1164, 2.1017, 1.9591], + device='cuda:3'), covar=tensor([0.4175, 0.3930, 0.2030, 0.2503, 0.4085, 0.2304, 0.5070, 0.3447], + device='cuda:3'), in_proj_covar=tensor([0.0928, 0.1004, 0.0736, 0.0949, 0.0908, 0.0841, 0.0859, 0.0805], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 06:48:39,711 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-03 06:48:41,208 INFO [train.py:903] (3/4) Epoch 26, batch 2750, loss[loss=0.2198, simple_loss=0.2991, pruned_loss=0.0703, over 19586.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2832, pruned_loss=0.05998, over 3829654.14 frames. ], batch size: 61, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:48:54,809 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.847e+02 4.981e+02 6.163e+02 7.639e+02 1.916e+03, threshold=1.233e+03, percent-clipped=5.0 +2023-04-03 06:49:03,502 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=173467.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:49:34,182 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=173492.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:49:44,193 INFO [train.py:903] (3/4) Epoch 26, batch 2800, loss[loss=0.215, simple_loss=0.2926, pruned_loss=0.06869, over 19604.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2834, pruned_loss=0.0604, over 3811102.11 frames. ], batch size: 57, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:49:52,990 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4604, 1.5164, 1.7436, 1.7361, 2.6287, 2.2567, 2.7945, 1.2266], + device='cuda:3'), covar=tensor([0.2540, 0.4347, 0.2745, 0.1948, 0.1497, 0.2161, 0.1388, 0.4641], + device='cuda:3'), in_proj_covar=tensor([0.0545, 0.0661, 0.0739, 0.0500, 0.0629, 0.0539, 0.0666, 0.0566], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 06:49:56,459 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=173509.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 06:49:56,869 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.56 vs. limit=5.0 +2023-04-03 06:50:00,282 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173512.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:50:48,002 INFO [train.py:903] (3/4) Epoch 26, batch 2850, loss[loss=0.2095, simple_loss=0.2945, pruned_loss=0.06223, over 19629.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2844, pruned_loss=0.06066, over 3817754.56 frames. ], batch size: 50, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:51:01,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.227e+02 4.954e+02 6.202e+02 8.183e+02 1.932e+03, threshold=1.240e+03, percent-clipped=7.0 +2023-04-03 06:51:23,220 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=173578.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:51:51,137 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 06:51:52,155 INFO [train.py:903] (3/4) Epoch 26, batch 2900, loss[loss=0.2523, simple_loss=0.3183, pruned_loss=0.0932, over 19607.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2851, pruned_loss=0.06115, over 3820166.81 frames. ], batch size: 57, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:51:57,199 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=173603.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:52:08,019 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9877, 2.0492, 2.3584, 2.5612, 1.9680, 2.5125, 2.3283, 2.1710], + device='cuda:3'), covar=tensor([0.4283, 0.4063, 0.1911, 0.2539, 0.4252, 0.2196, 0.4975, 0.3331], + device='cuda:3'), in_proj_covar=tensor([0.0928, 0.1003, 0.0736, 0.0946, 0.0907, 0.0840, 0.0859, 0.0805], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 06:52:27,628 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173627.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:52:34,772 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=173632.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:52:56,769 INFO [train.py:903] (3/4) Epoch 26, batch 2950, loss[loss=0.2032, simple_loss=0.2883, pruned_loss=0.05901, over 19780.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2849, pruned_loss=0.06121, over 3820699.74 frames. ], batch size: 56, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:53:10,750 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.057e+02 5.287e+02 6.719e+02 8.706e+02 2.181e+03, threshold=1.344e+03, percent-clipped=5.0 +2023-04-03 06:53:23,945 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173671.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:53:59,627 INFO [train.py:903] (3/4) Epoch 26, batch 3000, loss[loss=0.1973, simple_loss=0.2839, pruned_loss=0.0553, over 19688.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2841, pruned_loss=0.06086, over 3830366.00 frames. ], batch size: 59, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:53:59,627 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 06:54:12,257 INFO [train.py:937] (3/4) Epoch 26, validation: loss=0.1681, simple_loss=0.2675, pruned_loss=0.03435, over 944034.00 frames. +2023-04-03 06:54:12,259 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 06:54:17,259 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 06:55:16,507 INFO [train.py:903] (3/4) Epoch 26, batch 3050, loss[loss=0.172, simple_loss=0.2526, pruned_loss=0.04572, over 19401.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2844, pruned_loss=0.0609, over 3825496.13 frames. ], batch size: 48, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:55:30,851 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.198e+02 4.825e+02 5.950e+02 7.456e+02 1.374e+03, threshold=1.190e+03, percent-clipped=1.0 +2023-04-03 06:56:02,634 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173786.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:56:20,031 INFO [train.py:903] (3/4) Epoch 26, batch 3100, loss[loss=0.1731, simple_loss=0.2519, pruned_loss=0.04712, over 19763.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2834, pruned_loss=0.06032, over 3832235.37 frames. ], batch size: 48, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:56:32,224 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-04-03 06:56:49,601 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9059, 1.2884, 1.6275, 0.5685, 1.9463, 2.4421, 2.1464, 2.5955], + device='cuda:3'), covar=tensor([0.1638, 0.3850, 0.3375, 0.2830, 0.0678, 0.0273, 0.0341, 0.0386], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0328, 0.0361, 0.0269, 0.0252, 0.0193, 0.0218, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 06:57:23,192 INFO [train.py:903] (3/4) Epoch 26, batch 3150, loss[loss=0.1896, simple_loss=0.275, pruned_loss=0.05207, over 19586.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2843, pruned_loss=0.06118, over 3815381.97 frames. ], batch size: 52, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:57:23,615 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1255, 1.2395, 1.7313, 1.0336, 2.3770, 3.0846, 2.8264, 3.2798], + device='cuda:3'), covar=tensor([0.1599, 0.3897, 0.3237, 0.2617, 0.0591, 0.0217, 0.0243, 0.0331], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0328, 0.0361, 0.0269, 0.0252, 0.0193, 0.0218, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 06:57:26,865 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173853.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 06:57:37,138 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.975e+02 5.081e+02 6.233e+02 7.406e+02 2.417e+03, threshold=1.247e+03, percent-clipped=3.0 +2023-04-03 06:57:49,125 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 06:58:00,852 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-03 06:58:02,702 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.1294, 5.5992, 2.9915, 4.7742, 0.9397, 5.7123, 5.4859, 5.7156], + device='cuda:3'), covar=tensor([0.0364, 0.0789, 0.2009, 0.0724, 0.4274, 0.0553, 0.0864, 0.0943], + device='cuda:3'), in_proj_covar=tensor([0.0525, 0.0426, 0.0513, 0.0359, 0.0411, 0.0451, 0.0448, 0.0476], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:58:05,087 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=173883.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:58:25,322 INFO [train.py:903] (3/4) Epoch 26, batch 3200, loss[loss=0.2349, simple_loss=0.3085, pruned_loss=0.08066, over 18123.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06152, over 3813406.10 frames. ], batch size: 83, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:58:34,884 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9873, 1.6979, 1.5673, 1.8761, 1.6616, 1.6471, 1.4844, 1.8267], + device='cuda:3'), covar=tensor([0.1077, 0.1474, 0.1622, 0.1067, 0.1441, 0.0641, 0.1683, 0.0851], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0359, 0.0316, 0.0256, 0.0307, 0.0256, 0.0319, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:58:35,893 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=173908.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 06:59:25,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6068, 1.9830, 2.1061, 2.0730, 3.3416, 1.7450, 2.8601, 3.5656], + device='cuda:3'), covar=tensor([0.0442, 0.2255, 0.2319, 0.1617, 0.0501, 0.2065, 0.1705, 0.0280], + device='cuda:3'), in_proj_covar=tensor([0.0418, 0.0373, 0.0392, 0.0351, 0.0377, 0.0354, 0.0391, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 06:59:27,977 INFO [train.py:903] (3/4) Epoch 26, batch 3250, loss[loss=0.2219, simple_loss=0.3036, pruned_loss=0.07005, over 19329.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2855, pruned_loss=0.06193, over 3806420.85 frames. ], batch size: 66, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 06:59:42,863 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.716e+02 4.798e+02 6.185e+02 8.155e+02 2.789e+03, threshold=1.237e+03, percent-clipped=7.0 +2023-04-03 06:59:52,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=173968.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 07:00:01,804 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=173976.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:00:31,303 INFO [train.py:903] (3/4) Epoch 26, batch 3300, loss[loss=0.2165, simple_loss=0.2872, pruned_loss=0.0729, over 19610.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.285, pruned_loss=0.06181, over 3809330.62 frames. ], batch size: 50, lr: 3.16e-03, grad_scale: 8.0 +2023-04-03 07:00:35,708 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 07:00:38,017 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9676, 1.8028, 1.7835, 2.0107, 1.8813, 1.7057, 1.7261, 1.9104], + device='cuda:3'), covar=tensor([0.0873, 0.1270, 0.1191, 0.0821, 0.1027, 0.0526, 0.1242, 0.0655], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0358, 0.0316, 0.0256, 0.0306, 0.0256, 0.0319, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:01:14,204 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174033.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:01:17,518 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174035.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:01:26,023 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174042.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:01:35,839 INFO [train.py:903] (3/4) Epoch 26, batch 3350, loss[loss=0.2249, simple_loss=0.2914, pruned_loss=0.07919, over 19733.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2841, pruned_loss=0.06119, over 3822511.41 frames. ], batch size: 51, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:01:49,566 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.212e+02 4.999e+02 5.843e+02 7.881e+02 1.777e+03, threshold=1.169e+03, percent-clipped=4.0 +2023-04-03 07:01:56,819 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174067.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:02:28,303 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174091.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:02:39,919 INFO [train.py:903] (3/4) Epoch 26, batch 3400, loss[loss=0.1825, simple_loss=0.2588, pruned_loss=0.05309, over 19751.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2843, pruned_loss=0.06132, over 3815721.97 frames. ], batch size: 45, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:03:15,052 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 07:03:41,102 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 07:03:42,262 INFO [train.py:903] (3/4) Epoch 26, batch 3450, loss[loss=0.2057, simple_loss=0.2885, pruned_loss=0.06143, over 19656.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2846, pruned_loss=0.06152, over 3815350.97 frames. ], batch size: 55, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:03:53,287 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174157.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:03:57,816 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.388e+02 4.666e+02 5.900e+02 7.449e+02 1.550e+03, threshold=1.180e+03, percent-clipped=6.0 +2023-04-03 07:04:02,620 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1360, 1.1844, 1.4483, 1.3641, 2.7225, 1.0706, 2.1684, 3.0901], + device='cuda:3'), covar=tensor([0.0598, 0.3106, 0.3029, 0.1980, 0.0796, 0.2520, 0.1292, 0.0330], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0376, 0.0395, 0.0353, 0.0382, 0.0357, 0.0394, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:04:47,404 INFO [train.py:903] (3/4) Epoch 26, batch 3500, loss[loss=0.1889, simple_loss=0.2646, pruned_loss=0.05658, over 19613.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2851, pruned_loss=0.06203, over 3814574.68 frames. ], batch size: 50, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:05:18,052 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174224.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 07:05:49,655 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174249.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 07:05:50,383 INFO [train.py:903] (3/4) Epoch 26, batch 3550, loss[loss=0.208, simple_loss=0.2894, pruned_loss=0.06335, over 19542.00 frames. ], tot_loss[loss=0.2062, simple_loss=0.2864, pruned_loss=0.06299, over 3799929.40 frames. ], batch size: 56, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:06:03,179 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.093e+02 4.676e+02 6.063e+02 7.972e+02 1.969e+03, threshold=1.213e+03, percent-clipped=7.0 +2023-04-03 07:06:12,983 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5105, 1.5690, 1.8690, 1.7509, 2.7218, 2.3363, 2.9459, 1.3171], + device='cuda:3'), covar=tensor([0.2436, 0.4198, 0.2707, 0.1908, 0.1449, 0.2077, 0.1333, 0.4462], + device='cuda:3'), in_proj_covar=tensor([0.0545, 0.0664, 0.0740, 0.0499, 0.0628, 0.0539, 0.0664, 0.0567], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 07:06:17,607 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3045, 1.4973, 1.9630, 1.5126, 2.9342, 4.6607, 4.4993, 5.1159], + device='cuda:3'), covar=tensor([0.1654, 0.3818, 0.3344, 0.2565, 0.0712, 0.0216, 0.0176, 0.0222], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0327, 0.0361, 0.0269, 0.0250, 0.0193, 0.0218, 0.0270], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 07:06:53,369 INFO [train.py:903] (3/4) Epoch 26, batch 3600, loss[loss=0.1759, simple_loss=0.2674, pruned_loss=0.04223, over 19774.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2843, pruned_loss=0.06164, over 3819401.79 frames. ], batch size: 54, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:06:54,925 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174301.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:07:05,326 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174309.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:07:54,127 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174347.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:07:57,119 INFO [train.py:903] (3/4) Epoch 26, batch 3650, loss[loss=0.2291, simple_loss=0.3014, pruned_loss=0.07843, over 19690.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2838, pruned_loss=0.06135, over 3828486.93 frames. ], batch size: 53, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:08:12,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 5.064e+02 6.896e+02 8.623e+02 2.807e+03, threshold=1.379e+03, percent-clipped=9.0 +2023-04-03 07:08:26,880 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174372.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:08:28,386 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 07:08:32,716 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174377.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:08:35,179 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174379.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:09:02,303 INFO [train.py:903] (3/4) Epoch 26, batch 3700, loss[loss=0.2145, simple_loss=0.2854, pruned_loss=0.07178, over 19630.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2845, pruned_loss=0.06162, over 3829726.42 frames. ], batch size: 50, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:09:10,646 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174406.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:09:15,242 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1384, 2.8831, 2.2769, 2.2532, 2.0220, 2.5266, 1.0663, 2.1066], + device='cuda:3'), covar=tensor([0.0801, 0.0625, 0.0797, 0.1266, 0.1241, 0.1133, 0.1530, 0.1132], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0362, 0.0367, 0.0391, 0.0468, 0.0396, 0.0346, 0.0346], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 07:09:15,871 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-03 07:10:07,621 INFO [train.py:903] (3/4) Epoch 26, batch 3750, loss[loss=0.2488, simple_loss=0.3176, pruned_loss=0.09002, over 19796.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2843, pruned_loss=0.06122, over 3829131.28 frames. ], batch size: 56, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:10:20,548 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.722e+02 4.995e+02 5.757e+02 7.069e+02 1.518e+03, threshold=1.151e+03, percent-clipped=1.0 +2023-04-03 07:10:24,805 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.09 vs. limit=2.0 +2023-04-03 07:11:01,670 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174492.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:11:04,059 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174494.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:11:10,963 INFO [train.py:903] (3/4) Epoch 26, batch 3800, loss[loss=0.1729, simple_loss=0.2513, pruned_loss=0.04725, over 19383.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2843, pruned_loss=0.06103, over 3820068.71 frames. ], batch size: 47, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:11:12,312 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174501.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:11:42,948 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 07:11:55,976 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 07:12:14,059 INFO [train.py:903] (3/4) Epoch 26, batch 3850, loss[loss=0.2102, simple_loss=0.2862, pruned_loss=0.06716, over 19739.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2862, pruned_loss=0.06197, over 3819211.81 frames. ], batch size: 45, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:12:27,530 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.991e+02 5.118e+02 6.382e+02 8.415e+02 1.605e+03, threshold=1.276e+03, percent-clipped=5.0 +2023-04-03 07:13:15,048 INFO [train.py:903] (3/4) Epoch 26, batch 3900, loss[loss=0.1772, simple_loss=0.2555, pruned_loss=0.04949, over 19611.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2861, pruned_loss=0.06213, over 3827688.48 frames. ], batch size: 50, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:13:36,295 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174616.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:14:13,844 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174645.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:14:20,757 INFO [train.py:903] (3/4) Epoch 26, batch 3950, loss[loss=0.2076, simple_loss=0.2859, pruned_loss=0.06469, over 19587.00 frames. ], tot_loss[loss=0.2058, simple_loss=0.2869, pruned_loss=0.06238, over 3823440.78 frames. ], batch size: 52, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:14:24,354 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 07:14:24,512 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174653.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:14:34,032 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.720e+02 4.927e+02 6.085e+02 7.650e+02 1.385e+03, threshold=1.217e+03, percent-clipped=3.0 +2023-04-03 07:14:54,713 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.42 vs. limit=2.0 +2023-04-03 07:15:24,376 INFO [train.py:903] (3/4) Epoch 26, batch 4000, loss[loss=0.1781, simple_loss=0.2521, pruned_loss=0.05208, over 19757.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2862, pruned_loss=0.06195, over 3816546.88 frames. ], batch size: 46, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:15:54,776 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=174723.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:13,021 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 07:16:15,486 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1329, 1.8083, 1.8751, 1.6171, 4.7022, 1.2715, 2.6753, 4.9787], + device='cuda:3'), covar=tensor([0.0421, 0.2690, 0.2718, 0.2068, 0.0666, 0.2626, 0.1407, 0.0160], + device='cuda:3'), in_proj_covar=tensor([0.0418, 0.0372, 0.0390, 0.0349, 0.0377, 0.0353, 0.0389, 0.0409], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:16:23,701 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0375, 1.7685, 1.9432, 2.7817, 1.9160, 2.3799, 2.5687, 1.9859], + device='cuda:3'), covar=tensor([0.0851, 0.0959, 0.0978, 0.0768, 0.0902, 0.0714, 0.0815, 0.0706], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0223, 0.0227, 0.0240, 0.0224, 0.0212, 0.0187, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 07:16:24,951 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174748.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:26,990 INFO [train.py:903] (3/4) Epoch 26, batch 4050, loss[loss=0.1947, simple_loss=0.2772, pruned_loss=0.05614, over 19840.00 frames. ], tot_loss[loss=0.2048, simple_loss=0.2859, pruned_loss=0.06186, over 3829508.13 frames. ], batch size: 52, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:16:27,193 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=174750.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:27,483 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174750.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:40,317 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174760.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:41,107 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.277e+02 4.947e+02 6.251e+02 7.600e+02 1.203e+03, threshold=1.250e+03, percent-clipped=0.0 +2023-04-03 07:16:52,077 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174768.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:16:59,323 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174773.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:17:01,697 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174775.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:17:32,067 INFO [train.py:903] (3/4) Epoch 26, batch 4100, loss[loss=0.2342, simple_loss=0.3251, pruned_loss=0.07169, over 17615.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2846, pruned_loss=0.06134, over 3833853.60 frames. ], batch size: 101, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:17:50,339 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.29 vs. limit=5.0 +2023-04-03 07:18:07,596 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 07:18:38,176 INFO [train.py:903] (3/4) Epoch 26, batch 4150, loss[loss=0.1824, simple_loss=0.2663, pruned_loss=0.04928, over 19590.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2834, pruned_loss=0.06056, over 3834946.00 frames. ], batch size: 52, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:18:53,313 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.134e+02 5.046e+02 6.484e+02 8.542e+02 2.236e+03, threshold=1.297e+03, percent-clipped=8.0 +2023-04-03 07:18:57,193 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=174865.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:19:05,609 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=174872.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:19:26,294 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 07:19:38,551 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=174897.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:19:41,592 INFO [train.py:903] (3/4) Epoch 26, batch 4200, loss[loss=0.2063, simple_loss=0.2916, pruned_loss=0.0605, over 19594.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2851, pruned_loss=0.0614, over 3823866.30 frames. ], batch size: 61, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:19:42,860 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 07:19:46,872 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1348, 1.3736, 1.7928, 1.1933, 2.5617, 3.4103, 3.0772, 3.6082], + device='cuda:3'), covar=tensor([0.1556, 0.3745, 0.3207, 0.2481, 0.0550, 0.0191, 0.0236, 0.0301], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0329, 0.0362, 0.0271, 0.0252, 0.0194, 0.0219, 0.0271], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 07:20:44,732 INFO [train.py:903] (3/4) Epoch 26, batch 4250, loss[loss=0.1884, simple_loss=0.2675, pruned_loss=0.05463, over 19729.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2849, pruned_loss=0.06132, over 3822227.81 frames. ], batch size: 47, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:20:55,198 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 07:21:01,853 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.522e+02 4.804e+02 5.687e+02 7.207e+02 1.586e+03, threshold=1.137e+03, percent-clipped=5.0 +2023-04-03 07:21:07,833 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 07:21:10,193 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0408, 1.8101, 1.9587, 2.6533, 1.7000, 2.3108, 2.2608, 2.0757], + device='cuda:3'), covar=tensor([0.0827, 0.0930, 0.0933, 0.0797, 0.0970, 0.0774, 0.0917, 0.0683], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0227, 0.0240, 0.0225, 0.0212, 0.0188, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 07:21:49,466 INFO [train.py:903] (3/4) Epoch 26, batch 4300, loss[loss=0.1759, simple_loss=0.2643, pruned_loss=0.04376, over 19816.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2848, pruned_loss=0.06115, over 3816936.69 frames. ], batch size: 49, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:22:11,506 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175016.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:22:19,610 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.7065, 5.2693, 3.2294, 4.5488, 1.2232, 5.2890, 5.1262, 5.2324], + device='cuda:3'), covar=tensor([0.0383, 0.0754, 0.1666, 0.0653, 0.3850, 0.0544, 0.0779, 0.1039], + device='cuda:3'), in_proj_covar=tensor([0.0523, 0.0423, 0.0508, 0.0355, 0.0409, 0.0448, 0.0445, 0.0472], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:22:20,977 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175024.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:22:38,871 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 07:22:41,557 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175041.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:22:52,853 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175049.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:22:53,613 INFO [train.py:903] (3/4) Epoch 26, batch 4350, loss[loss=0.1916, simple_loss=0.2789, pruned_loss=0.05217, over 19848.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.284, pruned_loss=0.06075, over 3830453.35 frames. ], batch size: 52, lr: 3.15e-03, grad_scale: 4.0 +2023-04-03 07:23:08,531 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.209e+02 4.980e+02 6.586e+02 7.974e+02 2.340e+03, threshold=1.317e+03, percent-clipped=8.0 +2023-04-03 07:23:14,403 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175067.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:23:56,993 INFO [train.py:903] (3/4) Epoch 26, batch 4400, loss[loss=0.1891, simple_loss=0.2725, pruned_loss=0.05285, over 19837.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2837, pruned_loss=0.06063, over 3825200.70 frames. ], batch size: 52, lr: 3.15e-03, grad_scale: 8.0 +2023-04-03 07:24:17,272 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 07:24:23,150 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175121.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:24:26,278 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 07:24:55,280 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175146.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:24:59,528 INFO [train.py:903] (3/4) Epoch 26, batch 4450, loss[loss=0.25, simple_loss=0.3216, pruned_loss=0.08918, over 17493.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2845, pruned_loss=0.06134, over 3823349.14 frames. ], batch size: 101, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:25:08,212 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175157.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:25:14,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.113e+02 4.830e+02 6.413e+02 8.619e+02 2.132e+03, threshold=1.283e+03, percent-clipped=8.0 +2023-04-03 07:25:41,757 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175182.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:26:02,757 INFO [train.py:903] (3/4) Epoch 26, batch 4500, loss[loss=0.1906, simple_loss=0.276, pruned_loss=0.0526, over 19608.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2842, pruned_loss=0.06146, over 3815517.33 frames. ], batch size: 61, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:26:35,234 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-03 07:26:36,029 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175225.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:27:08,200 INFO [train.py:903] (3/4) Epoch 26, batch 4550, loss[loss=0.2566, simple_loss=0.3412, pruned_loss=0.08603, over 19527.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2838, pruned_loss=0.06127, over 3805603.11 frames. ], batch size: 56, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:27:15,053 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 07:27:23,523 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.068e+02 4.834e+02 6.204e+02 7.660e+02 2.009e+03, threshold=1.241e+03, percent-clipped=3.0 +2023-04-03 07:27:40,264 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 07:28:13,289 INFO [train.py:903] (3/4) Epoch 26, batch 4600, loss[loss=0.1926, simple_loss=0.2761, pruned_loss=0.05459, over 19669.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.284, pruned_loss=0.06123, over 3812266.95 frames. ], batch size: 55, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:29:16,052 INFO [train.py:903] (3/4) Epoch 26, batch 4650, loss[loss=0.2233, simple_loss=0.305, pruned_loss=0.07077, over 17478.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2842, pruned_loss=0.0611, over 3815303.96 frames. ], batch size: 101, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:29:29,905 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.676e+02 4.879e+02 5.869e+02 7.462e+02 1.692e+03, threshold=1.174e+03, percent-clipped=3.0 +2023-04-03 07:29:31,125 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 07:29:44,497 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 07:30:03,985 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175388.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:30:17,582 INFO [train.py:903] (3/4) Epoch 26, batch 4700, loss[loss=0.2525, simple_loss=0.3263, pruned_loss=0.08939, over 17370.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2852, pruned_loss=0.06196, over 3813629.94 frames. ], batch size: 101, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:30:19,595 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2051, 2.2413, 2.4425, 2.9868, 2.2910, 2.8706, 2.4412, 2.2152], + device='cuda:3'), covar=tensor([0.4319, 0.4398, 0.2032, 0.2718, 0.4701, 0.2354, 0.5149, 0.3597], + device='cuda:3'), in_proj_covar=tensor([0.0928, 0.1004, 0.0737, 0.0951, 0.0908, 0.0843, 0.0858, 0.0803], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 07:30:34,462 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8909, 4.3133, 4.6202, 4.6147, 1.7493, 4.3398, 3.7267, 4.3352], + device='cuda:3'), covar=tensor([0.1790, 0.0949, 0.0681, 0.0759, 0.6114, 0.1032, 0.0732, 0.1224], + device='cuda:3'), in_proj_covar=tensor([0.0812, 0.0775, 0.0977, 0.0862, 0.0852, 0.0744, 0.0582, 0.0907], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 07:30:35,691 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175412.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:30:40,993 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 07:31:07,019 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175438.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:31:21,517 INFO [train.py:903] (3/4) Epoch 26, batch 4750, loss[loss=0.1977, simple_loss=0.2827, pruned_loss=0.05637, over 19662.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2856, pruned_loss=0.0621, over 3806888.88 frames. ], batch size: 58, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:31:37,189 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.357e+02 4.945e+02 6.030e+02 7.655e+02 2.128e+03, threshold=1.206e+03, percent-clipped=6.0 +2023-04-03 07:31:38,786 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175463.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:32:22,862 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175499.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:32:23,788 INFO [train.py:903] (3/4) Epoch 26, batch 4800, loss[loss=0.2156, simple_loss=0.291, pruned_loss=0.07011, over 13085.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2847, pruned_loss=0.06165, over 3813262.01 frames. ], batch size: 137, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:32:25,152 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175501.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:33:28,514 INFO [train.py:903] (3/4) Epoch 26, batch 4850, loss[loss=0.1801, simple_loss=0.2729, pruned_loss=0.0436, over 19463.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06147, over 3809340.78 frames. ], batch size: 64, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:33:42,550 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.071e+02 4.553e+02 5.462e+02 6.461e+02 1.537e+03, threshold=1.092e+03, percent-clipped=2.0 +2023-04-03 07:33:49,690 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 07:33:51,024 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175569.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:34:12,021 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 07:34:18,041 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 07:34:18,069 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 07:34:27,160 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 07:34:30,589 INFO [train.py:903] (3/4) Epoch 26, batch 4900, loss[loss=0.2575, simple_loss=0.331, pruned_loss=0.092, over 19610.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2855, pruned_loss=0.06172, over 3809819.27 frames. ], batch size: 57, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:34:46,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 07:34:50,785 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175616.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:35:31,952 INFO [train.py:903] (3/4) Epoch 26, batch 4950, loss[loss=0.203, simple_loss=0.2883, pruned_loss=0.05883, over 19538.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2846, pruned_loss=0.06114, over 3823355.91 frames. ], batch size: 56, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:35:49,939 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.056e+02 4.878e+02 6.108e+02 7.489e+02 1.803e+03, threshold=1.222e+03, percent-clipped=10.0 +2023-04-03 07:35:49,976 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 07:36:13,140 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 07:36:15,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175684.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:36:36,321 INFO [train.py:903] (3/4) Epoch 26, batch 5000, loss[loss=0.2048, simple_loss=0.2783, pruned_loss=0.0657, over 19603.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2836, pruned_loss=0.0607, over 3822077.43 frames. ], batch size: 50, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:36:46,069 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 07:36:56,350 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 07:37:10,909 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175728.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:37:16,384 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175732.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:37:39,578 INFO [train.py:903] (3/4) Epoch 26, batch 5050, loss[loss=0.2132, simple_loss=0.2949, pruned_loss=0.06572, over 19770.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2834, pruned_loss=0.06005, over 3826244.15 frames. ], batch size: 54, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:37:41,475 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.67 vs. limit=5.0 +2023-04-03 07:37:46,846 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175756.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:37:49,654 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.64 vs. limit=5.0 +2023-04-03 07:37:53,680 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.913e+02 4.916e+02 5.717e+02 6.994e+02 1.273e+03, threshold=1.143e+03, percent-clipped=1.0 +2023-04-03 07:38:14,302 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 07:38:39,106 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0312, 3.7228, 2.4954, 3.3576, 0.8804, 3.6748, 3.5316, 3.6029], + device='cuda:3'), covar=tensor([0.0748, 0.1027, 0.1967, 0.0924, 0.3891, 0.0767, 0.1006, 0.1171], + device='cuda:3'), in_proj_covar=tensor([0.0523, 0.0423, 0.0507, 0.0358, 0.0409, 0.0449, 0.0445, 0.0473], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:38:42,427 INFO [train.py:903] (3/4) Epoch 26, batch 5100, loss[loss=0.2563, simple_loss=0.3294, pruned_loss=0.09154, over 19593.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2852, pruned_loss=0.06144, over 3824926.52 frames. ], batch size: 52, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:38:49,479 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 07:38:49,800 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4011, 2.1735, 1.6484, 1.4977, 2.0340, 1.3560, 1.2695, 1.8709], + device='cuda:3'), covar=tensor([0.1030, 0.0838, 0.1016, 0.0819, 0.0497, 0.1246, 0.0765, 0.0473], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0319, 0.0337, 0.0271, 0.0250, 0.0344, 0.0291, 0.0273], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:38:52,870 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 07:38:58,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 07:39:30,272 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2257, 1.2640, 1.2771, 1.1080, 1.0610, 1.1429, 0.1226, 0.3828], + device='cuda:3'), covar=tensor([0.0793, 0.0743, 0.0481, 0.0638, 0.1420, 0.0726, 0.1507, 0.1245], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0361, 0.0366, 0.0389, 0.0469, 0.0396, 0.0344, 0.0346], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 07:39:37,122 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=175843.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:39:42,013 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175847.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:39:45,161 INFO [train.py:903] (3/4) Epoch 26, batch 5150, loss[loss=0.1893, simple_loss=0.2607, pruned_loss=0.05901, over 19369.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06144, over 3820486.10 frames. ], batch size: 47, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:39:55,538 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 07:40:01,368 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 5.003e+02 6.332e+02 8.398e+02 1.509e+03, threshold=1.266e+03, percent-clipped=8.0 +2023-04-03 07:40:14,025 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175871.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:40:15,284 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175872.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:40:30,092 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 07:40:47,123 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175897.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:40:49,344 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175899.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:40:50,194 INFO [train.py:903] (3/4) Epoch 26, batch 5200, loss[loss=0.178, simple_loss=0.2605, pruned_loss=0.04772, over 19426.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2843, pruned_loss=0.06118, over 3815005.51 frames. ], batch size: 48, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:40:54,086 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5681, 1.1631, 1.4280, 1.3488, 2.2019, 1.1318, 2.0787, 2.5440], + device='cuda:3'), covar=tensor([0.0678, 0.2842, 0.2796, 0.1614, 0.0876, 0.1960, 0.1049, 0.0430], + device='cuda:3'), in_proj_covar=tensor([0.0417, 0.0371, 0.0389, 0.0349, 0.0376, 0.0353, 0.0388, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:41:02,247 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 07:41:41,511 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=175940.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:41:45,632 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 07:41:53,689 INFO [train.py:903] (3/4) Epoch 26, batch 5250, loss[loss=0.1986, simple_loss=0.2829, pruned_loss=0.05718, over 19444.00 frames. ], tot_loss[loss=0.202, simple_loss=0.283, pruned_loss=0.06049, over 3828943.64 frames. ], batch size: 70, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:42:03,288 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=175958.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:42:07,499 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.944e+02 4.675e+02 5.849e+02 7.641e+02 1.436e+03, threshold=1.170e+03, percent-clipped=2.0 +2023-04-03 07:42:09,026 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=175963.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:42:11,349 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=175965.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:42:25,590 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.00 vs. limit=5.0 +2023-04-03 07:42:54,939 INFO [train.py:903] (3/4) Epoch 26, batch 5300, loss[loss=0.2033, simple_loss=0.284, pruned_loss=0.06133, over 19574.00 frames. ], tot_loss[loss=0.203, simple_loss=0.284, pruned_loss=0.06097, over 3831317.87 frames. ], batch size: 52, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:43:08,838 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 07:43:09,634 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-03 07:43:57,534 INFO [train.py:903] (3/4) Epoch 26, batch 5350, loss[loss=0.2346, simple_loss=0.3158, pruned_loss=0.07667, over 19469.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2841, pruned_loss=0.06112, over 3830762.80 frames. ], batch size: 64, lr: 3.14e-03, grad_scale: 8.0 +2023-04-03 07:44:14,899 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.150e+02 4.809e+02 5.990e+02 7.353e+02 1.688e+03, threshold=1.198e+03, percent-clipped=9.0 +2023-04-03 07:44:27,825 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176072.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:44:30,115 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 07:45:03,474 INFO [train.py:903] (3/4) Epoch 26, batch 5400, loss[loss=0.2189, simple_loss=0.3137, pruned_loss=0.06207, over 19524.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2856, pruned_loss=0.0616, over 3824293.82 frames. ], batch size: 54, lr: 3.14e-03, grad_scale: 4.0 +2023-04-03 07:45:07,661 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176103.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:45:36,780 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176127.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:45:37,872 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176128.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:46:08,012 INFO [train.py:903] (3/4) Epoch 26, batch 5450, loss[loss=0.1778, simple_loss=0.2528, pruned_loss=0.05144, over 19735.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2856, pruned_loss=0.06193, over 3812626.32 frames. ], batch size: 46, lr: 3.14e-03, grad_scale: 4.0 +2023-04-03 07:46:10,696 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176152.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:46:23,248 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.815e+02 4.882e+02 5.855e+02 6.912e+02 1.680e+03, threshold=1.171e+03, percent-clipped=1.0 +2023-04-03 07:46:55,139 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176187.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:47:11,317 INFO [train.py:903] (3/4) Epoch 26, batch 5500, loss[loss=0.207, simple_loss=0.2967, pruned_loss=0.0587, over 19183.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2844, pruned_loss=0.06075, over 3819292.36 frames. ], batch size: 69, lr: 3.14e-03, grad_scale: 4.0 +2023-04-03 07:47:28,797 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176214.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:47:30,654 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 07:47:37,975 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9961, 1.6228, 1.6124, 1.8271, 1.5124, 1.6598, 1.5054, 1.8189], + device='cuda:3'), covar=tensor([0.1120, 0.1311, 0.1617, 0.1132, 0.1434, 0.0597, 0.1603, 0.0800], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0356, 0.0315, 0.0255, 0.0304, 0.0256, 0.0318, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 07:48:02,322 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176239.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:48:06,833 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176243.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:48:14,664 INFO [train.py:903] (3/4) Epoch 26, batch 5550, loss[loss=0.2739, simple_loss=0.3414, pruned_loss=0.1032, over 17293.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2839, pruned_loss=0.0607, over 3828556.57 frames. ], batch size: 101, lr: 3.14e-03, grad_scale: 4.0 +2023-04-03 07:48:17,144 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 07:48:29,013 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176260.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:48:32,235 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.168e+02 4.802e+02 5.930e+02 7.241e+02 1.738e+03, threshold=1.186e+03, percent-clipped=4.0 +2023-04-03 07:48:47,667 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176275.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:49:08,001 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 07:49:18,900 INFO [train.py:903] (3/4) Epoch 26, batch 5600, loss[loss=0.1885, simple_loss=0.2639, pruned_loss=0.0566, over 19091.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06028, over 3830537.23 frames. ], batch size: 42, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 07:49:28,230 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176307.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:49:45,712 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9295, 4.4083, 4.6909, 4.6751, 1.7200, 4.4124, 3.7973, 4.4036], + device='cuda:3'), covar=tensor([0.1691, 0.0864, 0.0620, 0.0714, 0.6167, 0.0932, 0.0755, 0.1075], + device='cuda:3'), in_proj_covar=tensor([0.0815, 0.0774, 0.0980, 0.0863, 0.0859, 0.0748, 0.0582, 0.0910], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 07:50:23,052 INFO [train.py:903] (3/4) Epoch 26, batch 5650, loss[loss=0.1512, simple_loss=0.2323, pruned_loss=0.03504, over 19716.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.284, pruned_loss=0.06109, over 3820193.12 frames. ], batch size: 46, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:50:26,196 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 07:50:32,662 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176358.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:50:39,207 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.976e+02 4.381e+02 5.619e+02 7.137e+02 2.187e+03, threshold=1.124e+03, percent-clipped=4.0 +2023-04-03 07:51:02,583 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 07:51:24,804 INFO [train.py:903] (3/4) Epoch 26, batch 5700, loss[loss=0.2021, simple_loss=0.2908, pruned_loss=0.0567, over 19781.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2848, pruned_loss=0.06123, over 3823124.38 frames. ], batch size: 56, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:51:52,472 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176422.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:52:19,184 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176443.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:52:22,356 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 07:52:26,812 INFO [train.py:903] (3/4) Epoch 26, batch 5750, loss[loss=0.1646, simple_loss=0.2469, pruned_loss=0.04112, over 19800.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2851, pruned_loss=0.06116, over 3828518.76 frames. ], batch size: 47, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:52:30,380 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 07:52:33,966 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 07:52:44,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.352e+02 5.112e+02 6.171e+02 7.862e+02 1.795e+03, threshold=1.234e+03, percent-clipped=7.0 +2023-04-03 07:52:50,854 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176468.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:53:09,514 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176484.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:53:28,523 INFO [train.py:903] (3/4) Epoch 26, batch 5800, loss[loss=0.1943, simple_loss=0.2867, pruned_loss=0.05097, over 19526.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2852, pruned_loss=0.06134, over 3812861.66 frames. ], batch size: 54, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:54:32,081 INFO [train.py:903] (3/4) Epoch 26, batch 5850, loss[loss=0.1653, simple_loss=0.2518, pruned_loss=0.03939, over 19477.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2853, pruned_loss=0.06153, over 3815284.37 frames. ], batch size: 49, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:54:48,246 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.299e+02 4.827e+02 6.114e+02 8.553e+02 2.097e+03, threshold=1.223e+03, percent-clipped=6.0 +2023-04-03 07:55:29,962 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 07:55:33,096 INFO [train.py:903] (3/4) Epoch 26, batch 5900, loss[loss=0.227, simple_loss=0.2998, pruned_loss=0.07715, over 19357.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2845, pruned_loss=0.06178, over 3821119.72 frames. ], batch size: 70, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:55:37,923 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176604.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:55:49,994 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176614.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:55:51,770 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 07:55:55,483 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176619.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:56:22,772 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176639.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:56:27,332 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176643.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:56:35,021 INFO [train.py:903] (3/4) Epoch 26, batch 5950, loss[loss=0.214, simple_loss=0.2909, pruned_loss=0.06859, over 19667.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2844, pruned_loss=0.06119, over 3834865.89 frames. ], batch size: 59, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 07:56:51,414 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.421e+02 5.079e+02 6.315e+02 7.489e+02 1.732e+03, threshold=1.263e+03, percent-clipped=4.0 +2023-04-03 07:57:12,971 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176678.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:57:18,414 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176683.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:57:38,458 INFO [train.py:903] (3/4) Epoch 26, batch 6000, loss[loss=0.1904, simple_loss=0.2658, pruned_loss=0.05755, over 19399.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.285, pruned_loss=0.0613, over 3825038.73 frames. ], batch size: 48, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 07:57:38,458 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 07:57:51,359 INFO [train.py:937] (3/4) Epoch 26, validation: loss=0.1675, simple_loss=0.2672, pruned_loss=0.03393, over 944034.00 frames. +2023-04-03 07:57:51,360 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 07:57:55,565 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=176703.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:58:16,481 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176719.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:58:36,024 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176734.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 07:58:44,226 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=5.42 vs. limit=5.0 +2023-04-03 07:58:54,251 INFO [train.py:903] (3/4) Epoch 26, batch 6050, loss[loss=0.2556, simple_loss=0.3412, pruned_loss=0.08496, over 19338.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2858, pruned_loss=0.06184, over 3811903.97 frames. ], batch size: 70, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 07:59:04,018 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4470, 3.1840, 2.4363, 2.5266, 2.2394, 2.7567, 1.2489, 2.2335], + device='cuda:3'), covar=tensor([0.0675, 0.0598, 0.0788, 0.1198, 0.1151, 0.1069, 0.1560, 0.1108], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0358, 0.0363, 0.0386, 0.0463, 0.0391, 0.0341, 0.0341], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 07:59:11,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.381e+02 5.094e+02 6.239e+02 7.368e+02 1.384e+03, threshold=1.248e+03, percent-clipped=1.0 +2023-04-03 07:59:57,921 INFO [train.py:903] (3/4) Epoch 26, batch 6100, loss[loss=0.2013, simple_loss=0.2676, pruned_loss=0.06747, over 16440.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.2854, pruned_loss=0.06161, over 3814471.51 frames. ], batch size: 36, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:00:33,086 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176828.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:01:00,302 INFO [train.py:903] (3/4) Epoch 26, batch 6150, loss[loss=0.1632, simple_loss=0.238, pruned_loss=0.04419, over 19732.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2839, pruned_loss=0.0608, over 3833035.68 frames. ], batch size: 46, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 08:01:18,088 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.177e+02 4.899e+02 5.851e+02 7.446e+02 2.190e+03, threshold=1.170e+03, percent-clipped=4.0 +2023-04-03 08:01:21,469 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 08:01:25,360 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=176871.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:02:01,466 INFO [train.py:903] (3/4) Epoch 26, batch 6200, loss[loss=0.2479, simple_loss=0.3287, pruned_loss=0.08355, over 19543.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2847, pruned_loss=0.06143, over 3823449.28 frames. ], batch size: 56, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 08:02:54,413 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=176943.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:03:02,385 INFO [train.py:903] (3/4) Epoch 26, batch 6250, loss[loss=0.216, simple_loss=0.3048, pruned_loss=0.06357, over 19681.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2844, pruned_loss=0.06094, over 3832626.26 frames. ], batch size: 59, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 08:03:20,781 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.905e+02 4.899e+02 6.025e+02 7.517e+02 2.005e+03, threshold=1.205e+03, percent-clipped=5.0 +2023-04-03 08:03:24,766 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1622, 3.1583, 1.9154, 1.9123, 2.8685, 1.6918, 1.5974, 2.2980], + device='cuda:3'), covar=tensor([0.1336, 0.0723, 0.1074, 0.0920, 0.0552, 0.1327, 0.1010, 0.0667], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0319, 0.0337, 0.0272, 0.0249, 0.0346, 0.0291, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:03:29,813 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 08:03:33,703 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176975.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:03:37,034 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9556, 1.5834, 1.7074, 1.5321, 3.5427, 1.1339, 2.4270, 4.0778], + device='cuda:3'), covar=tensor([0.0487, 0.2744, 0.2757, 0.2003, 0.0643, 0.2659, 0.1311, 0.0181], + device='cuda:3'), in_proj_covar=tensor([0.0419, 0.0374, 0.0391, 0.0351, 0.0378, 0.0356, 0.0391, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:03:47,548 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=176987.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:03:51,232 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=176990.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:04:03,343 INFO [train.py:903] (3/4) Epoch 26, batch 6300, loss[loss=0.2135, simple_loss=0.2796, pruned_loss=0.07369, over 19411.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2834, pruned_loss=0.06081, over 3823845.19 frames. ], batch size: 48, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 08:04:04,701 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177000.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:04:23,698 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177015.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:04:37,548 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=177027.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:04:56,729 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.43 vs. limit=2.0 +2023-04-03 08:05:06,334 INFO [train.py:903] (3/4) Epoch 26, batch 6350, loss[loss=0.2168, simple_loss=0.2807, pruned_loss=0.07644, over 19750.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2835, pruned_loss=0.06106, over 3827308.21 frames. ], batch size: 47, lr: 3.13e-03, grad_scale: 4.0 +2023-04-03 08:05:26,140 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.448e+02 5.061e+02 6.104e+02 7.524e+02 1.291e+03, threshold=1.221e+03, percent-clipped=2.0 +2023-04-03 08:05:27,891 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1898, 1.8133, 1.4438, 1.2646, 1.6326, 1.2317, 1.1358, 1.6529], + device='cuda:3'), covar=tensor([0.0866, 0.0799, 0.1086, 0.0841, 0.0578, 0.1309, 0.0710, 0.0433], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0317, 0.0334, 0.0270, 0.0248, 0.0343, 0.0290, 0.0272], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:06:11,963 INFO [train.py:903] (3/4) Epoch 26, batch 6400, loss[loss=0.1851, simple_loss=0.2688, pruned_loss=0.05072, over 19598.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2835, pruned_loss=0.06089, over 3822599.89 frames. ], batch size: 57, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:06:14,743 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=177102.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:06:38,229 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.74 vs. limit=2.0 +2023-04-03 08:06:45,698 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177127.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:06:48,162 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177129.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:07:05,480 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=177142.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:07:14,960 INFO [train.py:903] (3/4) Epoch 26, batch 6450, loss[loss=0.1975, simple_loss=0.2844, pruned_loss=0.05523, over 18895.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2844, pruned_loss=0.06116, over 3826215.70 frames. ], batch size: 74, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:07:33,737 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.467e+02 4.651e+02 5.846e+02 7.696e+02 2.286e+03, threshold=1.169e+03, percent-clipped=3.0 +2023-04-03 08:07:56,651 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 08:08:15,588 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177199.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:08:16,334 INFO [train.py:903] (3/4) Epoch 26, batch 6500, loss[loss=0.2314, simple_loss=0.3078, pruned_loss=0.07752, over 17549.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2847, pruned_loss=0.06174, over 3823094.71 frames. ], batch size: 101, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:08:17,677 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 08:08:27,444 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6129, 1.7226, 1.9560, 1.9601, 1.4844, 1.9187, 1.9535, 1.8142], + device='cuda:3'), covar=tensor([0.4180, 0.3645, 0.1990, 0.2434, 0.3927, 0.2252, 0.5136, 0.3393], + device='cuda:3'), in_proj_covar=tensor([0.0926, 0.1002, 0.0735, 0.0948, 0.0904, 0.0841, 0.0855, 0.0801], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 08:08:38,043 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=177215.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:08:49,220 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177224.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:09:20,907 INFO [train.py:903] (3/4) Epoch 26, batch 6550, loss[loss=0.2133, simple_loss=0.2947, pruned_loss=0.06599, over 19493.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2851, pruned_loss=0.06208, over 3826821.04 frames. ], batch size: 64, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:09:39,910 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.007e+02 4.796e+02 6.270e+02 7.966e+02 1.683e+03, threshold=1.254e+03, percent-clipped=5.0 +2023-04-03 08:10:25,190 INFO [train.py:903] (3/4) Epoch 26, batch 6600, loss[loss=0.2372, simple_loss=0.3154, pruned_loss=0.07956, over 18126.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2846, pruned_loss=0.0618, over 3819668.02 frames. ], batch size: 83, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:10:34,194 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.17 vs. limit=5.0 +2023-04-03 08:11:02,469 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=177330.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:11:27,589 INFO [train.py:903] (3/4) Epoch 26, batch 6650, loss[loss=0.1782, simple_loss=0.2674, pruned_loss=0.0445, over 19611.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2842, pruned_loss=0.0613, over 3835246.41 frames. ], batch size: 57, lr: 3.13e-03, grad_scale: 8.0 +2023-04-03 08:11:37,315 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177358.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:11:47,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.622e+02 4.339e+02 5.694e+02 7.782e+02 1.307e+03, threshold=1.139e+03, percent-clipped=1.0 +2023-04-03 08:12:11,033 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177383.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:12:28,608 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177398.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:12:30,480 INFO [train.py:903] (3/4) Epoch 26, batch 6700, loss[loss=0.1905, simple_loss=0.2789, pruned_loss=0.05106, over 19646.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2836, pruned_loss=0.06079, over 3825799.27 frames. ], batch size: 58, lr: 3.12e-03, grad_scale: 8.0 +2023-04-03 08:13:01,394 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177423.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:13:31,551 INFO [train.py:903] (3/4) Epoch 26, batch 6750, loss[loss=0.1924, simple_loss=0.2666, pruned_loss=0.05915, over 19734.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2837, pruned_loss=0.06066, over 3832718.14 frames. ], batch size: 47, lr: 3.12e-03, grad_scale: 8.0 +2023-04-03 08:13:48,517 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.769e+02 4.721e+02 5.880e+02 7.244e+02 1.873e+03, threshold=1.176e+03, percent-clipped=5.0 +2023-04-03 08:13:55,626 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=177471.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:13:57,777 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=177473.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:14:00,226 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1314, 1.0742, 1.1194, 1.2155, 0.9098, 1.2155, 1.1865, 1.1658], + device='cuda:3'), covar=tensor([0.0863, 0.0916, 0.0980, 0.0642, 0.0941, 0.0823, 0.0817, 0.0758], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0224, 0.0228, 0.0240, 0.0226, 0.0213, 0.0188, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 08:14:28,156 INFO [train.py:903] (3/4) Epoch 26, batch 6800, loss[loss=0.2302, simple_loss=0.3146, pruned_loss=0.07291, over 19297.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2845, pruned_loss=0.06121, over 3826247.89 frames. ], batch size: 66, lr: 3.12e-03, grad_scale: 8.0 +2023-04-03 08:14:46,325 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4705, 2.5483, 2.6730, 3.1475, 2.5988, 3.0386, 2.7020, 2.5270], + device='cuda:3'), covar=tensor([0.3608, 0.3270, 0.1610, 0.2079, 0.3344, 0.1737, 0.3837, 0.2753], + device='cuda:3'), in_proj_covar=tensor([0.0923, 0.1002, 0.0734, 0.0946, 0.0902, 0.0841, 0.0854, 0.0801], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 08:15:14,893 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 08:15:15,373 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 08:15:18,382 INFO [train.py:903] (3/4) Epoch 27, batch 0, loss[loss=0.2306, simple_loss=0.3079, pruned_loss=0.07663, over 19734.00 frames. ], tot_loss[loss=0.2306, simple_loss=0.3079, pruned_loss=0.07663, over 19734.00 frames. ], batch size: 51, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:15:18,382 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 08:15:30,263 INFO [train.py:937] (3/4) Epoch 27, validation: loss=0.1666, simple_loss=0.2668, pruned_loss=0.03317, over 944034.00 frames. +2023-04-03 08:15:30,264 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 08:15:42,939 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 08:16:15,168 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.146e+02 4.975e+02 6.244e+02 7.696e+02 2.158e+03, threshold=1.249e+03, percent-clipped=8.0 +2023-04-03 08:16:24,559 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177571.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:16:33,682 INFO [train.py:903] (3/4) Epoch 27, batch 50, loss[loss=0.1724, simple_loss=0.2539, pruned_loss=0.04548, over 19680.00 frames. ], tot_loss[loss=0.2064, simple_loss=0.2876, pruned_loss=0.06254, over 864131.17 frames. ], batch size: 53, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:16:44,159 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=177586.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:16:44,214 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177586.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:16:46,452 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=177588.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:17:06,226 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 08:17:15,812 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177611.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:17:35,978 INFO [train.py:903] (3/4) Epoch 27, batch 100, loss[loss=0.2421, simple_loss=0.3179, pruned_loss=0.08319, over 19471.00 frames. ], tot_loss[loss=0.2065, simple_loss=0.2879, pruned_loss=0.06251, over 1523907.71 frames. ], batch size: 64, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:17:47,459 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 08:17:53,818 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2601, 1.3004, 1.2667, 1.1112, 1.1396, 1.1284, 0.1000, 0.3748], + device='cuda:3'), covar=tensor([0.0750, 0.0710, 0.0488, 0.0694, 0.1379, 0.0740, 0.1458, 0.1244], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0360, 0.0363, 0.0387, 0.0467, 0.0392, 0.0342, 0.0344], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:18:23,249 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.041e+02 5.210e+02 6.703e+02 8.227e+02 2.617e+03, threshold=1.341e+03, percent-clipped=11.0 +2023-04-03 08:18:39,597 INFO [train.py:903] (3/4) Epoch 27, batch 150, loss[loss=0.2045, simple_loss=0.2932, pruned_loss=0.0579, over 19567.00 frames. ], tot_loss[loss=0.2061, simple_loss=0.2872, pruned_loss=0.06252, over 2044439.04 frames. ], batch size: 61, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:18:44,583 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177682.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:18:57,584 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-03 08:19:40,063 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 08:19:42,203 INFO [train.py:903] (3/4) Epoch 27, batch 200, loss[loss=0.1689, simple_loss=0.2584, pruned_loss=0.03973, over 19586.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2867, pruned_loss=0.06193, over 2452606.79 frames. ], batch size: 52, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:20:29,448 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.575e+02 4.383e+02 5.368e+02 7.234e+02 1.640e+03, threshold=1.074e+03, percent-clipped=1.0 +2023-04-03 08:20:41,538 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.80 vs. limit=5.0 +2023-04-03 08:20:46,587 INFO [train.py:903] (3/4) Epoch 27, batch 250, loss[loss=0.1823, simple_loss=0.2741, pruned_loss=0.04522, over 19634.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2851, pruned_loss=0.06114, over 2770387.37 frames. ], batch size: 50, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:21:09,866 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6175, 4.1237, 4.2911, 4.3156, 1.7236, 4.0273, 3.5611, 4.0409], + device='cuda:3'), covar=tensor([0.1752, 0.0862, 0.0697, 0.0739, 0.6162, 0.0994, 0.0675, 0.1191], + device='cuda:3'), in_proj_covar=tensor([0.0812, 0.0777, 0.0981, 0.0862, 0.0860, 0.0746, 0.0580, 0.0908], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 08:21:12,216 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177798.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:21:32,145 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=177815.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:21:50,878 INFO [train.py:903] (3/4) Epoch 27, batch 300, loss[loss=0.2035, simple_loss=0.2879, pruned_loss=0.05955, over 18363.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2838, pruned_loss=0.061, over 3005102.83 frames. ], batch size: 83, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:22:08,570 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177842.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:22:10,860 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=177844.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:22:36,373 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.100e+02 5.015e+02 6.251e+02 7.839e+02 1.329e+03, threshold=1.250e+03, percent-clipped=8.0 +2023-04-03 08:22:40,026 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177867.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:22:42,386 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=177869.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:22:52,702 INFO [train.py:903] (3/4) Epoch 27, batch 350, loss[loss=0.2051, simple_loss=0.2921, pruned_loss=0.05901, over 19498.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2854, pruned_loss=0.06172, over 3196004.07 frames. ], batch size: 64, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:23:00,617 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 08:23:00,929 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9060, 4.3858, 4.6855, 4.6710, 1.9125, 4.4121, 3.7844, 4.3833], + device='cuda:3'), covar=tensor([0.1780, 0.0787, 0.0583, 0.0735, 0.5904, 0.0873, 0.0709, 0.1189], + device='cuda:3'), in_proj_covar=tensor([0.0810, 0.0777, 0.0979, 0.0861, 0.0858, 0.0745, 0.0579, 0.0909], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 08:23:40,516 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=177915.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:23:56,872 INFO [train.py:903] (3/4) Epoch 27, batch 400, loss[loss=0.189, simple_loss=0.2781, pruned_loss=0.05002, over 19758.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2844, pruned_loss=0.06081, over 3340435.76 frames. ], batch size: 63, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:24:43,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.926e+02 4.795e+02 5.555e+02 6.674e+02 1.146e+03, threshold=1.111e+03, percent-clipped=0.0 +2023-04-03 08:24:57,928 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-04-03 08:24:58,340 INFO [train.py:903] (3/4) Epoch 27, batch 450, loss[loss=0.2942, simple_loss=0.3441, pruned_loss=0.1222, over 13765.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2851, pruned_loss=0.06184, over 3424239.76 frames. ], batch size: 135, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:25:39,717 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 08:25:40,954 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 08:25:59,729 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=178026.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:26:02,522 INFO [train.py:903] (3/4) Epoch 27, batch 500, loss[loss=0.2638, simple_loss=0.3301, pruned_loss=0.09873, over 19771.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2857, pruned_loss=0.06186, over 3527081.34 frames. ], batch size: 56, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:26:06,491 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=178030.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:26:06,599 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6933, 1.8422, 2.1585, 1.9827, 3.4965, 2.7709, 3.7656, 1.8063], + device='cuda:3'), covar=tensor([0.2460, 0.4287, 0.2723, 0.1849, 0.1351, 0.2055, 0.1382, 0.4214], + device='cuda:3'), in_proj_covar=tensor([0.0549, 0.0665, 0.0744, 0.0503, 0.0630, 0.0544, 0.0665, 0.0568], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:26:48,570 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.047e+02 5.151e+02 6.325e+02 8.134e+02 1.856e+03, threshold=1.265e+03, percent-clipped=5.0 +2023-04-03 08:26:51,474 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0734, 2.8461, 2.1971, 2.2115, 2.0155, 2.5048, 1.1317, 2.0724], + device='cuda:3'), covar=tensor([0.0794, 0.0686, 0.0725, 0.1194, 0.1203, 0.1187, 0.1488, 0.1073], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0363, 0.0367, 0.0390, 0.0472, 0.0395, 0.0346, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:27:07,128 INFO [train.py:903] (3/4) Epoch 27, batch 550, loss[loss=0.2008, simple_loss=0.2853, pruned_loss=0.05816, over 19681.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2835, pruned_loss=0.06066, over 3601795.42 frames. ], batch size: 58, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:27:15,659 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0849, 5.1082, 5.8728, 5.9385, 2.0734, 5.5519, 4.7271, 5.5379], + device='cuda:3'), covar=tensor([0.1830, 0.0845, 0.0601, 0.0661, 0.6328, 0.0819, 0.0650, 0.1224], + device='cuda:3'), in_proj_covar=tensor([0.0815, 0.0781, 0.0984, 0.0865, 0.0861, 0.0749, 0.0580, 0.0913], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 08:28:09,845 INFO [train.py:903] (3/4) Epoch 27, batch 600, loss[loss=0.1908, simple_loss=0.2723, pruned_loss=0.05462, over 19363.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2828, pruned_loss=0.0604, over 3651582.51 frames. ], batch size: 47, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:28:15,816 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9359, 4.3702, 4.6776, 4.6765, 1.9120, 4.3599, 3.8373, 4.4029], + device='cuda:3'), covar=tensor([0.1855, 0.1112, 0.0594, 0.0717, 0.6404, 0.1086, 0.0712, 0.1077], + device='cuda:3'), in_proj_covar=tensor([0.0818, 0.0781, 0.0987, 0.0868, 0.0864, 0.0751, 0.0583, 0.0916], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 08:28:25,333 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=178141.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:28:26,330 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=178142.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:28:48,732 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=178159.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:28:54,525 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 08:28:55,502 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.215e+02 5.259e+02 6.338e+02 7.640e+02 1.730e+03, threshold=1.268e+03, percent-clipped=4.0 +2023-04-03 08:29:11,522 INFO [train.py:903] (3/4) Epoch 27, batch 650, loss[loss=0.203, simple_loss=0.2813, pruned_loss=0.06231, over 19610.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2838, pruned_loss=0.06087, over 3700325.86 frames. ], batch size: 50, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:29:19,040 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1939, 2.0371, 1.8158, 2.0555, 1.7984, 1.8254, 1.7228, 2.0579], + device='cuda:3'), covar=tensor([0.1019, 0.1296, 0.1595, 0.1105, 0.1483, 0.0576, 0.1560, 0.0734], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0360, 0.0320, 0.0258, 0.0309, 0.0259, 0.0323, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 08:30:11,959 INFO [train.py:903] (3/4) Epoch 27, batch 700, loss[loss=0.1842, simple_loss=0.2776, pruned_loss=0.04541, over 19549.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2837, pruned_loss=0.06051, over 3737261.67 frames. ], batch size: 61, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:30:49,315 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=178257.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:30:58,538 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.944e+02 4.694e+02 6.359e+02 8.615e+02 1.569e+03, threshold=1.272e+03, percent-clipped=7.0 +2023-04-03 08:31:11,157 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=178274.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:31:16,249 INFO [train.py:903] (3/4) Epoch 27, batch 750, loss[loss=0.2219, simple_loss=0.3051, pruned_loss=0.06936, over 19789.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2834, pruned_loss=0.06067, over 3738568.27 frames. ], batch size: 56, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:31:26,140 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=178286.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:31:56,981 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=178311.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:32:17,797 INFO [train.py:903] (3/4) Epoch 27, batch 800, loss[loss=0.1833, simple_loss=0.2583, pruned_loss=0.05417, over 19717.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.283, pruned_loss=0.06068, over 3765990.24 frames. ], batch size: 46, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:32:31,916 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 08:33:04,022 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.236e+02 4.845e+02 6.131e+02 7.065e+02 2.334e+03, threshold=1.226e+03, percent-clipped=1.0 +2023-04-03 08:33:20,261 INFO [train.py:903] (3/4) Epoch 27, batch 850, loss[loss=0.1886, simple_loss=0.2791, pruned_loss=0.04909, over 19659.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2826, pruned_loss=0.06069, over 3774623.70 frames. ], batch size: 58, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:33:46,345 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=178397.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:34:12,683 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 08:34:16,564 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=178422.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:34:22,834 INFO [train.py:903] (3/4) Epoch 27, batch 900, loss[loss=0.2805, simple_loss=0.3408, pruned_loss=0.1101, over 18778.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2831, pruned_loss=0.06063, over 3794614.15 frames. ], batch size: 74, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:34:30,685 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.60 vs. limit=5.0 +2023-04-03 08:35:10,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.108e+02 4.416e+02 5.422e+02 6.593e+02 1.258e+03, threshold=1.084e+03, percent-clipped=1.0 +2023-04-03 08:35:17,045 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178470.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:35:28,161 INFO [train.py:903] (3/4) Epoch 27, batch 950, loss[loss=0.1839, simple_loss=0.2658, pruned_loss=0.05097, over 19750.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2821, pruned_loss=0.06, over 3805823.34 frames. ], batch size: 51, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:35:30,650 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 08:36:11,791 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=178513.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:36:32,191 INFO [train.py:903] (3/4) Epoch 27, batch 1000, loss[loss=0.213, simple_loss=0.2807, pruned_loss=0.07267, over 16048.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2827, pruned_loss=0.06037, over 3798216.68 frames. ], batch size: 35, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:36:34,991 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=178530.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:36:45,479 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=178538.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:37:06,264 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=178555.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:37:19,295 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.158e+02 4.971e+02 6.338e+02 8.822e+02 2.004e+03, threshold=1.268e+03, percent-clipped=12.0 +2023-04-03 08:37:25,264 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 08:37:35,828 INFO [train.py:903] (3/4) Epoch 27, batch 1050, loss[loss=0.1701, simple_loss=0.2557, pruned_loss=0.04222, over 19584.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2827, pruned_loss=0.06037, over 3818358.40 frames. ], batch size: 52, lr: 3.06e-03, grad_scale: 8.0 +2023-04-03 08:37:48,984 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2941, 1.5159, 1.9429, 1.5804, 2.8953, 4.5518, 4.4635, 5.0190], + device='cuda:3'), covar=tensor([0.1651, 0.3902, 0.3438, 0.2432, 0.0670, 0.0205, 0.0175, 0.0189], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0329, 0.0362, 0.0271, 0.0253, 0.0195, 0.0218, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 08:38:09,365 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 08:38:38,428 INFO [train.py:903] (3/4) Epoch 27, batch 1100, loss[loss=0.2191, simple_loss=0.299, pruned_loss=0.06958, over 18794.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2832, pruned_loss=0.06006, over 3832678.72 frames. ], batch size: 74, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:38:50,329 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178637.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:39:24,188 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6145, 1.6331, 1.8579, 1.7607, 2.5768, 2.3923, 2.8347, 1.2690], + device='cuda:3'), covar=tensor([0.2448, 0.4324, 0.2734, 0.1975, 0.1707, 0.2133, 0.1472, 0.4819], + device='cuda:3'), in_proj_covar=tensor([0.0552, 0.0667, 0.0745, 0.0504, 0.0633, 0.0544, 0.0665, 0.0568], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:39:27,056 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.060e+02 4.758e+02 5.833e+02 7.524e+02 1.653e+03, threshold=1.167e+03, percent-clipped=2.0 +2023-04-03 08:39:42,281 INFO [train.py:903] (3/4) Epoch 27, batch 1150, loss[loss=0.2014, simple_loss=0.2886, pruned_loss=0.05712, over 18754.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2826, pruned_loss=0.05955, over 3833740.96 frames. ], batch size: 74, lr: 3.05e-03, grad_scale: 4.0 +2023-04-03 08:40:25,352 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178712.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:40:30,194 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178715.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 08:40:47,635 INFO [train.py:903] (3/4) Epoch 27, batch 1200, loss[loss=0.1664, simple_loss=0.2417, pruned_loss=0.04553, over 19757.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2832, pruned_loss=0.0599, over 3825597.69 frames. ], batch size: 45, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:41:04,868 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3023, 2.1712, 2.1078, 1.9269, 1.7211, 1.8624, 0.6870, 1.3295], + device='cuda:3'), covar=tensor([0.0652, 0.0685, 0.0517, 0.0947, 0.1296, 0.1075, 0.1480, 0.1186], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0367, 0.0370, 0.0394, 0.0474, 0.0398, 0.0348, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:41:18,726 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 08:41:38,085 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.626e+02 4.854e+02 6.177e+02 8.249e+02 1.593e+03, threshold=1.235e+03, percent-clipped=5.0 +2023-04-03 08:41:53,228 INFO [train.py:903] (3/4) Epoch 27, batch 1250, loss[loss=0.1723, simple_loss=0.2481, pruned_loss=0.04827, over 19358.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2818, pruned_loss=0.0592, over 3831710.34 frames. ], batch size: 47, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:42:03,118 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4583, 2.4099, 2.2818, 2.5584, 2.3278, 2.0945, 2.1413, 2.4141], + device='cuda:3'), covar=tensor([0.0835, 0.1251, 0.1169, 0.0777, 0.1176, 0.0496, 0.1246, 0.0575], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0361, 0.0319, 0.0258, 0.0309, 0.0258, 0.0323, 0.0265], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 08:42:05,330 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178788.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:42:40,245 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=178814.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:42:47,795 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3464, 1.9959, 1.5015, 1.3962, 1.8263, 1.2482, 1.3666, 1.7703], + device='cuda:3'), covar=tensor([0.0990, 0.0901, 0.1147, 0.0885, 0.0594, 0.1359, 0.0698, 0.0471], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0321, 0.0337, 0.0273, 0.0250, 0.0346, 0.0294, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:42:56,590 INFO [train.py:903] (3/4) Epoch 27, batch 1300, loss[loss=0.2138, simple_loss=0.2833, pruned_loss=0.07212, over 19732.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.282, pruned_loss=0.0594, over 3825235.82 frames. ], batch size: 51, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:43:01,013 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 08:43:34,985 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=1.98 vs. limit=5.0 +2023-04-03 08:43:44,695 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.050e+02 5.004e+02 6.015e+02 7.494e+02 1.649e+03, threshold=1.203e+03, percent-clipped=2.0 +2023-04-03 08:43:58,891 INFO [train.py:903] (3/4) Epoch 27, batch 1350, loss[loss=0.1898, simple_loss=0.2801, pruned_loss=0.04978, over 19789.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2822, pruned_loss=0.05932, over 3837503.23 frames. ], batch size: 56, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:44:11,444 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2597, 1.8075, 1.4947, 1.1609, 1.6214, 1.1586, 1.2477, 1.7364], + device='cuda:3'), covar=tensor([0.0800, 0.0830, 0.1069, 0.0997, 0.0667, 0.1352, 0.0670, 0.0419], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0322, 0.0338, 0.0274, 0.0251, 0.0347, 0.0295, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:44:41,959 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=178912.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 08:45:03,473 INFO [train.py:903] (3/4) Epoch 27, batch 1400, loss[loss=0.2123, simple_loss=0.2946, pruned_loss=0.06498, over 19531.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2822, pruned_loss=0.0592, over 3838155.32 frames. ], batch size: 54, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:45:04,999 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=178929.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:45:50,939 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.415e+02 4.777e+02 6.195e+02 8.137e+02 1.607e+03, threshold=1.239e+03, percent-clipped=6.0 +2023-04-03 08:46:03,713 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 08:46:04,755 INFO [train.py:903] (3/4) Epoch 27, batch 1450, loss[loss=0.2224, simple_loss=0.2868, pruned_loss=0.07894, over 19740.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2832, pruned_loss=0.06, over 3835894.49 frames. ], batch size: 51, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:46:09,313 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=178981.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:46:20,296 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.74 vs. limit=2.0 +2023-04-03 08:47:07,115 INFO [train.py:903] (3/4) Epoch 27, batch 1500, loss[loss=0.2303, simple_loss=0.3022, pruned_loss=0.07925, over 13516.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2828, pruned_loss=0.06005, over 3843080.45 frames. ], batch size: 135, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:47:42,378 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179056.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:47:45,651 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179059.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 08:47:54,377 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.273e+02 4.749e+02 5.669e+02 7.445e+02 1.551e+03, threshold=1.134e+03, percent-clipped=3.0 +2023-04-03 08:48:08,105 INFO [train.py:903] (3/4) Epoch 27, batch 1550, loss[loss=0.1962, simple_loss=0.2779, pruned_loss=0.05723, over 19595.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2845, pruned_loss=0.06122, over 3827189.41 frames. ], batch size: 52, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:48:32,743 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179096.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:49:12,296 INFO [train.py:903] (3/4) Epoch 27, batch 1600, loss[loss=0.1647, simple_loss=0.2446, pruned_loss=0.04235, over 19415.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2829, pruned_loss=0.06008, over 3838834.23 frames. ], batch size: 48, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:49:18,229 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179132.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:49:37,308 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 08:49:59,054 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.135e+02 4.994e+02 6.186e+02 7.700e+02 1.707e+03, threshold=1.237e+03, percent-clipped=5.0 +2023-04-03 08:50:06,443 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179171.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:50:09,903 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179174.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 08:50:14,086 INFO [train.py:903] (3/4) Epoch 27, batch 1650, loss[loss=0.2058, simple_loss=0.2903, pruned_loss=0.06066, over 19524.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2815, pruned_loss=0.05949, over 3840785.64 frames. ], batch size: 54, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:50:23,752 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179185.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:50:55,268 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179210.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:51:17,033 INFO [train.py:903] (3/4) Epoch 27, batch 1700, loss[loss=0.1931, simple_loss=0.28, pruned_loss=0.05313, over 17458.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2812, pruned_loss=0.05939, over 3847026.52 frames. ], batch size: 101, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:51:41,270 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179247.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:51:52,273 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179256.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 08:51:58,254 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 08:52:04,052 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.943e+02 5.162e+02 6.407e+02 7.875e+02 2.392e+03, threshold=1.281e+03, percent-clipped=6.0 +2023-04-03 08:52:18,192 INFO [train.py:903] (3/4) Epoch 27, batch 1750, loss[loss=0.2022, simple_loss=0.2783, pruned_loss=0.06304, over 19755.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2828, pruned_loss=0.06042, over 3832476.77 frames. ], batch size: 51, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:53:13,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9820, 1.9435, 1.8898, 1.7796, 1.6267, 1.8078, 1.0124, 1.4309], + device='cuda:3'), covar=tensor([0.0534, 0.0642, 0.0447, 0.0709, 0.0990, 0.0838, 0.1302, 0.0884], + device='cuda:3'), in_proj_covar=tensor([0.0364, 0.0364, 0.0367, 0.0391, 0.0472, 0.0397, 0.0345, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 08:53:21,038 INFO [train.py:903] (3/4) Epoch 27, batch 1800, loss[loss=0.1676, simple_loss=0.2487, pruned_loss=0.04323, over 15944.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2831, pruned_loss=0.06081, over 3816333.86 frames. ], batch size: 35, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:53:21,459 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0042, 1.7193, 1.6212, 1.9053, 1.6343, 1.6599, 1.5149, 1.8561], + device='cuda:3'), covar=tensor([0.1081, 0.1436, 0.1531, 0.1054, 0.1393, 0.0610, 0.1616, 0.0762], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0359, 0.0318, 0.0256, 0.0307, 0.0257, 0.0320, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 08:53:51,557 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179352.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:54:08,233 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 4.924e+02 6.082e+02 7.444e+02 1.664e+03, threshold=1.216e+03, percent-clipped=4.0 +2023-04-03 08:54:16,177 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179371.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 08:54:21,557 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 08:54:23,167 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179377.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:54:23,933 INFO [train.py:903] (3/4) Epoch 27, batch 1850, loss[loss=0.2184, simple_loss=0.3024, pruned_loss=0.06721, over 19612.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2833, pruned_loss=0.06052, over 3830710.91 frames. ], batch size: 61, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:54:25,271 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179379.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:55:00,548 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 08:55:21,046 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9535, 2.0015, 1.8682, 2.9129, 2.1398, 2.7152, 1.9983, 1.6897], + device='cuda:3'), covar=tensor([0.4735, 0.4460, 0.2884, 0.3418, 0.4692, 0.2566, 0.6419, 0.4982], + device='cuda:3'), in_proj_covar=tensor([0.0927, 0.1004, 0.0737, 0.0948, 0.0905, 0.0844, 0.0855, 0.0802], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 08:55:25,821 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179427.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:55:26,557 INFO [train.py:903] (3/4) Epoch 27, batch 1900, loss[loss=0.1977, simple_loss=0.2841, pruned_loss=0.0556, over 19713.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2835, pruned_loss=0.0603, over 3836731.21 frames. ], batch size: 63, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:55:26,998 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1632, 1.4434, 1.8155, 1.0841, 2.4185, 3.0802, 2.7914, 3.2745], + device='cuda:3'), covar=tensor([0.1540, 0.3529, 0.3074, 0.2580, 0.0580, 0.0226, 0.0255, 0.0292], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0330, 0.0363, 0.0270, 0.0253, 0.0195, 0.0218, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 08:55:29,287 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179430.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 08:55:45,919 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 08:55:49,549 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 08:55:56,566 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179452.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:56:00,940 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179455.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 08:56:10,083 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179462.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 08:56:14,505 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.488e+02 5.214e+02 6.120e+02 7.486e+02 1.853e+03, threshold=1.224e+03, percent-clipped=1.0 +2023-04-03 08:56:15,775 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 08:56:28,553 INFO [train.py:903] (3/4) Epoch 27, batch 1950, loss[loss=0.1587, simple_loss=0.245, pruned_loss=0.03617, over 19473.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.283, pruned_loss=0.06, over 3840168.00 frames. ], batch size: 49, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:57:02,060 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179503.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:57:13,666 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.50 vs. limit=2.0 +2023-04-03 08:57:31,829 INFO [train.py:903] (3/4) Epoch 27, batch 2000, loss[loss=0.2096, simple_loss=0.2947, pruned_loss=0.06221, over 19681.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.283, pruned_loss=0.05984, over 3847533.86 frames. ], batch size: 60, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:57:32,247 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179528.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 08:58:19,141 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.349e+02 5.376e+02 6.611e+02 8.547e+02 2.231e+03, threshold=1.322e+03, percent-clipped=11.0 +2023-04-03 08:58:33,829 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 08:58:34,814 INFO [train.py:903] (3/4) Epoch 27, batch 2050, loss[loss=0.212, simple_loss=0.2934, pruned_loss=0.06535, over 19521.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.2842, pruned_loss=0.06062, over 3840942.26 frames. ], batch size: 56, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 08:58:53,466 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 08:58:54,673 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 08:59:14,222 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 08:59:37,263 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=179627.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 08:59:37,893 INFO [train.py:903] (3/4) Epoch 27, batch 2100, loss[loss=0.2421, simple_loss=0.3228, pruned_loss=0.0807, over 17147.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2838, pruned_loss=0.06012, over 3838532.15 frames. ], batch size: 101, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 09:00:03,485 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9181, 1.7066, 1.7470, 2.4482, 1.8783, 2.0929, 2.2082, 1.8933], + device='cuda:3'), covar=tensor([0.0741, 0.0906, 0.0958, 0.0651, 0.0829, 0.0832, 0.0865, 0.0704], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0224, 0.0227, 0.0239, 0.0226, 0.0213, 0.0188, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 09:00:07,552 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 09:00:07,917 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=179652.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 09:00:10,155 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179654.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:00:25,319 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.840e+02 4.609e+02 5.792e+02 6.874e+02 1.495e+03, threshold=1.158e+03, percent-clipped=2.0 +2023-04-03 09:00:28,911 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 09:00:39,305 INFO [train.py:903] (3/4) Epoch 27, batch 2150, loss[loss=0.206, simple_loss=0.2975, pruned_loss=0.05728, over 19617.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2836, pruned_loss=0.0606, over 3840684.77 frames. ], batch size: 67, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 09:00:42,472 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 09:01:37,657 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179723.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:01:43,430 INFO [train.py:903] (3/4) Epoch 27, batch 2200, loss[loss=0.217, simple_loss=0.299, pruned_loss=0.06748, over 19135.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2843, pruned_loss=0.06095, over 3841264.65 frames. ], batch size: 69, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 09:02:30,599 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.069e+02 5.235e+02 6.025e+02 7.394e+02 1.358e+03, threshold=1.205e+03, percent-clipped=1.0 +2023-04-03 09:02:46,342 INFO [train.py:903] (3/4) Epoch 27, batch 2250, loss[loss=0.2601, simple_loss=0.332, pruned_loss=0.0941, over 19665.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2838, pruned_loss=0.06066, over 3837932.57 frames. ], batch size: 58, lr: 3.05e-03, grad_scale: 8.0 +2023-04-03 09:03:21,828 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179806.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:03:33,123 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179814.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:03:50,157 INFO [train.py:903] (3/4) Epoch 27, batch 2300, loss[loss=0.192, simple_loss=0.2787, pruned_loss=0.05263, over 19699.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2849, pruned_loss=0.0611, over 3833340.73 frames. ], batch size: 59, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:04:02,183 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179838.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:04:03,018 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 09:04:38,194 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.712e+02 5.173e+02 6.408e+02 8.011e+02 2.024e+03, threshold=1.282e+03, percent-clipped=7.0 +2023-04-03 09:04:51,777 INFO [train.py:903] (3/4) Epoch 27, batch 2350, loss[loss=0.1798, simple_loss=0.2605, pruned_loss=0.04953, over 19626.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2846, pruned_loss=0.06142, over 3839121.28 frames. ], batch size: 50, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:05:34,683 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 09:05:45,187 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=179921.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:05:48,595 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179924.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:05:50,590 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 09:05:53,778 INFO [train.py:903] (3/4) Epoch 27, batch 2400, loss[loss=0.2229, simple_loss=0.3106, pruned_loss=0.06766, over 19669.00 frames. ], tot_loss[loss=0.2042, simple_loss=0.2853, pruned_loss=0.06156, over 3827338.62 frames. ], batch size: 58, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:06:11,791 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=179942.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:06:41,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.340e+02 4.979e+02 6.417e+02 8.310e+02 2.182e+03, threshold=1.283e+03, percent-clipped=4.0 +2023-04-03 09:06:57,082 INFO [train.py:903] (3/4) Epoch 27, batch 2450, loss[loss=0.1798, simple_loss=0.2629, pruned_loss=0.04841, over 19864.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2858, pruned_loss=0.06211, over 3806890.51 frames. ], batch size: 52, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:07:21,329 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=179998.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:07:46,433 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-04-03 09:08:00,988 INFO [train.py:903] (3/4) Epoch 27, batch 2500, loss[loss=0.2365, simple_loss=0.3102, pruned_loss=0.08137, over 19596.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.2852, pruned_loss=0.06174, over 3791629.43 frames. ], batch size: 61, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:08:19,354 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180044.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:08:47,565 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0538, 1.3409, 1.6450, 0.9682, 2.3017, 3.0805, 2.7739, 3.3003], + device='cuda:3'), covar=tensor([0.1680, 0.3868, 0.3464, 0.2811, 0.0647, 0.0222, 0.0271, 0.0298], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0331, 0.0364, 0.0270, 0.0254, 0.0196, 0.0220, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 09:08:48,295 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.631e+02 4.714e+02 5.802e+02 6.784e+02 1.958e+03, threshold=1.160e+03, percent-clipped=2.0 +2023-04-03 09:09:02,320 INFO [train.py:903] (3/4) Epoch 27, batch 2550, loss[loss=0.1687, simple_loss=0.2442, pruned_loss=0.04657, over 19789.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.284, pruned_loss=0.06096, over 3814244.18 frames. ], batch size: 48, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:09:03,695 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7961, 4.2569, 4.4732, 4.4894, 1.7818, 4.2389, 3.6647, 4.1859], + device='cuda:3'), covar=tensor([0.1681, 0.1040, 0.0658, 0.0693, 0.5959, 0.1061, 0.0724, 0.1120], + device='cuda:3'), in_proj_covar=tensor([0.0817, 0.0781, 0.0989, 0.0867, 0.0865, 0.0750, 0.0582, 0.0915], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 09:09:22,052 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180094.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:09:47,093 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180113.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:09:53,124 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0934, 3.3460, 1.8637, 2.0246, 2.9951, 1.7322, 1.5450, 2.1742], + device='cuda:3'), covar=tensor([0.1428, 0.0708, 0.1211, 0.0937, 0.0626, 0.1342, 0.0998, 0.0753], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0321, 0.0338, 0.0273, 0.0250, 0.0344, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:09:54,432 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180119.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:09:58,713 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 09:10:04,541 INFO [train.py:903] (3/4) Epoch 27, batch 2600, loss[loss=0.2126, simple_loss=0.2977, pruned_loss=0.06378, over 19527.00 frames. ], tot_loss[loss=0.204, simple_loss=0.285, pruned_loss=0.06146, over 3797979.37 frames. ], batch size: 56, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:10:15,938 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0665, 1.8991, 1.7203, 2.0202, 1.7688, 1.7827, 1.6656, 1.9051], + device='cuda:3'), covar=tensor([0.0993, 0.1378, 0.1488, 0.1102, 0.1378, 0.0555, 0.1388, 0.0793], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0359, 0.0317, 0.0258, 0.0307, 0.0256, 0.0321, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:10:42,772 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180158.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:10:52,190 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.132e+02 5.258e+02 6.240e+02 7.647e+02 1.495e+03, threshold=1.248e+03, percent-clipped=4.0 +2023-04-03 09:11:07,747 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180177.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 09:11:08,382 INFO [train.py:903] (3/4) Epoch 27, batch 2650, loss[loss=0.1735, simple_loss=0.2502, pruned_loss=0.04842, over 19755.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2847, pruned_loss=0.06146, over 3794598.91 frames. ], batch size: 46, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:11:18,552 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6320, 4.2382, 2.6303, 3.6774, 0.8788, 4.2572, 4.0467, 4.1808], + device='cuda:3'), covar=tensor([0.0623, 0.0989, 0.2053, 0.0921, 0.4021, 0.0659, 0.0916, 0.1235], + device='cuda:3'), in_proj_covar=tensor([0.0528, 0.0427, 0.0511, 0.0358, 0.0407, 0.0455, 0.0447, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:11:29,719 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 09:11:38,057 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180202.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:11:59,670 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-03 09:12:04,724 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180223.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:12:11,766 INFO [train.py:903] (3/4) Epoch 27, batch 2700, loss[loss=0.226, simple_loss=0.3087, pruned_loss=0.07166, over 19757.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2852, pruned_loss=0.06141, over 3799561.13 frames. ], batch size: 63, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:12:33,808 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 09:12:39,112 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2878, 2.3755, 2.5492, 2.8913, 2.4046, 2.8029, 2.5547, 2.4333], + device='cuda:3'), covar=tensor([0.3569, 0.3293, 0.1658, 0.2076, 0.3433, 0.1868, 0.3811, 0.2608], + device='cuda:3'), in_proj_covar=tensor([0.0934, 0.1012, 0.0742, 0.0954, 0.0912, 0.0852, 0.0859, 0.0809], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 09:13:00,979 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.426e+02 5.233e+02 6.139e+02 8.955e+02 2.009e+03, threshold=1.228e+03, percent-clipped=9.0 +2023-04-03 09:13:02,196 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180268.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:13:08,003 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180273.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:13:11,344 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4786, 2.1709, 1.6092, 1.4855, 1.9547, 1.3551, 1.3714, 1.8784], + device='cuda:3'), covar=tensor([0.1077, 0.0888, 0.1162, 0.0894, 0.0628, 0.1343, 0.0745, 0.0472], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0323, 0.0341, 0.0275, 0.0253, 0.0347, 0.0295, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:13:13,279 INFO [train.py:903] (3/4) Epoch 27, batch 2750, loss[loss=0.2115, simple_loss=0.2961, pruned_loss=0.06347, over 17314.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2834, pruned_loss=0.06068, over 3805290.07 frames. ], batch size: 101, lr: 3.04e-03, grad_scale: 4.0 +2023-04-03 09:13:20,962 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.08 vs. limit=2.0 +2023-04-03 09:13:22,799 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180286.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:14:15,231 INFO [train.py:903] (3/4) Epoch 27, batch 2800, loss[loss=0.2106, simple_loss=0.2928, pruned_loss=0.06419, over 19759.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2825, pruned_loss=0.06023, over 3810329.67 frames. ], batch size: 63, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:14:49,797 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180354.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:15:04,816 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.309e+02 4.838e+02 6.554e+02 8.151e+02 1.805e+03, threshold=1.311e+03, percent-clipped=3.0 +2023-04-03 09:15:08,577 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180369.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:15:18,466 INFO [train.py:903] (3/4) Epoch 27, batch 2850, loss[loss=0.2016, simple_loss=0.2887, pruned_loss=0.05721, over 19476.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2827, pruned_loss=0.06052, over 3803495.47 frames. ], batch size: 49, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:15:24,610 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-04-03 09:15:26,280 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180383.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 09:15:33,053 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180388.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:15:40,147 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180394.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:15:48,634 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180401.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:16:20,689 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 09:16:22,936 INFO [train.py:903] (3/4) Epoch 27, batch 2900, loss[loss=0.1739, simple_loss=0.264, pruned_loss=0.04188, over 19664.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2827, pruned_loss=0.06003, over 3826322.30 frames. ], batch size: 55, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:17:13,071 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.869e+02 4.922e+02 6.355e+02 8.136e+02 1.738e+03, threshold=1.271e+03, percent-clipped=4.0 +2023-04-03 09:17:15,679 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3120, 3.8298, 3.9512, 3.9556, 1.6533, 3.7591, 3.2621, 3.7296], + device='cuda:3'), covar=tensor([0.1801, 0.1036, 0.0682, 0.0792, 0.5768, 0.1054, 0.0786, 0.1126], + device='cuda:3'), in_proj_covar=tensor([0.0813, 0.0777, 0.0987, 0.0865, 0.0860, 0.0749, 0.0580, 0.0915], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 09:17:26,070 INFO [train.py:903] (3/4) Epoch 27, batch 2950, loss[loss=0.2113, simple_loss=0.2911, pruned_loss=0.06569, over 19683.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2827, pruned_loss=0.05977, over 3835327.56 frames. ], batch size: 53, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:17:45,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8646, 1.9199, 2.2709, 1.9817, 2.8262, 3.2203, 3.1233, 3.3808], + device='cuda:3'), covar=tensor([0.1335, 0.3070, 0.2732, 0.2306, 0.1079, 0.0320, 0.0211, 0.0388], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0332, 0.0365, 0.0271, 0.0255, 0.0197, 0.0220, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 09:17:56,638 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180503.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:18:27,174 INFO [train.py:903] (3/4) Epoch 27, batch 3000, loss[loss=0.1861, simple_loss=0.2667, pruned_loss=0.0528, over 19751.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2836, pruned_loss=0.06001, over 3836321.14 frames. ], batch size: 51, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:18:27,174 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 09:18:39,750 INFO [train.py:937] (3/4) Epoch 27, validation: loss=0.1667, simple_loss=0.2664, pruned_loss=0.03355, over 944034.00 frames. +2023-04-03 09:18:39,751 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 09:18:41,291 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180529.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:18:43,443 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 09:19:12,395 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180554.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:19:29,346 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180567.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:19:30,287 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.939e+02 4.587e+02 5.827e+02 7.595e+02 1.750e+03, threshold=1.165e+03, percent-clipped=2.0 +2023-04-03 09:19:41,899 INFO [train.py:903] (3/4) Epoch 27, batch 3050, loss[loss=0.1906, simple_loss=0.2659, pruned_loss=0.0577, over 19764.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2835, pruned_loss=0.06044, over 3846895.35 frames. ], batch size: 48, lr: 3.04e-03, grad_scale: 4.0 +2023-04-03 09:19:54,664 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180588.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:20:44,771 INFO [train.py:903] (3/4) Epoch 27, batch 3100, loss[loss=0.219, simple_loss=0.3006, pruned_loss=0.06874, over 19661.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2824, pruned_loss=0.05975, over 3842595.69 frames. ], batch size: 58, lr: 3.04e-03, grad_scale: 4.0 +2023-04-03 09:20:56,826 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4097, 1.5275, 1.7138, 1.6670, 2.4669, 2.1480, 2.6040, 1.0472], + device='cuda:3'), covar=tensor([0.2639, 0.4545, 0.2964, 0.2013, 0.1603, 0.2306, 0.1509, 0.5007], + device='cuda:3'), in_proj_covar=tensor([0.0551, 0.0668, 0.0749, 0.0505, 0.0633, 0.0545, 0.0668, 0.0569], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 09:20:59,825 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180639.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 09:21:20,360 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180657.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:21:29,225 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180664.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 09:21:33,518 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 4.719e+02 5.890e+02 7.652e+02 1.677e+03, threshold=1.178e+03, percent-clipped=3.0 +2023-04-03 09:21:46,247 INFO [train.py:903] (3/4) Epoch 27, batch 3150, loss[loss=0.1942, simple_loss=0.2849, pruned_loss=0.05172, over 18126.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2825, pruned_loss=0.05954, over 3840324.30 frames. ], batch size: 84, lr: 3.04e-03, grad_scale: 4.0 +2023-04-03 09:21:52,951 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180682.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:21:52,990 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180682.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:22:11,325 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180698.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:22:15,854 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 09:22:49,869 INFO [train.py:903] (3/4) Epoch 27, batch 3200, loss[loss=0.2069, simple_loss=0.2857, pruned_loss=0.06403, over 19749.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.284, pruned_loss=0.06026, over 3833985.83 frames. ], batch size: 54, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:23:16,637 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180750.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:23:28,924 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180759.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:23:39,759 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.293e+02 5.255e+02 6.624e+02 9.042e+02 3.460e+03, threshold=1.325e+03, percent-clipped=12.0 +2023-04-03 09:23:51,461 INFO [train.py:903] (3/4) Epoch 27, batch 3250, loss[loss=0.1991, simple_loss=0.2808, pruned_loss=0.05865, over 19285.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2831, pruned_loss=0.06009, over 3836351.12 frames. ], batch size: 66, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:23:59,371 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180784.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:24:36,461 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=180813.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:24:45,537 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.8151, 5.3060, 3.0383, 4.5627, 1.3376, 5.3896, 5.2348, 5.3606], + device='cuda:3'), covar=tensor([0.0396, 0.0837, 0.1938, 0.0854, 0.3800, 0.0519, 0.0822, 0.1138], + device='cuda:3'), in_proj_covar=tensor([0.0527, 0.0425, 0.0511, 0.0357, 0.0405, 0.0452, 0.0447, 0.0475], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:24:54,213 INFO [train.py:903] (3/4) Epoch 27, batch 3300, loss[loss=0.2338, simple_loss=0.3131, pruned_loss=0.07725, over 19398.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2843, pruned_loss=0.06083, over 3821070.97 frames. ], batch size: 66, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:25:00,020 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 09:25:07,333 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-03 09:25:13,339 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=180843.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:25:43,899 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.436e+02 5.181e+02 6.566e+02 8.708e+02 1.878e+03, threshold=1.313e+03, percent-clipped=7.0 +2023-04-03 09:25:56,111 INFO [train.py:903] (3/4) Epoch 27, batch 3350, loss[loss=0.2389, simple_loss=0.3129, pruned_loss=0.08242, over 13521.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2842, pruned_loss=0.0607, over 3815239.40 frames. ], batch size: 136, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:26:51,579 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0106, 4.4060, 4.7234, 4.7890, 1.9215, 4.4409, 3.8727, 4.3940], + device='cuda:3'), covar=tensor([0.1750, 0.1048, 0.0682, 0.0666, 0.6023, 0.0979, 0.0726, 0.1389], + device='cuda:3'), in_proj_covar=tensor([0.0822, 0.0783, 0.0997, 0.0875, 0.0865, 0.0756, 0.0585, 0.0923], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 09:27:00,212 INFO [train.py:903] (3/4) Epoch 27, batch 3400, loss[loss=0.1771, simple_loss=0.27, pruned_loss=0.04205, over 19785.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2842, pruned_loss=0.06092, over 3828281.22 frames. ], batch size: 54, lr: 3.04e-03, grad_scale: 8.0 +2023-04-03 09:27:05,046 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=180932.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:27:11,994 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=180938.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:27:43,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=180963.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:27:50,457 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.175e+02 4.823e+02 5.841e+02 8.287e+02 1.627e+03, threshold=1.168e+03, percent-clipped=5.0 +2023-04-03 09:27:59,651 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-04-03 09:28:02,290 INFO [train.py:903] (3/4) Epoch 27, batch 3450, loss[loss=0.1926, simple_loss=0.2726, pruned_loss=0.05628, over 19617.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2846, pruned_loss=0.06122, over 3800007.88 frames. ], batch size: 50, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:28:06,935 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 09:28:18,944 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.60 vs. limit=5.0 +2023-04-03 09:29:04,452 INFO [train.py:903] (3/4) Epoch 27, batch 3500, loss[loss=0.2058, simple_loss=0.2935, pruned_loss=0.05909, over 17290.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.284, pruned_loss=0.06116, over 3806389.67 frames. ], batch size: 101, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:29:28,501 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=181047.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:29:53,846 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.012e+02 5.177e+02 6.114e+02 8.226e+02 1.424e+03, threshold=1.223e+03, percent-clipped=6.0 +2023-04-03 09:29:55,530 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=181069.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:30:06,527 INFO [train.py:903] (3/4) Epoch 27, batch 3550, loss[loss=0.2443, simple_loss=0.3262, pruned_loss=0.08118, over 19599.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2835, pruned_loss=0.06074, over 3826473.42 frames. ], batch size: 57, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:30:27,217 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=181094.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:30:27,439 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=181094.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:31:09,188 INFO [train.py:903] (3/4) Epoch 27, batch 3600, loss[loss=0.247, simple_loss=0.3376, pruned_loss=0.07819, over 19563.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2846, pruned_loss=0.06142, over 3826035.47 frames. ], batch size: 61, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:32:00,758 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.317e+02 4.881e+02 5.766e+02 7.270e+02 1.755e+03, threshold=1.153e+03, percent-clipped=5.0 +2023-04-03 09:32:12,503 INFO [train.py:903] (3/4) Epoch 27, batch 3650, loss[loss=0.2099, simple_loss=0.2959, pruned_loss=0.06195, over 19485.00 frames. ], tot_loss[loss=0.2049, simple_loss=0.2857, pruned_loss=0.06207, over 3821594.53 frames. ], batch size: 64, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:32:24,261 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=181187.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:32:33,466 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.4116, 4.0254, 2.8108, 3.6040, 1.0463, 4.0121, 3.8848, 3.9540], + device='cuda:3'), covar=tensor([0.0679, 0.1148, 0.1874, 0.0898, 0.3937, 0.0710, 0.0926, 0.1116], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0428, 0.0514, 0.0358, 0.0408, 0.0453, 0.0448, 0.0476], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:32:52,782 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=181209.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:33:15,871 INFO [train.py:903] (3/4) Epoch 27, batch 3700, loss[loss=0.1534, simple_loss=0.2443, pruned_loss=0.03122, over 19621.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2846, pruned_loss=0.06134, over 3818803.62 frames. ], batch size: 50, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:34:06,266 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.101e+02 5.357e+02 6.375e+02 8.389e+02 2.807e+03, threshold=1.275e+03, percent-clipped=10.0 +2023-04-03 09:34:17,475 INFO [train.py:903] (3/4) Epoch 27, batch 3750, loss[loss=0.1807, simple_loss=0.2648, pruned_loss=0.04837, over 19773.00 frames. ], tot_loss[loss=0.2052, simple_loss=0.2857, pruned_loss=0.06233, over 3817353.29 frames. ], batch size: 54, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:34:48,314 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=181302.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:34:49,639 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=181303.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:35:04,207 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=181315.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:35:20,116 INFO [train.py:903] (3/4) Epoch 27, batch 3800, loss[loss=0.1879, simple_loss=0.2589, pruned_loss=0.0585, over 19100.00 frames. ], tot_loss[loss=0.2053, simple_loss=0.2858, pruned_loss=0.06242, over 3799124.39 frames. ], batch size: 42, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:35:20,561 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=181328.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:35:48,534 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 09:36:10,636 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.827e+02 5.403e+02 6.773e+02 8.761e+02 1.536e+03, threshold=1.355e+03, percent-clipped=8.0 +2023-04-03 09:36:22,053 INFO [train.py:903] (3/4) Epoch 27, batch 3850, loss[loss=0.1954, simple_loss=0.2772, pruned_loss=0.05682, over 19854.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2861, pruned_loss=0.06269, over 3804429.20 frames. ], batch size: 52, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:37:25,663 INFO [train.py:903] (3/4) Epoch 27, batch 3900, loss[loss=0.2129, simple_loss=0.2954, pruned_loss=0.06522, over 18759.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2855, pruned_loss=0.06189, over 3819858.17 frames. ], batch size: 74, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:38:13,057 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=181465.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:38:17,126 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.394e+02 5.123e+02 6.226e+02 8.139e+02 1.802e+03, threshold=1.245e+03, percent-clipped=4.0 +2023-04-03 09:38:28,446 INFO [train.py:903] (3/4) Epoch 27, batch 3950, loss[loss=0.2264, simple_loss=0.3106, pruned_loss=0.07111, over 19820.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2847, pruned_loss=0.06132, over 3824240.62 frames. ], batch size: 52, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:38:30,592 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 09:38:43,821 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=181490.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:39:30,606 INFO [train.py:903] (3/4) Epoch 27, batch 4000, loss[loss=0.1641, simple_loss=0.2413, pruned_loss=0.04348, over 19759.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2841, pruned_loss=0.06096, over 3814727.73 frames. ], batch size: 47, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:39:55,687 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=181548.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:40:08,375 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=181558.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:40:14,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 09:40:21,508 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.200e+02 5.331e+02 6.990e+02 9.796e+02 2.756e+03, threshold=1.398e+03, percent-clipped=11.0 +2023-04-03 09:40:32,606 INFO [train.py:903] (3/4) Epoch 27, batch 4050, loss[loss=0.1689, simple_loss=0.2577, pruned_loss=0.04002, over 19850.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2839, pruned_loss=0.06099, over 3802618.73 frames. ], batch size: 52, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:40:38,826 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=181583.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:41:35,049 INFO [train.py:903] (3/4) Epoch 27, batch 4100, loss[loss=0.2305, simple_loss=0.31, pruned_loss=0.07551, over 19654.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2845, pruned_loss=0.0614, over 3805778.24 frames. ], batch size: 60, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:42:07,712 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 09:42:13,760 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=181659.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:42:26,970 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.417e+02 5.142e+02 6.045e+02 7.647e+02 1.406e+03, threshold=1.209e+03, percent-clipped=1.0 +2023-04-03 09:42:35,660 INFO [train.py:903] (3/4) Epoch 27, batch 4150, loss[loss=0.1967, simple_loss=0.2809, pruned_loss=0.05629, over 19292.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2845, pruned_loss=0.06171, over 3819200.90 frames. ], batch size: 66, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:43:07,931 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4765, 1.3502, 1.5056, 1.6043, 3.1129, 1.0931, 2.4707, 3.4615], + device='cuda:3'), covar=tensor([0.0505, 0.2758, 0.2857, 0.1676, 0.0680, 0.2436, 0.1117, 0.0259], + device='cuda:3'), in_proj_covar=tensor([0.0420, 0.0375, 0.0392, 0.0350, 0.0381, 0.0354, 0.0390, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:43:21,534 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3437, 1.3916, 1.5520, 1.4902, 1.7382, 1.8353, 1.7387, 0.5175], + device='cuda:3'), covar=tensor([0.2535, 0.4399, 0.2797, 0.2102, 0.1796, 0.2433, 0.1565, 0.5300], + device='cuda:3'), in_proj_covar=tensor([0.0553, 0.0669, 0.0751, 0.0506, 0.0635, 0.0547, 0.0670, 0.0571], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 09:43:38,702 INFO [train.py:903] (3/4) Epoch 27, batch 4200, loss[loss=0.1786, simple_loss=0.2627, pruned_loss=0.04725, over 19363.00 frames. ], tot_loss[loss=0.2033, simple_loss=0.2838, pruned_loss=0.06137, over 3822076.70 frames. ], batch size: 47, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:43:42,087 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 09:43:55,242 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4682, 1.3940, 1.6069, 1.5292, 3.0938, 1.1970, 2.4275, 3.4950], + device='cuda:3'), covar=tensor([0.0526, 0.2878, 0.2753, 0.1841, 0.0662, 0.2403, 0.1231, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0377, 0.0394, 0.0352, 0.0382, 0.0356, 0.0392, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:44:30,939 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.172e+02 5.083e+02 6.482e+02 7.948e+02 1.533e+03, threshold=1.296e+03, percent-clipped=4.0 +2023-04-03 09:44:36,209 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=181774.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:44:40,467 INFO [train.py:903] (3/4) Epoch 27, batch 4250, loss[loss=0.242, simple_loss=0.3222, pruned_loss=0.08096, over 19660.00 frames. ], tot_loss[loss=0.2044, simple_loss=0.2853, pruned_loss=0.06181, over 3822012.35 frames. ], batch size: 60, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:44:55,291 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 09:45:08,385 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 09:45:43,719 INFO [train.py:903] (3/4) Epoch 27, batch 4300, loss[loss=0.2165, simple_loss=0.2988, pruned_loss=0.06711, over 17194.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2847, pruned_loss=0.06125, over 3809706.71 frames. ], batch size: 101, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:45:46,616 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9635, 3.6136, 2.4999, 3.2348, 0.9993, 3.5747, 3.4152, 3.5395], + device='cuda:3'), covar=tensor([0.0909, 0.1083, 0.1985, 0.0936, 0.3818, 0.0824, 0.1140, 0.1412], + device='cuda:3'), in_proj_covar=tensor([0.0527, 0.0427, 0.0512, 0.0357, 0.0406, 0.0452, 0.0448, 0.0477], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:46:16,690 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9468, 4.4462, 4.6636, 4.6452, 1.7606, 4.3573, 3.8150, 4.4165], + device='cuda:3'), covar=tensor([0.1758, 0.0776, 0.0617, 0.0709, 0.6165, 0.0949, 0.0710, 0.1054], + device='cuda:3'), in_proj_covar=tensor([0.0823, 0.0784, 0.0994, 0.0870, 0.0864, 0.0755, 0.0588, 0.0922], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 09:46:36,878 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.524e+02 4.917e+02 6.450e+02 8.224e+02 1.543e+03, threshold=1.290e+03, percent-clipped=3.0 +2023-04-03 09:46:40,243 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 09:46:47,394 INFO [train.py:903] (3/4) Epoch 27, batch 4350, loss[loss=0.2187, simple_loss=0.3082, pruned_loss=0.06457, over 19566.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2844, pruned_loss=0.06092, over 3813515.39 frames. ], batch size: 61, lr: 3.03e-03, grad_scale: 4.0 +2023-04-03 09:46:59,920 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3430, 3.0879, 2.2665, 2.7826, 0.8476, 3.0286, 2.9136, 2.9732], + device='cuda:3'), covar=tensor([0.1165, 0.1360, 0.2086, 0.1095, 0.3905, 0.1021, 0.1285, 0.1564], + device='cuda:3'), in_proj_covar=tensor([0.0528, 0.0428, 0.0513, 0.0358, 0.0408, 0.0454, 0.0450, 0.0479], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:47:05,494 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=181892.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:47:25,502 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-04-03 09:47:49,877 INFO [train.py:903] (3/4) Epoch 27, batch 4400, loss[loss=0.2016, simple_loss=0.2896, pruned_loss=0.05684, over 19588.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2853, pruned_loss=0.06148, over 3804970.14 frames. ], batch size: 61, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:48:14,994 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 09:48:24,191 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 09:48:42,706 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.593e+02 4.919e+02 6.507e+02 9.105e+02 1.976e+03, threshold=1.301e+03, percent-clipped=10.0 +2023-04-03 09:48:52,975 INFO [train.py:903] (3/4) Epoch 27, batch 4450, loss[loss=0.2254, simple_loss=0.3031, pruned_loss=0.0738, over 18192.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2853, pruned_loss=0.06137, over 3798980.92 frames. ], batch size: 83, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:48:53,729 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-03 09:48:59,282 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4363, 2.4823, 2.6827, 3.2634, 2.4622, 3.0590, 2.6512, 2.5370], + device='cuda:3'), covar=tensor([0.3932, 0.4033, 0.1787, 0.2421, 0.4216, 0.2131, 0.4579, 0.3198], + device='cuda:3'), in_proj_covar=tensor([0.0930, 0.1006, 0.0739, 0.0948, 0.0907, 0.0848, 0.0855, 0.0805], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 09:49:16,016 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=181996.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:49:30,768 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=182007.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:49:37,369 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0495, 3.6806, 2.6553, 3.2834, 1.0369, 3.6618, 3.5447, 3.6174], + device='cuda:3'), covar=tensor([0.0795, 0.1178, 0.1892, 0.0967, 0.3779, 0.0766, 0.1047, 0.1244], + device='cuda:3'), in_proj_covar=tensor([0.0528, 0.0427, 0.0513, 0.0357, 0.0407, 0.0453, 0.0449, 0.0477], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:49:56,877 INFO [train.py:903] (3/4) Epoch 27, batch 4500, loss[loss=0.2049, simple_loss=0.2721, pruned_loss=0.06886, over 19365.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2845, pruned_loss=0.06119, over 3812076.20 frames. ], batch size: 47, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:49:59,836 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=182030.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:50:31,966 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=182055.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:50:49,955 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.312e+02 5.059e+02 6.218e+02 7.785e+02 2.105e+03, threshold=1.244e+03, percent-clipped=5.0 +2023-04-03 09:51:00,199 INFO [train.py:903] (3/4) Epoch 27, batch 4550, loss[loss=0.1721, simple_loss=0.2499, pruned_loss=0.04715, over 19770.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.284, pruned_loss=0.06074, over 3815519.32 frames. ], batch size: 47, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:51:09,766 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 09:51:32,197 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 09:51:35,759 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9073, 1.3359, 1.0676, 0.9826, 1.1718, 1.0136, 1.0051, 1.2635], + device='cuda:3'), covar=tensor([0.0601, 0.0949, 0.1185, 0.0803, 0.0644, 0.1339, 0.0603, 0.0518], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0321, 0.0337, 0.0273, 0.0251, 0.0345, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:52:02,910 INFO [train.py:903] (3/4) Epoch 27, batch 4600, loss[loss=0.1587, simple_loss=0.2422, pruned_loss=0.03759, over 19622.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.284, pruned_loss=0.06084, over 3820591.33 frames. ], batch size: 50, lr: 3.03e-03, grad_scale: 8.0 +2023-04-03 09:52:54,752 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.336e+02 4.830e+02 5.724e+02 7.323e+02 1.391e+03, threshold=1.145e+03, percent-clipped=2.0 +2023-04-03 09:53:05,187 INFO [train.py:903] (3/4) Epoch 27, batch 4650, loss[loss=0.1852, simple_loss=0.258, pruned_loss=0.05616, over 19752.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2841, pruned_loss=0.06104, over 3819462.38 frames. ], batch size: 45, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:53:22,611 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 09:53:34,165 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 09:53:57,765 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5173, 1.7024, 2.0104, 1.8454, 3.1132, 2.5873, 3.4413, 1.8236], + device='cuda:3'), covar=tensor([0.2655, 0.4496, 0.2962, 0.2023, 0.1585, 0.2222, 0.1556, 0.4238], + device='cuda:3'), in_proj_covar=tensor([0.0552, 0.0667, 0.0751, 0.0505, 0.0636, 0.0544, 0.0668, 0.0570], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 09:54:07,704 INFO [train.py:903] (3/4) Epoch 27, batch 4700, loss[loss=0.2134, simple_loss=0.2988, pruned_loss=0.06394, over 19527.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2835, pruned_loss=0.06077, over 3813162.37 frames. ], batch size: 56, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:54:30,872 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 09:54:51,539 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=182263.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:54:59,120 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.994e+02 4.730e+02 5.895e+02 7.660e+02 1.174e+03, threshold=1.179e+03, percent-clipped=2.0 +2023-04-03 09:55:10,410 INFO [train.py:903] (3/4) Epoch 27, batch 4750, loss[loss=0.1842, simple_loss=0.2604, pruned_loss=0.05399, over 19781.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2831, pruned_loss=0.06062, over 3820256.78 frames. ], batch size: 47, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:55:22,536 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=182288.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:55:36,026 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=182299.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:56:12,222 INFO [train.py:903] (3/4) Epoch 27, batch 4800, loss[loss=0.2278, simple_loss=0.3109, pruned_loss=0.07233, over 19686.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2832, pruned_loss=0.06046, over 3824322.74 frames. ], batch size: 60, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:56:26,909 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=182340.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:56:27,636 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-03 09:56:42,893 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0896, 2.0320, 1.7751, 2.0949, 1.8666, 1.8188, 1.8201, 2.0701], + device='cuda:3'), covar=tensor([0.1106, 0.1398, 0.1530, 0.1086, 0.1401, 0.0575, 0.1493, 0.0705], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0358, 0.0317, 0.0257, 0.0307, 0.0256, 0.0321, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 09:57:03,560 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.957e+02 4.862e+02 5.780e+02 7.243e+02 1.108e+03, threshold=1.156e+03, percent-clipped=0.0 +2023-04-03 09:57:13,593 INFO [train.py:903] (3/4) Epoch 27, batch 4850, loss[loss=0.1537, simple_loss=0.2352, pruned_loss=0.03614, over 19777.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2831, pruned_loss=0.06058, over 3835318.07 frames. ], batch size: 47, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:57:36,901 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 09:57:58,319 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 09:57:58,638 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9602, 4.4833, 4.6981, 4.6875, 1.9240, 4.3896, 3.8185, 4.4409], + device='cuda:3'), covar=tensor([0.1802, 0.0797, 0.0586, 0.0730, 0.6024, 0.0882, 0.0660, 0.1051], + device='cuda:3'), in_proj_covar=tensor([0.0819, 0.0782, 0.0990, 0.0869, 0.0860, 0.0753, 0.0587, 0.0919], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 09:58:03,901 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 09:58:03,922 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 09:58:13,220 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 09:58:14,425 INFO [train.py:903] (3/4) Epoch 27, batch 4900, loss[loss=0.2085, simple_loss=0.2988, pruned_loss=0.05916, over 19532.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2835, pruned_loss=0.06053, over 3834237.38 frames. ], batch size: 64, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:58:34,876 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 09:58:50,308 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=182455.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 09:59:07,215 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.920e+02 4.786e+02 5.871e+02 7.388e+02 1.622e+03, threshold=1.174e+03, percent-clipped=2.0 +2023-04-03 09:59:18,868 INFO [train.py:903] (3/4) Epoch 27, batch 4950, loss[loss=0.2011, simple_loss=0.2842, pruned_loss=0.05897, over 19565.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2836, pruned_loss=0.06026, over 3824293.81 frames. ], batch size: 52, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 09:59:36,547 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 10:00:00,867 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 10:00:21,847 INFO [train.py:903] (3/4) Epoch 27, batch 5000, loss[loss=0.1969, simple_loss=0.2844, pruned_loss=0.05468, over 19387.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.2842, pruned_loss=0.06057, over 3826368.01 frames. ], batch size: 70, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:00:27,404 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=182532.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:00:32,542 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 10:00:44,435 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 10:01:15,154 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.165e+02 4.878e+02 5.976e+02 7.448e+02 1.686e+03, threshold=1.195e+03, percent-clipped=3.0 +2023-04-03 10:01:25,320 INFO [train.py:903] (3/4) Epoch 27, batch 5050, loss[loss=0.1736, simple_loss=0.2493, pruned_loss=0.04898, over 19744.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2841, pruned_loss=0.06033, over 3830675.49 frames. ], batch size: 46, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:02:03,643 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 10:02:27,488 INFO [train.py:903] (3/4) Epoch 27, batch 5100, loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05558, over 19519.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2842, pruned_loss=0.06022, over 3826549.63 frames. ], batch size: 64, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:02:31,202 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8222, 4.3018, 4.6065, 4.5596, 1.8018, 4.2872, 3.7377, 4.3381], + device='cuda:3'), covar=tensor([0.1765, 0.1206, 0.0590, 0.0691, 0.6173, 0.1161, 0.0706, 0.1068], + device='cuda:3'), in_proj_covar=tensor([0.0811, 0.0777, 0.0982, 0.0864, 0.0854, 0.0746, 0.0582, 0.0912], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 10:02:37,760 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 10:02:41,996 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 10:02:46,622 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 10:02:46,783 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=182643.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:03:19,756 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.602e+02 5.110e+02 6.408e+02 8.268e+02 2.195e+03, threshold=1.282e+03, percent-clipped=9.0 +2023-04-03 10:03:30,279 INFO [train.py:903] (3/4) Epoch 27, batch 5150, loss[loss=0.1962, simple_loss=0.2857, pruned_loss=0.05337, over 19658.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2856, pruned_loss=0.06106, over 3817776.00 frames. ], batch size: 58, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:03:44,261 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 10:04:12,440 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=182711.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:04:21,099 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 10:04:24,623 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1104, 5.2057, 5.9561, 5.9815, 2.1782, 5.5897, 4.7621, 5.6066], + device='cuda:3'), covar=tensor([0.1762, 0.0816, 0.0571, 0.0568, 0.6200, 0.0822, 0.0650, 0.1200], + device='cuda:3'), in_proj_covar=tensor([0.0812, 0.0779, 0.0984, 0.0867, 0.0856, 0.0748, 0.0583, 0.0913], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 10:04:26,014 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5560, 2.2961, 1.7450, 1.6117, 2.1184, 1.4230, 1.4777, 1.9168], + device='cuda:3'), covar=tensor([0.1184, 0.0776, 0.1092, 0.0874, 0.0616, 0.1276, 0.0778, 0.0581], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0319, 0.0335, 0.0271, 0.0250, 0.0343, 0.0291, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:04:34,479 INFO [train.py:903] (3/4) Epoch 27, batch 5200, loss[loss=0.2213, simple_loss=0.3136, pruned_loss=0.06452, over 19615.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2846, pruned_loss=0.06055, over 3813713.17 frames. ], batch size: 57, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:04:45,076 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=182736.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:04:50,584 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 10:05:11,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=182758.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:05:19,100 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-04-03 10:05:28,498 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.060e+02 4.726e+02 5.796e+02 7.282e+02 1.371e+03, threshold=1.159e+03, percent-clipped=1.0 +2023-04-03 10:05:35,278 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 10:05:37,678 INFO [train.py:903] (3/4) Epoch 27, batch 5250, loss[loss=0.2052, simple_loss=0.2989, pruned_loss=0.05571, over 19657.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.2846, pruned_loss=0.06037, over 3802463.02 frames. ], batch size: 55, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:05:47,603 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.43 vs. limit=2.0 +2023-04-03 10:05:51,929 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1553, 2.0413, 1.8839, 1.7208, 1.6926, 1.6877, 0.6250, 1.0747], + device='cuda:3'), covar=tensor([0.0641, 0.0718, 0.0544, 0.0886, 0.1202, 0.1002, 0.1464, 0.1184], + device='cuda:3'), in_proj_covar=tensor([0.0365, 0.0363, 0.0370, 0.0393, 0.0472, 0.0397, 0.0347, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 10:06:39,518 INFO [train.py:903] (3/4) Epoch 27, batch 5300, loss[loss=0.1878, simple_loss=0.2536, pruned_loss=0.06105, over 19749.00 frames. ], tot_loss[loss=0.203, simple_loss=0.285, pruned_loss=0.06055, over 3806253.39 frames. ], batch size: 46, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:06:57,248 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.64 vs. limit=5.0 +2023-04-03 10:06:57,448 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 10:06:59,529 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.08 vs. limit=5.0 +2023-04-03 10:07:34,149 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.440e+02 4.985e+02 5.848e+02 7.587e+02 2.195e+03, threshold=1.170e+03, percent-clipped=4.0 +2023-04-03 10:07:37,805 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=182874.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:07:39,975 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=182876.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:07:42,160 INFO [train.py:903] (3/4) Epoch 27, batch 5350, loss[loss=0.1943, simple_loss=0.2773, pruned_loss=0.05563, over 19532.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2848, pruned_loss=0.06052, over 3799124.15 frames. ], batch size: 54, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:07:44,851 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=182880.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:08:16,938 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 10:08:46,436 INFO [train.py:903] (3/4) Epoch 27, batch 5400, loss[loss=0.214, simple_loss=0.2971, pruned_loss=0.06547, over 19529.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.284, pruned_loss=0.06023, over 3816814.31 frames. ], batch size: 56, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:09:41,328 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.624e+02 4.932e+02 6.009e+02 7.341e+02 1.388e+03, threshold=1.202e+03, percent-clipped=2.0 +2023-04-03 10:09:49,115 INFO [train.py:903] (3/4) Epoch 27, batch 5450, loss[loss=0.1798, simple_loss=0.2644, pruned_loss=0.04759, over 19665.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2828, pruned_loss=0.05969, over 3808732.60 frames. ], batch size: 53, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:10:04,451 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=182991.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:10:35,569 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=183014.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:10:40,368 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.48 vs. limit=2.0 +2023-04-03 10:10:51,004 INFO [train.py:903] (3/4) Epoch 27, batch 5500, loss[loss=0.2157, simple_loss=0.3068, pruned_loss=0.06235, over 19663.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2831, pruned_loss=0.05993, over 3814225.77 frames. ], batch size: 55, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:11:05,227 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=183039.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:11:14,836 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 10:11:33,638 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.63 vs. limit=5.0 +2023-04-03 10:11:45,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.090e+02 5.032e+02 6.184e+02 7.491e+02 1.557e+03, threshold=1.237e+03, percent-clipped=5.0 +2023-04-03 10:11:52,113 INFO [train.py:903] (3/4) Epoch 27, batch 5550, loss[loss=0.2129, simple_loss=0.2895, pruned_loss=0.06819, over 19574.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2836, pruned_loss=0.06015, over 3813067.71 frames. ], batch size: 52, lr: 3.02e-03, grad_scale: 4.0 +2023-04-03 10:11:52,643 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5231, 1.5142, 1.7107, 1.7222, 2.2904, 2.2201, 2.3590, 0.9621], + device='cuda:3'), covar=tensor([0.2538, 0.4576, 0.2883, 0.2056, 0.1578, 0.2289, 0.1427, 0.4904], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0672, 0.0758, 0.0509, 0.0638, 0.0548, 0.0673, 0.0575], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:11:57,905 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 10:12:46,680 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 10:12:56,191 INFO [train.py:903] (3/4) Epoch 27, batch 5600, loss[loss=0.2522, simple_loss=0.3248, pruned_loss=0.08981, over 19696.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2833, pruned_loss=0.06003, over 3812053.32 frames. ], batch size: 63, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:13:48,464 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0008, 2.0644, 2.3336, 2.5941, 2.0483, 2.4846, 2.2909, 2.1645], + device='cuda:3'), covar=tensor([0.3951, 0.3768, 0.1745, 0.2300, 0.3785, 0.1989, 0.4519, 0.3131], + device='cuda:3'), in_proj_covar=tensor([0.0933, 0.1012, 0.0742, 0.0950, 0.0910, 0.0849, 0.0859, 0.0807], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:13:51,541 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.195e+02 4.790e+02 6.069e+02 7.863e+02 1.689e+03, threshold=1.214e+03, percent-clipped=3.0 +2023-04-03 10:13:56,159 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.9735, 2.9146, 2.6790, 2.9420, 2.7890, 2.6710, 2.5674, 2.9372], + device='cuda:3'), covar=tensor([0.0851, 0.1415, 0.1260, 0.1019, 0.1240, 0.0468, 0.1237, 0.0598], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0357, 0.0317, 0.0257, 0.0306, 0.0256, 0.0318, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:13:59,319 INFO [train.py:903] (3/4) Epoch 27, batch 5650, loss[loss=0.1834, simple_loss=0.2691, pruned_loss=0.04884, over 19534.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2828, pruned_loss=0.06028, over 3807049.94 frames. ], batch size: 54, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:14:35,068 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7229, 1.7523, 1.6809, 1.4240, 1.4436, 1.4532, 0.2648, 0.7074], + device='cuda:3'), covar=tensor([0.0719, 0.0661, 0.0467, 0.0738, 0.1353, 0.0822, 0.1457, 0.1229], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0364, 0.0372, 0.0394, 0.0472, 0.0398, 0.0347, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 10:14:36,254 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4517, 1.5659, 1.8079, 1.6824, 2.7636, 2.2032, 2.9361, 1.4851], + device='cuda:3'), covar=tensor([0.2702, 0.4722, 0.3069, 0.2202, 0.1592, 0.2432, 0.1465, 0.4635], + device='cuda:3'), in_proj_covar=tensor([0.0555, 0.0672, 0.0758, 0.0510, 0.0639, 0.0548, 0.0673, 0.0577], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:14:44,836 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 10:14:49,514 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=183218.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:14:56,365 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=183224.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:15:01,000 INFO [train.py:903] (3/4) Epoch 27, batch 5700, loss[loss=0.1652, simple_loss=0.2471, pruned_loss=0.04163, over 19609.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2837, pruned_loss=0.06038, over 3806645.13 frames. ], batch size: 50, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:15:25,157 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=183247.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:15:54,499 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.213e+02 5.103e+02 5.954e+02 7.593e+02 1.308e+03, threshold=1.191e+03, percent-clipped=1.0 +2023-04-03 10:15:54,976 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=183272.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:15:59,205 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 10:16:01,598 INFO [train.py:903] (3/4) Epoch 27, batch 5750, loss[loss=0.1997, simple_loss=0.28, pruned_loss=0.0597, over 19578.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2831, pruned_loss=0.06042, over 3823645.41 frames. ], batch size: 52, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:16:08,285 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 10:16:13,692 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 10:17:05,409 INFO [train.py:903] (3/4) Epoch 27, batch 5800, loss[loss=0.2069, simple_loss=0.29, pruned_loss=0.06189, over 19315.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2835, pruned_loss=0.06064, over 3840255.16 frames. ], batch size: 66, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:17:12,709 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=183333.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:17:19,497 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=183339.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:17:59,580 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.172e+02 5.364e+02 7.027e+02 9.052e+02 1.891e+03, threshold=1.405e+03, percent-clipped=4.0 +2023-04-03 10:18:07,742 INFO [train.py:903] (3/4) Epoch 27, batch 5850, loss[loss=0.2089, simple_loss=0.2933, pruned_loss=0.06223, over 19703.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2843, pruned_loss=0.06099, over 3833316.09 frames. ], batch size: 59, lr: 3.02e-03, grad_scale: 8.0 +2023-04-03 10:18:12,169 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 10:18:12,952 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4839, 2.6827, 2.2199, 2.6472, 2.4790, 2.2089, 2.2124, 2.4931], + device='cuda:3'), covar=tensor([0.1032, 0.1377, 0.1372, 0.1009, 0.1239, 0.0527, 0.1299, 0.0686], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0355, 0.0316, 0.0256, 0.0305, 0.0255, 0.0317, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:18:31,484 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-03 10:18:46,969 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0480, 2.1688, 2.4036, 2.6983, 2.1103, 2.5998, 2.3717, 2.2164], + device='cuda:3'), covar=tensor([0.4252, 0.3870, 0.1924, 0.2492, 0.4175, 0.2178, 0.4950, 0.3314], + device='cuda:3'), in_proj_covar=tensor([0.0930, 0.1008, 0.0740, 0.0946, 0.0907, 0.0847, 0.0858, 0.0803], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:19:07,136 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=183426.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:19:08,989 INFO [train.py:903] (3/4) Epoch 27, batch 5900, loss[loss=0.1921, simple_loss=0.2863, pruned_loss=0.04899, over 19707.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2824, pruned_loss=0.0601, over 3838838.00 frames. ], batch size: 59, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:19:09,033 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 10:19:32,072 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 10:20:03,306 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.958e+02 4.868e+02 5.663e+02 7.388e+02 1.454e+03, threshold=1.133e+03, percent-clipped=1.0 +2023-04-03 10:20:10,243 INFO [train.py:903] (3/4) Epoch 27, batch 5950, loss[loss=0.2151, simple_loss=0.2986, pruned_loss=0.0658, over 19466.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2833, pruned_loss=0.06048, over 3844486.09 frames. ], batch size: 64, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:21:12,623 INFO [train.py:903] (3/4) Epoch 27, batch 6000, loss[loss=0.1761, simple_loss=0.2587, pruned_loss=0.04674, over 19720.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2823, pruned_loss=0.05993, over 3851891.39 frames. ], batch size: 51, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:21:12,624 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 10:21:25,588 INFO [train.py:937] (3/4) Epoch 27, validation: loss=0.1675, simple_loss=0.2669, pruned_loss=0.03401, over 944034.00 frames. +2023-04-03 10:21:25,589 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 10:21:39,339 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 10:22:22,442 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.170e+02 5.376e+02 6.105e+02 7.773e+02 1.848e+03, threshold=1.221e+03, percent-clipped=6.0 +2023-04-03 10:22:28,409 INFO [train.py:903] (3/4) Epoch 27, batch 6050, loss[loss=0.2794, simple_loss=0.3366, pruned_loss=0.1111, over 13105.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2823, pruned_loss=0.05989, over 3841937.93 frames. ], batch size: 136, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:22:43,124 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=183589.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:22:50,964 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=183595.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:22:55,539 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0399, 3.2567, 1.8606, 2.0381, 3.0035, 1.6331, 1.3884, 2.1960], + device='cuda:3'), covar=tensor([0.1458, 0.0646, 0.1305, 0.0942, 0.0543, 0.1428, 0.1144, 0.0783], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0319, 0.0336, 0.0273, 0.0250, 0.0344, 0.0292, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:23:08,789 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.65 vs. limit=2.0 +2023-04-03 10:23:14,128 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=183614.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:23:20,836 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=183620.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:23:29,936 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.67 vs. limit=5.0 +2023-04-03 10:23:31,577 INFO [train.py:903] (3/4) Epoch 27, batch 6100, loss[loss=0.1829, simple_loss=0.2601, pruned_loss=0.05287, over 18703.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2827, pruned_loss=0.06013, over 3838379.78 frames. ], batch size: 41, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:23:37,192 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=183632.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:24:00,157 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7747, 1.9235, 2.1841, 2.2582, 1.7619, 2.1815, 2.1329, 1.9595], + device='cuda:3'), covar=tensor([0.4200, 0.3595, 0.2014, 0.2344, 0.3810, 0.2168, 0.5008, 0.3415], + device='cuda:3'), in_proj_covar=tensor([0.0935, 0.1011, 0.0743, 0.0950, 0.0912, 0.0850, 0.0862, 0.0805], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:24:27,763 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.311e+02 5.176e+02 6.035e+02 7.763e+02 1.396e+03, threshold=1.207e+03, percent-clipped=4.0 +2023-04-03 10:24:28,029 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9160, 4.4592, 2.8985, 3.8687, 0.8701, 4.4662, 4.2765, 4.4296], + device='cuda:3'), covar=tensor([0.0615, 0.1079, 0.1899, 0.0893, 0.4319, 0.0666, 0.0989, 0.1163], + device='cuda:3'), in_proj_covar=tensor([0.0527, 0.0428, 0.0516, 0.0357, 0.0410, 0.0454, 0.0451, 0.0479], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:24:33,565 INFO [train.py:903] (3/4) Epoch 27, batch 6150, loss[loss=0.1812, simple_loss=0.2733, pruned_loss=0.04456, over 19781.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.282, pruned_loss=0.05952, over 3846959.56 frames. ], batch size: 56, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:24:39,685 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0472, 3.6954, 2.6870, 3.2978, 0.9149, 3.6802, 3.5425, 3.6058], + device='cuda:3'), covar=tensor([0.0829, 0.1287, 0.1928, 0.0999, 0.4162, 0.0769, 0.1029, 0.1231], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0429, 0.0517, 0.0357, 0.0411, 0.0455, 0.0452, 0.0480], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:24:53,484 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9452, 2.0979, 2.2132, 2.1163, 3.6601, 1.7612, 3.0093, 3.7461], + device='cuda:3'), covar=tensor([0.0462, 0.2276, 0.2345, 0.1690, 0.0577, 0.2247, 0.1524, 0.0255], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0379, 0.0397, 0.0353, 0.0385, 0.0358, 0.0396, 0.0416], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:25:01,288 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 10:25:10,556 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2096, 1.2977, 1.7120, 1.1564, 2.5548, 3.4125, 3.1345, 3.5609], + device='cuda:3'), covar=tensor([0.1560, 0.3906, 0.3389, 0.2633, 0.0634, 0.0202, 0.0216, 0.0288], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0334, 0.0365, 0.0273, 0.0255, 0.0197, 0.0220, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 10:25:34,995 INFO [train.py:903] (3/4) Epoch 27, batch 6200, loss[loss=0.2408, simple_loss=0.3265, pruned_loss=0.07749, over 19731.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2825, pruned_loss=0.05985, over 3839723.24 frames. ], batch size: 63, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:26:22,170 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-03 10:26:28,522 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=183770.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:26:28,796 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2187, 1.2860, 1.2464, 1.0595, 1.1040, 1.1265, 0.0867, 0.3848], + device='cuda:3'), covar=tensor([0.0722, 0.0701, 0.0495, 0.0639, 0.1358, 0.0674, 0.1470, 0.1213], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0358, 0.0366, 0.0388, 0.0466, 0.0391, 0.0342, 0.0345], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 10:26:31,934 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.388e+02 4.839e+02 5.826e+02 7.637e+02 1.855e+03, threshold=1.165e+03, percent-clipped=4.0 +2023-04-03 10:26:37,746 INFO [train.py:903] (3/4) Epoch 27, batch 6250, loss[loss=0.201, simple_loss=0.2847, pruned_loss=0.0587, over 19659.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.283, pruned_loss=0.06042, over 3834207.92 frames. ], batch size: 55, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:27:08,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 10:27:40,817 INFO [train.py:903] (3/4) Epoch 27, batch 6300, loss[loss=0.2054, simple_loss=0.2827, pruned_loss=0.06402, over 19402.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.283, pruned_loss=0.06089, over 3823758.32 frames. ], batch size: 48, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:28:20,290 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=183860.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:28:36,672 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.559e+02 4.928e+02 5.882e+02 7.199e+02 1.705e+03, threshold=1.176e+03, percent-clipped=2.0 +2023-04-03 10:28:43,614 INFO [train.py:903] (3/4) Epoch 27, batch 6350, loss[loss=0.1691, simple_loss=0.2648, pruned_loss=0.03664, over 19601.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2832, pruned_loss=0.06086, over 3807866.42 frames. ], batch size: 57, lr: 3.01e-03, grad_scale: 4.0 +2023-04-03 10:28:51,874 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=183885.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:29:43,978 INFO [train.py:903] (3/4) Epoch 27, batch 6400, loss[loss=0.1847, simple_loss=0.2673, pruned_loss=0.05109, over 19719.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2827, pruned_loss=0.06046, over 3806464.04 frames. ], batch size: 51, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:30:21,983 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1801, 5.2439, 5.9825, 5.9923, 2.0161, 5.6222, 4.7840, 5.7107], + device='cuda:3'), covar=tensor([0.1777, 0.0773, 0.0603, 0.0618, 0.6277, 0.0858, 0.0680, 0.1137], + device='cuda:3'), in_proj_covar=tensor([0.0823, 0.0786, 0.0995, 0.0875, 0.0867, 0.0757, 0.0589, 0.0926], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 10:30:39,966 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.119e+02 4.757e+02 6.035e+02 7.575e+02 1.901e+03, threshold=1.207e+03, percent-clipped=7.0 +2023-04-03 10:30:43,705 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=183976.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:30:45,870 INFO [train.py:903] (3/4) Epoch 27, batch 6450, loss[loss=0.1617, simple_loss=0.2438, pruned_loss=0.03977, over 19311.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2835, pruned_loss=0.06088, over 3811255.41 frames. ], batch size: 44, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:31:28,476 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 10:31:50,531 INFO [train.py:903] (3/4) Epoch 27, batch 6500, loss[loss=0.1558, simple_loss=0.2347, pruned_loss=0.03845, over 19728.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2839, pruned_loss=0.06082, over 3800610.73 frames. ], batch size: 46, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:31:52,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 10:32:40,180 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=184068.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:32:46,400 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.099e+02 4.542e+02 5.565e+02 7.286e+02 1.442e+03, threshold=1.113e+03, percent-clipped=3.0 +2023-04-03 10:32:53,095 INFO [train.py:903] (3/4) Epoch 27, batch 6550, loss[loss=0.1772, simple_loss=0.2585, pruned_loss=0.04798, over 19803.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2834, pruned_loss=0.06049, over 3815977.26 frames. ], batch size: 49, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:33:08,717 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=184091.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:33:47,754 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 10:33:55,001 INFO [train.py:903] (3/4) Epoch 27, batch 6600, loss[loss=0.2224, simple_loss=0.3084, pruned_loss=0.06827, over 18296.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2825, pruned_loss=0.06013, over 3820296.17 frames. ], batch size: 83, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:34:11,923 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=184141.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:34:43,118 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=184166.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:34:50,799 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.027e+02 4.833e+02 5.798e+02 7.178e+02 1.551e+03, threshold=1.160e+03, percent-clipped=2.0 +2023-04-03 10:34:58,097 INFO [train.py:903] (3/4) Epoch 27, batch 6650, loss[loss=0.2044, simple_loss=0.2912, pruned_loss=0.05878, over 18896.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2821, pruned_loss=0.05956, over 3831896.31 frames. ], batch size: 74, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:35:30,804 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=184204.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:35:32,185 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1497, 1.2513, 1.7270, 1.2324, 2.5547, 3.3537, 3.0559, 3.5691], + device='cuda:3'), covar=tensor([0.1734, 0.4092, 0.3517, 0.2759, 0.0641, 0.0225, 0.0240, 0.0270], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0331, 0.0362, 0.0271, 0.0253, 0.0195, 0.0219, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 10:35:56,593 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-03 10:35:59,888 INFO [train.py:903] (3/4) Epoch 27, batch 6700, loss[loss=0.2144, simple_loss=0.2996, pruned_loss=0.06466, over 18775.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2837, pruned_loss=0.06022, over 3833560.75 frames. ], batch size: 74, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:36:02,643 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.73 vs. limit=2.0 +2023-04-03 10:36:52,206 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.305e+02 5.122e+02 6.084e+02 8.596e+02 2.606e+03, threshold=1.217e+03, percent-clipped=9.0 +2023-04-03 10:36:57,982 INFO [train.py:903] (3/4) Epoch 27, batch 6750, loss[loss=0.2098, simple_loss=0.2951, pruned_loss=0.06229, over 17542.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2836, pruned_loss=0.06053, over 3831349.17 frames. ], batch size: 101, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:37:44,310 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=184319.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:37:54,072 INFO [train.py:903] (3/4) Epoch 27, batch 6800, loss[loss=0.1987, simple_loss=0.2849, pruned_loss=0.05624, over 19652.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2852, pruned_loss=0.06199, over 3811059.87 frames. ], batch size: 55, lr: 3.01e-03, grad_scale: 8.0 +2023-04-03 10:38:15,913 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=184347.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:38:16,784 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8668, 4.0084, 4.4220, 4.4273, 2.7307, 4.1038, 3.7850, 4.1942], + device='cuda:3'), covar=tensor([0.1436, 0.3173, 0.0614, 0.0743, 0.4399, 0.1384, 0.0644, 0.1035], + device='cuda:3'), in_proj_covar=tensor([0.0827, 0.0787, 0.1000, 0.0878, 0.0869, 0.0762, 0.0590, 0.0930], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 10:38:39,138 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 10:38:40,151 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 10:38:43,606 INFO [train.py:903] (3/4) Epoch 28, batch 0, loss[loss=0.2165, simple_loss=0.302, pruned_loss=0.06546, over 19583.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.302, pruned_loss=0.06546, over 19583.00 frames. ], batch size: 61, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:38:43,606 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 10:38:54,480 INFO [train.py:937] (3/4) Epoch 28, validation: loss=0.1665, simple_loss=0.2666, pruned_loss=0.03316, over 944034.00 frames. +2023-04-03 10:38:54,481 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 10:39:08,338 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 10:39:14,457 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=184372.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:39:15,175 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.715e+02 5.190e+02 6.304e+02 8.212e+02 1.288e+03, threshold=1.261e+03, percent-clipped=2.0 +2023-04-03 10:39:21,680 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=184377.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:39:42,083 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4240, 1.4514, 1.8504, 1.7066, 2.5797, 2.1429, 2.7167, 1.1755], + device='cuda:3'), covar=tensor([0.2893, 0.4937, 0.3087, 0.2202, 0.1758, 0.2582, 0.1708, 0.5340], + device='cuda:3'), in_proj_covar=tensor([0.0555, 0.0670, 0.0753, 0.0507, 0.0639, 0.0546, 0.0670, 0.0573], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 10:39:57,727 INFO [train.py:903] (3/4) Epoch 28, batch 50, loss[loss=0.2629, simple_loss=0.3261, pruned_loss=0.09984, over 13209.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2824, pruned_loss=0.06029, over 866102.15 frames. ], batch size: 135, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:40:04,866 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=184412.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:40:32,251 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 10:40:58,084 INFO [train.py:903] (3/4) Epoch 28, batch 100, loss[loss=0.2309, simple_loss=0.3089, pruned_loss=0.0764, over 19759.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2813, pruned_loss=0.05956, over 1523617.42 frames. ], batch size: 63, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:41:08,342 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 10:41:18,602 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.133e+02 4.572e+02 5.776e+02 7.316e+02 1.195e+03, threshold=1.155e+03, percent-clipped=0.0 +2023-04-03 10:41:56,033 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=184504.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:41:58,109 INFO [train.py:903] (3/4) Epoch 28, batch 150, loss[loss=0.1977, simple_loss=0.2848, pruned_loss=0.05531, over 19757.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2808, pruned_loss=0.05965, over 2038332.96 frames. ], batch size: 63, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:42:21,067 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5379, 2.1630, 1.6795, 1.5031, 1.9872, 1.3367, 1.3937, 1.9645], + device='cuda:3'), covar=tensor([0.0995, 0.0831, 0.1125, 0.0945, 0.0620, 0.1396, 0.0770, 0.0488], + device='cuda:3'), in_proj_covar=tensor([0.0299, 0.0317, 0.0334, 0.0270, 0.0248, 0.0341, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:42:24,573 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=184527.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:42:57,421 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 10:42:58,595 INFO [train.py:903] (3/4) Epoch 28, batch 200, loss[loss=0.159, simple_loss=0.2336, pruned_loss=0.04222, over 19717.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2809, pruned_loss=0.05908, over 2441141.06 frames. ], batch size: 46, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:43:19,481 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.222e+02 4.969e+02 6.258e+02 7.516e+02 2.266e+03, threshold=1.252e+03, percent-clipped=4.0 +2023-04-03 10:43:22,291 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=184575.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:43:51,813 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6049, 1.3625, 1.4854, 1.5233, 3.1698, 1.2836, 2.4251, 3.6589], + device='cuda:3'), covar=tensor([0.0542, 0.2895, 0.3100, 0.1937, 0.0731, 0.2508, 0.1376, 0.0247], + device='cuda:3'), in_proj_covar=tensor([0.0427, 0.0380, 0.0398, 0.0354, 0.0385, 0.0359, 0.0397, 0.0417], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 10:43:51,898 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=184600.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:43:59,858 INFO [train.py:903] (3/4) Epoch 28, batch 250, loss[loss=0.2046, simple_loss=0.2699, pruned_loss=0.06963, over 19029.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2831, pruned_loss=0.06003, over 2748795.65 frames. ], batch size: 42, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:45:01,682 INFO [train.py:903] (3/4) Epoch 28, batch 300, loss[loss=0.2347, simple_loss=0.3061, pruned_loss=0.08161, over 19769.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2811, pruned_loss=0.05895, over 2999710.31 frames. ], batch size: 54, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:45:22,239 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 4.936e+02 6.396e+02 8.049e+02 1.564e+03, threshold=1.279e+03, percent-clipped=7.0 +2023-04-03 10:46:02,785 INFO [train.py:903] (3/4) Epoch 28, batch 350, loss[loss=0.2071, simple_loss=0.2913, pruned_loss=0.06149, over 17323.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2814, pruned_loss=0.05921, over 3176831.71 frames. ], batch size: 101, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:46:06,329 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 10:46:20,196 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=184721.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:46:39,110 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2946, 2.1390, 2.1135, 1.9276, 1.7763, 1.8669, 0.7347, 1.2503], + device='cuda:3'), covar=tensor([0.0654, 0.0663, 0.0477, 0.0785, 0.1213, 0.0894, 0.1363, 0.1100], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0364, 0.0370, 0.0393, 0.0471, 0.0395, 0.0346, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 10:47:04,862 INFO [train.py:903] (3/4) Epoch 28, batch 400, loss[loss=0.198, simple_loss=0.2788, pruned_loss=0.05859, over 19760.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2818, pruned_loss=0.05912, over 3337441.87 frames. ], batch size: 54, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:47:24,991 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.223e+02 4.910e+02 5.909e+02 7.518e+02 1.907e+03, threshold=1.182e+03, percent-clipped=3.0 +2023-04-03 10:47:38,241 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=184783.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:48:05,126 INFO [train.py:903] (3/4) Epoch 28, batch 450, loss[loss=0.2183, simple_loss=0.2979, pruned_loss=0.06939, over 19519.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2811, pruned_loss=0.05897, over 3445775.29 frames. ], batch size: 54, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:48:07,846 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=184808.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:48:38,459 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 10:48:39,481 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 10:48:43,531 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=184836.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:48:56,809 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=184848.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:49:03,975 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2876, 2.1303, 1.8995, 1.8306, 1.6413, 1.7840, 0.5113, 1.2612], + device='cuda:3'), covar=tensor([0.0683, 0.0696, 0.0637, 0.0970, 0.1391, 0.1073, 0.1576, 0.1246], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0365, 0.0371, 0.0395, 0.0474, 0.0397, 0.0348, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 10:49:06,862 INFO [train.py:903] (3/4) Epoch 28, batch 500, loss[loss=0.2278, simple_loss=0.2952, pruned_loss=0.08017, over 19389.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2812, pruned_loss=0.05908, over 3543619.83 frames. ], batch size: 47, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:49:28,423 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.110e+02 5.109e+02 6.404e+02 8.256e+02 1.456e+03, threshold=1.281e+03, percent-clipped=5.0 +2023-04-03 10:49:41,485 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=184884.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:50:09,155 INFO [train.py:903] (3/4) Epoch 28, batch 550, loss[loss=0.1739, simple_loss=0.2513, pruned_loss=0.04826, over 19750.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2816, pruned_loss=0.05906, over 3616775.36 frames. ], batch size: 45, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:51:11,432 INFO [train.py:903] (3/4) Epoch 28, batch 600, loss[loss=0.2696, simple_loss=0.3404, pruned_loss=0.09944, over 19667.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2814, pruned_loss=0.05908, over 3673233.30 frames. ], batch size: 58, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:51:12,920 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=184957.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:51:19,893 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=184963.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:51:31,267 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.601e+02 5.007e+02 6.276e+02 8.222e+02 1.849e+03, threshold=1.255e+03, percent-clipped=3.0 +2023-04-03 10:51:50,595 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 10:52:14,346 INFO [train.py:903] (3/4) Epoch 28, batch 650, loss[loss=0.1901, simple_loss=0.2785, pruned_loss=0.05086, over 18764.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2823, pruned_loss=0.05952, over 3707798.61 frames. ], batch size: 74, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:53:16,119 INFO [train.py:903] (3/4) Epoch 28, batch 700, loss[loss=0.1804, simple_loss=0.2577, pruned_loss=0.05157, over 19383.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.282, pruned_loss=0.05938, over 3740183.81 frames. ], batch size: 48, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:53:31,593 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6721, 1.5311, 1.6176, 2.0340, 1.7101, 1.8033, 1.8637, 1.6664], + device='cuda:3'), covar=tensor([0.0755, 0.0833, 0.0862, 0.0634, 0.0950, 0.0716, 0.0852, 0.0665], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0226, 0.0228, 0.0240, 0.0227, 0.0215, 0.0189, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 10:53:38,017 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.143e+02 4.589e+02 5.550e+02 7.126e+02 1.317e+03, threshold=1.110e+03, percent-clipped=1.0 +2023-04-03 10:53:45,177 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=185078.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:54:01,399 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185092.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:54:19,781 INFO [train.py:903] (3/4) Epoch 28, batch 750, loss[loss=0.2195, simple_loss=0.2861, pruned_loss=0.07648, over 19463.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2826, pruned_loss=0.05986, over 3768532.80 frames. ], batch size: 49, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:54:34,061 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185117.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:55:10,749 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=185147.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:55:20,953 INFO [train.py:903] (3/4) Epoch 28, batch 800, loss[loss=0.1948, simple_loss=0.2844, pruned_loss=0.05261, over 18048.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2825, pruned_loss=0.0597, over 3787428.94 frames. ], batch size: 83, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:55:34,889 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 10:55:41,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.136e+02 5.493e+02 6.558e+02 7.858e+02 2.224e+03, threshold=1.312e+03, percent-clipped=8.0 +2023-04-03 10:55:53,722 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2349, 1.3631, 2.0779, 1.6632, 3.1842, 4.7131, 4.6301, 5.1488], + device='cuda:3'), covar=tensor([0.1708, 0.3995, 0.3251, 0.2373, 0.0604, 0.0233, 0.0178, 0.0178], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0331, 0.0363, 0.0271, 0.0253, 0.0195, 0.0219, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 10:56:24,245 INFO [train.py:903] (3/4) Epoch 28, batch 850, loss[loss=0.1935, simple_loss=0.2706, pruned_loss=0.05821, over 19396.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2819, pruned_loss=0.0592, over 3802775.68 frames. ], batch size: 48, lr: 2.95e-03, grad_scale: 8.0 +2023-04-03 10:56:39,434 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185219.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:56:51,612 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=185228.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:57:11,581 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185244.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:57:15,603 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 10:57:24,912 INFO [train.py:903] (3/4) Epoch 28, batch 900, loss[loss=0.1918, simple_loss=0.2731, pruned_loss=0.05522, over 19536.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2822, pruned_loss=0.05959, over 3812762.18 frames. ], batch size: 54, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 10:57:47,704 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.300e+02 4.798e+02 5.781e+02 7.336e+02 1.381e+03, threshold=1.156e+03, percent-clipped=1.0 +2023-04-03 10:58:21,556 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=185301.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:58:28,126 INFO [train.py:903] (3/4) Epoch 28, batch 950, loss[loss=0.1788, simple_loss=0.2499, pruned_loss=0.05387, over 19778.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2819, pruned_loss=0.05932, over 3813282.08 frames. ], batch size: 46, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 10:58:29,318 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 10:58:40,880 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-03 10:59:15,319 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=185343.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 10:59:32,235 INFO [train.py:903] (3/4) Epoch 28, batch 1000, loss[loss=0.1825, simple_loss=0.2713, pruned_loss=0.04689, over 18383.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2823, pruned_loss=0.05948, over 3815431.87 frames. ], batch size: 84, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 10:59:53,728 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.982e+02 4.891e+02 5.853e+02 7.878e+02 2.572e+03, threshold=1.171e+03, percent-clipped=6.0 +2023-04-03 11:00:23,365 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 11:00:34,735 INFO [train.py:903] (3/4) Epoch 28, batch 1050, loss[loss=0.205, simple_loss=0.2945, pruned_loss=0.05772, over 19307.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2827, pruned_loss=0.05963, over 3818228.83 frames. ], batch size: 66, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:00:47,037 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=185416.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:00:53,766 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=185422.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:01:02,624 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 11:01:35,159 INFO [train.py:903] (3/4) Epoch 28, batch 1100, loss[loss=0.1605, simple_loss=0.239, pruned_loss=0.04103, over 19787.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2838, pruned_loss=0.06062, over 3808771.09 frames. ], batch size: 48, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:01:57,231 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.219e+02 4.947e+02 6.329e+02 8.288e+02 1.903e+03, threshold=1.266e+03, percent-clipped=3.0 +2023-04-03 11:02:13,034 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=185486.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:02:18,712 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=185491.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:02:35,899 INFO [train.py:903] (3/4) Epoch 28, batch 1150, loss[loss=0.1735, simple_loss=0.2531, pruned_loss=0.04694, over 19625.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2841, pruned_loss=0.06079, over 3798271.28 frames. ], batch size: 50, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:02:55,995 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 11:03:15,694 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=185537.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:03:40,780 INFO [train.py:903] (3/4) Epoch 28, batch 1200, loss[loss=0.2033, simple_loss=0.2843, pruned_loss=0.06113, over 19716.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2843, pruned_loss=0.06065, over 3808340.98 frames. ], batch size: 51, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:04:01,640 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.374e+02 5.383e+02 6.799e+02 8.653e+02 1.626e+03, threshold=1.360e+03, percent-clipped=3.0 +2023-04-03 11:04:11,755 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 11:04:34,073 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185599.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:04:41,622 INFO [train.py:903] (3/4) Epoch 28, batch 1250, loss[loss=0.2206, simple_loss=0.2899, pruned_loss=0.07562, over 19785.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2842, pruned_loss=0.06087, over 3820520.08 frames. ], batch size: 54, lr: 2.94e-03, grad_scale: 16.0 +2023-04-03 11:04:41,975 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=185606.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:05:03,725 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185624.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:05:44,893 INFO [train.py:903] (3/4) Epoch 28, batch 1300, loss[loss=0.1662, simple_loss=0.2481, pruned_loss=0.0421, over 19746.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2836, pruned_loss=0.06037, over 3824111.07 frames. ], batch size: 51, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:06:04,634 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185672.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:06:06,494 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.213e+02 4.542e+02 5.530e+02 7.592e+02 1.164e+03, threshold=1.106e+03, percent-clipped=0.0 +2023-04-03 11:06:35,949 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185697.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:06:46,284 INFO [train.py:903] (3/4) Epoch 28, batch 1350, loss[loss=0.1675, simple_loss=0.2465, pruned_loss=0.04422, over 19762.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2827, pruned_loss=0.06001, over 3834137.93 frames. ], batch size: 47, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:07:24,108 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5552, 1.2552, 1.2232, 1.4985, 1.1133, 1.3482, 1.2447, 1.4283], + device='cuda:3'), covar=tensor([0.1236, 0.1254, 0.1685, 0.1082, 0.1374, 0.0673, 0.1721, 0.0891], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0361, 0.0320, 0.0259, 0.0308, 0.0259, 0.0323, 0.0266], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 11:07:31,232 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2171, 1.3504, 1.7686, 1.1757, 2.5644, 3.3571, 3.0630, 3.5744], + device='cuda:3'), covar=tensor([0.1607, 0.3823, 0.3355, 0.2630, 0.0647, 0.0230, 0.0266, 0.0350], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0332, 0.0365, 0.0272, 0.0255, 0.0196, 0.0220, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 11:07:35,958 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3591, 1.2818, 1.6800, 1.2034, 2.5146, 3.3266, 3.0336, 3.5636], + device='cuda:3'), covar=tensor([0.1505, 0.4013, 0.3634, 0.2700, 0.0669, 0.0225, 0.0261, 0.0315], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0332, 0.0365, 0.0272, 0.0255, 0.0196, 0.0220, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 11:07:48,953 INFO [train.py:903] (3/4) Epoch 28, batch 1400, loss[loss=0.2152, simple_loss=0.293, pruned_loss=0.06874, over 19529.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2826, pruned_loss=0.06013, over 3829923.43 frames. ], batch size: 64, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:08:11,867 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.444e+02 4.899e+02 6.106e+02 7.699e+02 1.518e+03, threshold=1.221e+03, percent-clipped=6.0 +2023-04-03 11:08:15,711 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=185777.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:08:35,156 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185793.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:08:49,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 11:08:51,547 INFO [train.py:903] (3/4) Epoch 28, batch 1450, loss[loss=0.2211, simple_loss=0.3067, pruned_loss=0.06777, over 17945.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2824, pruned_loss=0.0601, over 3809679.58 frames. ], batch size: 83, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:08:53,248 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0797, 2.1624, 2.3240, 2.7793, 2.1173, 2.5961, 2.3084, 2.1666], + device='cuda:3'), covar=tensor([0.4233, 0.4049, 0.1948, 0.2441, 0.4225, 0.2258, 0.5003, 0.3447], + device='cuda:3'), in_proj_covar=tensor([0.0935, 0.1012, 0.0742, 0.0950, 0.0910, 0.0850, 0.0860, 0.0808], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 11:09:00,618 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-03 11:09:05,685 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=185817.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 11:09:06,967 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185818.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:09:20,222 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=185830.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:09:38,505 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2650, 1.2319, 1.2485, 1.3916, 1.0490, 1.3948, 1.3844, 1.3190], + device='cuda:3'), covar=tensor([0.0944, 0.0994, 0.1097, 0.0638, 0.0908, 0.0866, 0.0808, 0.0808], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0225, 0.0228, 0.0240, 0.0227, 0.0215, 0.0188, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 11:09:54,217 INFO [train.py:903] (3/4) Epoch 28, batch 1500, loss[loss=0.2451, simple_loss=0.3294, pruned_loss=0.08043, over 18756.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2838, pruned_loss=0.06117, over 3813615.31 frames. ], batch size: 74, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:10:02,016 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=185862.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:10:15,979 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.811e+02 5.017e+02 6.227e+02 8.666e+02 1.816e+03, threshold=1.245e+03, percent-clipped=11.0 +2023-04-03 11:10:33,549 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=185887.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:10:56,447 INFO [train.py:903] (3/4) Epoch 28, batch 1550, loss[loss=0.174, simple_loss=0.2557, pruned_loss=0.04616, over 19349.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.283, pruned_loss=0.06066, over 3814968.03 frames. ], batch size: 47, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:11:45,524 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=185945.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:11:58,295 INFO [train.py:903] (3/4) Epoch 28, batch 1600, loss[loss=0.2486, simple_loss=0.3213, pruned_loss=0.08795, over 19461.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.2833, pruned_loss=0.06109, over 3816485.13 frames. ], batch size: 70, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:12:20,803 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 11:12:23,176 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.991e+02 4.773e+02 5.899e+02 6.974e+02 1.687e+03, threshold=1.180e+03, percent-clipped=2.0 +2023-04-03 11:13:03,460 INFO [train.py:903] (3/4) Epoch 28, batch 1650, loss[loss=0.2061, simple_loss=0.287, pruned_loss=0.06266, over 19577.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2829, pruned_loss=0.06094, over 3820116.89 frames. ], batch size: 52, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:13:54,022 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-03 11:14:07,006 INFO [train.py:903] (3/4) Epoch 28, batch 1700, loss[loss=0.1777, simple_loss=0.2453, pruned_loss=0.05502, over 19070.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2834, pruned_loss=0.06094, over 3825651.79 frames. ], batch size: 42, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:14:29,606 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.897e+02 4.751e+02 5.670e+02 7.170e+02 1.723e+03, threshold=1.134e+03, percent-clipped=7.0 +2023-04-03 11:14:44,410 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 11:15:08,855 INFO [train.py:903] (3/4) Epoch 28, batch 1750, loss[loss=0.2552, simple_loss=0.3263, pruned_loss=0.0921, over 19491.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2828, pruned_loss=0.06027, over 3818446.20 frames. ], batch size: 64, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:15:15,098 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2606, 2.1128, 1.9567, 1.7836, 1.7547, 1.7694, 0.6716, 1.1813], + device='cuda:3'), covar=tensor([0.0735, 0.0700, 0.0613, 0.0997, 0.1300, 0.1035, 0.1477, 0.1214], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0366, 0.0371, 0.0393, 0.0471, 0.0396, 0.0346, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 11:15:24,173 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186118.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:15:28,463 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186121.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:15:46,553 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-04-03 11:16:11,525 INFO [train.py:903] (3/4) Epoch 28, batch 1800, loss[loss=0.1855, simple_loss=0.2615, pruned_loss=0.05475, over 19413.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2829, pruned_loss=0.06033, over 3807196.65 frames. ], batch size: 48, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:16:13,506 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.74 vs. limit=2.0 +2023-04-03 11:16:18,384 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186161.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 11:16:36,642 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.127e+02 5.184e+02 6.048e+02 8.514e+02 1.613e+03, threshold=1.210e+03, percent-clipped=4.0 +2023-04-03 11:16:38,082 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186176.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:17:08,560 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 11:17:09,016 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186201.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:17:15,218 INFO [train.py:903] (3/4) Epoch 28, batch 1850, loss[loss=0.2028, simple_loss=0.2889, pruned_loss=0.05839, over 19761.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2833, pruned_loss=0.06067, over 3791630.46 frames. ], batch size: 54, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:17:39,918 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=186226.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:17:46,428 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 11:17:52,432 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186236.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:18:17,536 INFO [train.py:903] (3/4) Epoch 28, batch 1900, loss[loss=0.2109, simple_loss=0.2916, pruned_loss=0.06513, over 19664.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2821, pruned_loss=0.0597, over 3803411.47 frames. ], batch size: 55, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:18:33,594 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 11:18:38,417 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 11:18:39,533 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.824e+02 4.995e+02 5.845e+02 7.128e+02 1.193e+03, threshold=1.169e+03, percent-clipped=0.0 +2023-04-03 11:18:41,912 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186276.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 11:18:42,905 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186277.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:19:02,971 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 11:19:19,217 INFO [train.py:903] (3/4) Epoch 28, batch 1950, loss[loss=0.1595, simple_loss=0.2385, pruned_loss=0.04023, over 19376.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2829, pruned_loss=0.05979, over 3815212.39 frames. ], batch size: 47, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:20:20,369 INFO [train.py:903] (3/4) Epoch 28, batch 2000, loss[loss=0.189, simple_loss=0.2771, pruned_loss=0.05045, over 19597.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2836, pruned_loss=0.06009, over 3815466.76 frames. ], batch size: 57, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:20:45,108 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.192e+02 5.060e+02 6.550e+02 8.587e+02 3.446e+03, threshold=1.310e+03, percent-clipped=8.0 +2023-04-03 11:21:20,868 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 11:21:24,394 INFO [train.py:903] (3/4) Epoch 28, batch 2050, loss[loss=0.1942, simple_loss=0.2833, pruned_loss=0.0526, over 19522.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2818, pruned_loss=0.05933, over 3823683.62 frames. ], batch size: 54, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:21:40,368 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 11:21:42,531 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 11:22:02,198 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 11:22:04,665 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186439.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:22:07,925 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1992, 0.9728, 0.9726, 1.1460, 0.8480, 1.0281, 0.9506, 1.0994], + device='cuda:3'), covar=tensor([0.0975, 0.1175, 0.1331, 0.0871, 0.1217, 0.0572, 0.1373, 0.0760], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0364, 0.0323, 0.0260, 0.0310, 0.0260, 0.0325, 0.0268], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 11:22:26,599 INFO [train.py:903] (3/4) Epoch 28, batch 2100, loss[loss=0.1831, simple_loss=0.2677, pruned_loss=0.04925, over 19588.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2823, pruned_loss=0.05974, over 3816103.89 frames. ], batch size: 52, lr: 2.94e-03, grad_scale: 8.0 +2023-04-03 11:22:35,102 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186462.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:22:49,640 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.238e+02 5.089e+02 6.089e+02 7.390e+02 1.324e+03, threshold=1.218e+03, percent-clipped=1.0 +2023-04-03 11:22:50,535 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-04-03 11:22:58,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 11:23:11,818 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186492.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:23:13,818 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186493.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:23:20,625 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 11:23:29,672 INFO [train.py:903] (3/4) Epoch 28, batch 2150, loss[loss=0.1996, simple_loss=0.283, pruned_loss=0.05811, over 19656.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2822, pruned_loss=0.05972, over 3822323.44 frames. ], batch size: 60, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:23:30,063 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8956, 1.4729, 1.5522, 1.8063, 4.4084, 1.2264, 2.6148, 4.7996], + device='cuda:3'), covar=tensor([0.0538, 0.3104, 0.3300, 0.2057, 0.0769, 0.2805, 0.1496, 0.0187], + device='cuda:3'), in_proj_covar=tensor([0.0422, 0.0378, 0.0396, 0.0353, 0.0382, 0.0357, 0.0395, 0.0417], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:23:42,677 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=186517.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:23:46,781 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186520.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:24:02,531 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186532.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 11:24:06,091 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186535.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:24:30,565 INFO [train.py:903] (3/4) Epoch 28, batch 2200, loss[loss=0.1933, simple_loss=0.279, pruned_loss=0.05375, over 19474.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2828, pruned_loss=0.06012, over 3829199.24 frames. ], batch size: 64, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:24:32,121 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186557.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:24:32,226 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=186557.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 11:24:55,772 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.530e+02 4.864e+02 5.746e+02 7.619e+02 1.717e+03, threshold=1.149e+03, percent-clipped=3.0 +2023-04-03 11:24:58,717 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186577.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:25:05,219 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.30 vs. limit=5.0 +2023-04-03 11:25:09,426 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186585.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:25:15,865 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-04-03 11:25:32,821 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186604.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:25:34,931 INFO [train.py:903] (3/4) Epoch 28, batch 2250, loss[loss=0.1995, simple_loss=0.2746, pruned_loss=0.06214, over 19747.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2812, pruned_loss=0.05925, over 3839461.36 frames. ], batch size: 51, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:25:35,614 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-03 11:25:54,494 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186621.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:26:12,671 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186635.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:26:39,389 INFO [train.py:903] (3/4) Epoch 28, batch 2300, loss[loss=0.1896, simple_loss=0.2736, pruned_loss=0.0528, over 19659.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2811, pruned_loss=0.05943, over 3849272.73 frames. ], batch size: 53, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:26:55,447 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 11:27:02,265 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.323e+02 4.896e+02 5.750e+02 7.152e+02 2.246e+03, threshold=1.150e+03, percent-clipped=6.0 +2023-04-03 11:27:42,237 INFO [train.py:903] (3/4) Epoch 28, batch 2350, loss[loss=0.2093, simple_loss=0.2938, pruned_loss=0.06243, over 19590.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2813, pruned_loss=0.0594, over 3841501.33 frames. ], batch size: 61, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:28:07,896 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186727.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:28:20,558 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186736.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:28:26,047 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 11:28:42,190 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 11:28:44,640 INFO [train.py:903] (3/4) Epoch 28, batch 2400, loss[loss=0.1723, simple_loss=0.2635, pruned_loss=0.04055, over 19728.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2814, pruned_loss=0.05914, over 3841329.64 frames. ], batch size: 51, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:29:08,814 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.283e+02 4.439e+02 5.729e+02 7.466e+02 1.887e+03, threshold=1.146e+03, percent-clipped=9.0 +2023-04-03 11:29:20,269 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186783.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:29:47,728 INFO [train.py:903] (3/4) Epoch 28, batch 2450, loss[loss=0.1962, simple_loss=0.2856, pruned_loss=0.05337, over 19523.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2819, pruned_loss=0.05918, over 3830574.30 frames. ], batch size: 56, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:30:22,247 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186833.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:30:26,698 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186837.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:30:43,209 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7389, 1.5911, 1.5970, 2.1595, 1.5850, 1.9978, 2.0098, 1.8189], + device='cuda:3'), covar=tensor([0.0819, 0.0887, 0.1013, 0.0751, 0.0933, 0.0764, 0.0816, 0.0683], + device='cuda:3'), in_proj_covar=tensor([0.0211, 0.0222, 0.0226, 0.0239, 0.0225, 0.0212, 0.0186, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 11:30:48,876 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.1156, 2.8877, 2.3167, 2.2649, 2.0541, 2.4714, 1.1333, 2.1352], + device='cuda:3'), covar=tensor([0.0713, 0.0648, 0.0753, 0.1196, 0.1237, 0.1249, 0.1475, 0.1147], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0366, 0.0371, 0.0393, 0.0473, 0.0398, 0.0347, 0.0350], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 11:30:50,779 INFO [train.py:903] (3/4) Epoch 28, batch 2500, loss[loss=0.16, simple_loss=0.2427, pruned_loss=0.03864, over 19761.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.281, pruned_loss=0.05864, over 3830666.02 frames. ], batch size: 47, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:30:54,437 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186858.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:30:54,553 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=186858.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:31:15,068 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.341e+02 4.860e+02 5.866e+02 7.110e+02 2.029e+03, threshold=1.173e+03, percent-clipped=7.0 +2023-04-03 11:31:19,830 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186879.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:31:35,109 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186891.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:31:45,080 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186898.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:31:49,213 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186901.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:31:54,930 INFO [train.py:903] (3/4) Epoch 28, batch 2550, loss[loss=0.2143, simple_loss=0.2944, pruned_loss=0.06711, over 19670.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2809, pruned_loss=0.05882, over 3819499.78 frames. ], batch size: 60, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:32:06,745 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=186916.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:32:22,505 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186929.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:32:22,651 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=186929.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:32:47,706 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=186948.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:32:50,818 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 11:32:52,343 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186952.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:32:56,644 INFO [train.py:903] (3/4) Epoch 28, batch 2600, loss[loss=0.1958, simple_loss=0.2792, pruned_loss=0.0562, over 19671.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2815, pruned_loss=0.05963, over 3809152.47 frames. ], batch size: 55, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:33:20,687 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.371e+02 5.127e+02 5.898e+02 7.788e+02 1.720e+03, threshold=1.180e+03, percent-clipped=7.0 +2023-04-03 11:33:38,273 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4684, 1.2548, 1.3866, 1.4566, 2.2325, 1.2171, 1.9624, 2.4546], + device='cuda:3'), covar=tensor([0.0557, 0.2232, 0.2191, 0.1387, 0.0631, 0.1822, 0.1615, 0.0435], + device='cuda:3'), in_proj_covar=tensor([0.0421, 0.0377, 0.0396, 0.0352, 0.0382, 0.0357, 0.0394, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:33:43,121 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=186992.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:33:45,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=186994.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:33:59,847 INFO [train.py:903] (3/4) Epoch 28, batch 2650, loss[loss=0.1755, simple_loss=0.2535, pruned_loss=0.04878, over 14737.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2823, pruned_loss=0.06016, over 3795144.57 frames. ], batch size: 32, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:34:12,812 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187016.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:34:14,011 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187017.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:34:22,402 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 11:34:46,307 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187043.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:34:47,571 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187044.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:35:02,638 INFO [train.py:903] (3/4) Epoch 28, batch 2700, loss[loss=0.201, simple_loss=0.284, pruned_loss=0.05902, over 19589.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2832, pruned_loss=0.06039, over 3806515.51 frames. ], batch size: 61, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:35:02,900 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187056.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:35:13,332 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187063.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:35:22,297 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187071.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:35:26,784 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.432e+02 4.685e+02 5.845e+02 7.450e+02 1.306e+03, threshold=1.169e+03, percent-clipped=4.0 +2023-04-03 11:36:06,485 INFO [train.py:903] (3/4) Epoch 28, batch 2750, loss[loss=0.1757, simple_loss=0.2518, pruned_loss=0.04981, over 19762.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2827, pruned_loss=0.05994, over 3823989.10 frames. ], batch size: 47, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:37:06,454 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187154.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:37:08,435 INFO [train.py:903] (3/4) Epoch 28, batch 2800, loss[loss=0.1764, simple_loss=0.2497, pruned_loss=0.05154, over 19725.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2818, pruned_loss=0.05967, over 3824791.45 frames. ], batch size: 46, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:37:31,617 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.065e+02 4.906e+02 5.640e+02 7.444e+02 1.445e+03, threshold=1.128e+03, percent-clipped=1.0 +2023-04-03 11:37:36,920 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187179.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:37:46,960 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187186.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:38:05,782 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187202.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:38:10,182 INFO [train.py:903] (3/4) Epoch 28, batch 2850, loss[loss=0.2543, simple_loss=0.3151, pruned_loss=0.09671, over 19773.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2821, pruned_loss=0.05995, over 3827114.42 frames. ], batch size: 54, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:38:13,801 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187208.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:38:44,982 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187233.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:38:46,482 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-04-03 11:39:05,636 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187250.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:39:11,665 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 11:39:12,697 INFO [train.py:903] (3/4) Epoch 28, batch 2900, loss[loss=0.2166, simple_loss=0.3105, pruned_loss=0.06134, over 19672.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.281, pruned_loss=0.05933, over 3826844.58 frames. ], batch size: 58, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:39:32,591 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187272.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:39:33,456 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187273.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:39:35,532 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.364e+02 5.012e+02 6.172e+02 7.442e+02 2.226e+03, threshold=1.234e+03, percent-clipped=8.0 +2023-04-03 11:39:35,980 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187275.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:39:54,724 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.88 vs. limit=5.0 +2023-04-03 11:40:02,602 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187297.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:40:06,121 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187300.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:40:13,518 INFO [train.py:903] (3/4) Epoch 28, batch 2950, loss[loss=0.2089, simple_loss=0.2982, pruned_loss=0.05983, over 19656.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2822, pruned_loss=0.06013, over 3808857.69 frames. ], batch size: 58, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:40:26,542 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187317.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:40:28,985 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187319.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:40:35,752 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187325.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:40:59,985 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187344.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:41:13,833 INFO [train.py:903] (3/4) Epoch 28, batch 3000, loss[loss=0.2503, simple_loss=0.3177, pruned_loss=0.09141, over 12839.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2836, pruned_loss=0.06104, over 3802285.03 frames. ], batch size: 136, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:41:13,833 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 11:41:26,709 INFO [train.py:937] (3/4) Epoch 28, validation: loss=0.1673, simple_loss=0.2667, pruned_loss=0.03394, over 944034.00 frames. +2023-04-03 11:41:26,709 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 11:41:29,122 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 11:41:35,521 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1050, 2.2069, 2.4216, 2.7815, 2.1245, 2.6903, 2.4310, 2.2168], + device='cuda:3'), covar=tensor([0.4325, 0.4185, 0.2060, 0.2366, 0.4246, 0.2233, 0.5139, 0.3580], + device='cuda:3'), in_proj_covar=tensor([0.0935, 0.1013, 0.0742, 0.0949, 0.0911, 0.0853, 0.0861, 0.0809], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 11:41:41,220 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0922, 1.6892, 1.9686, 2.0942, 4.6181, 1.3320, 2.8425, 5.1074], + device='cuda:3'), covar=tensor([0.0478, 0.2877, 0.2809, 0.1840, 0.0742, 0.2709, 0.1384, 0.0167], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0381, 0.0401, 0.0355, 0.0385, 0.0361, 0.0399, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:41:49,225 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.693e+02 4.898e+02 6.411e+02 7.995e+02 1.373e+03, threshold=1.282e+03, percent-clipped=5.0 +2023-04-03 11:42:05,148 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187387.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:42:06,466 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187388.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:42:08,740 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187390.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:42:21,160 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187400.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:42:27,782 INFO [train.py:903] (3/4) Epoch 28, batch 3050, loss[loss=0.2172, simple_loss=0.3002, pruned_loss=0.06704, over 19296.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2825, pruned_loss=0.06039, over 3798283.49 frames. ], batch size: 66, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:42:56,177 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-04-03 11:43:13,143 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187442.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:43:29,503 INFO [train.py:903] (3/4) Epoch 28, batch 3100, loss[loss=0.1837, simple_loss=0.2694, pruned_loss=0.04895, over 19577.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2816, pruned_loss=0.05987, over 3810722.76 frames. ], batch size: 52, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:43:43,950 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187467.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:43:54,450 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.395e+02 4.792e+02 5.797e+02 7.675e+02 1.223e+03, threshold=1.159e+03, percent-clipped=0.0 +2023-04-03 11:44:27,074 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187502.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:44:32,164 INFO [train.py:903] (3/4) Epoch 28, batch 3150, loss[loss=0.2215, simple_loss=0.296, pruned_loss=0.07351, over 13072.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2813, pruned_loss=0.05944, over 3815595.45 frames. ], batch size: 136, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:44:44,944 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187515.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:44:54,641 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 11:45:01,662 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187530.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:45:06,302 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7663, 4.3193, 4.4973, 4.5082, 1.8200, 4.2717, 3.6743, 4.2458], + device='cuda:3'), covar=tensor([0.1834, 0.0900, 0.0648, 0.0728, 0.6004, 0.1040, 0.0755, 0.1146], + device='cuda:3'), in_proj_covar=tensor([0.0829, 0.0792, 0.1004, 0.0881, 0.0867, 0.0769, 0.0594, 0.0932], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 11:45:10,628 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187537.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:45:15,263 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=187541.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:45:35,285 INFO [train.py:903] (3/4) Epoch 28, batch 3200, loss[loss=0.2324, simple_loss=0.3081, pruned_loss=0.07833, over 17203.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2821, pruned_loss=0.05957, over 3819099.53 frames. ], batch size: 101, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:45:35,960 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 11:45:55,556 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187573.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:45:57,382 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.995e+02 4.914e+02 6.228e+02 8.001e+02 2.182e+03, threshold=1.246e+03, percent-clipped=10.0 +2023-04-03 11:46:27,908 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187598.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:46:36,793 INFO [train.py:903] (3/4) Epoch 28, batch 3250, loss[loss=0.1747, simple_loss=0.257, pruned_loss=0.04617, over 19613.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2827, pruned_loss=0.05973, over 3828023.92 frames. ], batch size: 50, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:46:55,616 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 11:47:22,872 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.09 vs. limit=5.0 +2023-04-03 11:47:23,903 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187644.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:47:30,157 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.41 vs. limit=2.0 +2023-04-03 11:47:37,435 INFO [train.py:903] (3/4) Epoch 28, batch 3300, loss[loss=0.2179, simple_loss=0.2963, pruned_loss=0.06977, over 19578.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2834, pruned_loss=0.06047, over 3828162.07 frames. ], batch size: 61, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:47:37,453 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 11:47:54,269 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187669.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:47:56,603 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0369, 1.8510, 1.6793, 2.0465, 1.6716, 1.7025, 1.6401, 1.9242], + device='cuda:3'), covar=tensor([0.1063, 0.1503, 0.1527, 0.1054, 0.1480, 0.0622, 0.1592, 0.0786], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0360, 0.0319, 0.0257, 0.0306, 0.0256, 0.0322, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:47:57,868 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4035, 2.3519, 2.2654, 2.5576, 2.3792, 2.0485, 2.1707, 2.3915], + device='cuda:3'), covar=tensor([0.0852, 0.1250, 0.1194, 0.0829, 0.1084, 0.0508, 0.1225, 0.0596], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0360, 0.0319, 0.0257, 0.0306, 0.0256, 0.0322, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:48:01,982 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.234e+02 4.770e+02 6.093e+02 7.830e+02 1.620e+03, threshold=1.219e+03, percent-clipped=4.0 +2023-04-03 11:48:08,592 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.11 vs. limit=5.0 +2023-04-03 11:48:10,620 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 11:48:40,829 INFO [train.py:903] (3/4) Epoch 28, batch 3350, loss[loss=0.205, simple_loss=0.2942, pruned_loss=0.05787, over 19605.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2836, pruned_loss=0.06043, over 3826403.34 frames. ], batch size: 57, lr: 2.93e-03, grad_scale: 8.0 +2023-04-03 11:49:14,752 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187734.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:49:42,975 INFO [train.py:903] (3/4) Epoch 28, batch 3400, loss[loss=0.2131, simple_loss=0.3, pruned_loss=0.06312, over 19676.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2832, pruned_loss=0.06023, over 3837954.78 frames. ], batch size: 59, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:49:45,557 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187758.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:49:54,718 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1861, 0.9010, 0.9685, 1.1309, 0.8183, 1.0048, 0.9192, 1.0776], + device='cuda:3'), covar=tensor([0.0832, 0.1111, 0.1114, 0.0733, 0.1069, 0.0527, 0.1236, 0.0661], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0359, 0.0318, 0.0255, 0.0304, 0.0255, 0.0320, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:50:02,064 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=187771.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:50:07,730 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.151e+02 4.801e+02 5.865e+02 7.575e+02 1.695e+03, threshold=1.173e+03, percent-clipped=1.0 +2023-04-03 11:50:16,882 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187783.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:50:34,172 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=187796.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:50:46,600 INFO [train.py:903] (3/4) Epoch 28, batch 3450, loss[loss=0.2028, simple_loss=0.2979, pruned_loss=0.05387, over 19569.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2828, pruned_loss=0.05988, over 3830388.30 frames. ], batch size: 61, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:50:47,821 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 11:51:41,401 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187849.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:51:49,297 INFO [train.py:903] (3/4) Epoch 28, batch 3500, loss[loss=0.1972, simple_loss=0.2789, pruned_loss=0.05777, over 19549.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2832, pruned_loss=0.06008, over 3815867.30 frames. ], batch size: 52, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:52:12,001 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187874.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:52:14,010 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.602e+02 4.936e+02 6.004e+02 7.154e+02 1.224e+03, threshold=1.201e+03, percent-clipped=1.0 +2023-04-03 11:52:20,877 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187881.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:52:26,772 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=187885.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:52:50,771 INFO [train.py:903] (3/4) Epoch 28, batch 3550, loss[loss=0.2055, simple_loss=0.2868, pruned_loss=0.06205, over 19694.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.284, pruned_loss=0.06059, over 3820189.66 frames. ], batch size: 59, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:53:52,892 INFO [train.py:903] (3/4) Epoch 28, batch 3600, loss[loss=0.1986, simple_loss=0.2847, pruned_loss=0.05625, over 18701.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.2847, pruned_loss=0.06074, over 3811970.49 frames. ], batch size: 74, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 11:54:17,598 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.816e+02 4.904e+02 5.870e+02 7.567e+02 1.667e+03, threshold=1.174e+03, percent-clipped=3.0 +2023-04-03 11:54:34,351 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187989.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:54:42,622 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=187996.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:54:48,323 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=188000.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:54:56,060 INFO [train.py:903] (3/4) Epoch 28, batch 3650, loss[loss=0.1986, simple_loss=0.2846, pruned_loss=0.05629, over 19310.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2855, pruned_loss=0.06078, over 3801007.75 frames. ], batch size: 70, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 11:55:01,222 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.09 vs. limit=2.0 +2023-04-03 11:55:37,616 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1070, 2.1778, 2.4688, 2.6932, 2.1179, 2.6277, 2.4578, 2.2665], + device='cuda:3'), covar=tensor([0.4220, 0.4120, 0.1970, 0.2499, 0.4431, 0.2386, 0.4883, 0.3495], + device='cuda:3'), in_proj_covar=tensor([0.0940, 0.1017, 0.0745, 0.0954, 0.0914, 0.0858, 0.0865, 0.0812], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 11:55:57,938 INFO [train.py:903] (3/4) Epoch 28, batch 3700, loss[loss=0.2124, simple_loss=0.2979, pruned_loss=0.0635, over 19616.00 frames. ], tot_loss[loss=0.2036, simple_loss=0.2849, pruned_loss=0.06112, over 3804874.36 frames. ], batch size: 57, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 11:56:10,322 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-04-03 11:56:23,996 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.791e+02 5.013e+02 6.013e+02 7.793e+02 2.143e+03, threshold=1.203e+03, percent-clipped=3.0 +2023-04-03 11:56:59,875 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=188105.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:57:00,592 INFO [train.py:903] (3/4) Epoch 28, batch 3750, loss[loss=0.175, simple_loss=0.2613, pruned_loss=0.04436, over 19664.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2854, pruned_loss=0.06115, over 3800626.55 frames. ], batch size: 53, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:57:32,241 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=188130.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 11:58:03,790 INFO [train.py:903] (3/4) Epoch 28, batch 3800, loss[loss=0.1989, simple_loss=0.287, pruned_loss=0.05546, over 19492.00 frames. ], tot_loss[loss=0.2038, simple_loss=0.2853, pruned_loss=0.0612, over 3798920.51 frames. ], batch size: 64, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:58:28,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3783, 3.1026, 2.2166, 2.7420, 0.7511, 3.0714, 2.9106, 3.0296], + device='cuda:3'), covar=tensor([0.1024, 0.1233, 0.2065, 0.1084, 0.3930, 0.0935, 0.1203, 0.1456], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0430, 0.0515, 0.0361, 0.0409, 0.0456, 0.0452, 0.0484], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 11:58:30,486 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.894e+02 5.177e+02 6.017e+02 8.386e+02 1.721e+03, threshold=1.203e+03, percent-clipped=7.0 +2023-04-03 11:58:33,986 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 11:59:06,272 INFO [train.py:903] (3/4) Epoch 28, batch 3850, loss[loss=0.2048, simple_loss=0.2906, pruned_loss=0.05952, over 19724.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2843, pruned_loss=0.06082, over 3792584.02 frames. ], batch size: 63, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 11:59:35,001 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2757, 2.1119, 2.0074, 1.8381, 1.6350, 1.7911, 0.6079, 1.3052], + device='cuda:3'), covar=tensor([0.0687, 0.0724, 0.0579, 0.0979, 0.1304, 0.1118, 0.1491, 0.1197], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0367, 0.0374, 0.0395, 0.0475, 0.0400, 0.0347, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 11:59:56,134 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=188245.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:00:05,189 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=188252.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:00:09,176 INFO [train.py:903] (3/4) Epoch 28, batch 3900, loss[loss=0.182, simple_loss=0.2554, pruned_loss=0.05426, over 19772.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2825, pruned_loss=0.05988, over 3805716.34 frames. ], batch size: 47, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 12:00:09,615 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=188256.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:00:26,398 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=188270.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:00:34,023 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.127e+02 5.053e+02 6.264e+02 8.116e+02 1.975e+03, threshold=1.253e+03, percent-clipped=4.0 +2023-04-03 12:00:34,442 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=188277.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:00:39,114 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=188281.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:01:04,086 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=188301.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:01:09,681 INFO [train.py:903] (3/4) Epoch 28, batch 3950, loss[loss=0.2208, simple_loss=0.2953, pruned_loss=0.0732, over 19686.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2826, pruned_loss=0.06002, over 3818093.38 frames. ], batch size: 53, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 12:01:15,489 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 12:01:53,022 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-03 12:02:01,855 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6017, 2.1547, 1.6660, 1.6031, 2.0584, 1.4853, 1.5606, 1.9570], + device='cuda:3'), covar=tensor([0.1012, 0.0791, 0.0995, 0.0829, 0.0510, 0.1167, 0.0683, 0.0480], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0319, 0.0342, 0.0272, 0.0252, 0.0345, 0.0293, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:02:12,178 INFO [train.py:903] (3/4) Epoch 28, batch 4000, loss[loss=0.2087, simple_loss=0.2907, pruned_loss=0.06336, over 19731.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2824, pruned_loss=0.05977, over 3814187.90 frames. ], batch size: 63, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:02:38,279 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.378e+02 4.980e+02 6.272e+02 8.090e+02 1.579e+03, threshold=1.254e+03, percent-clipped=5.0 +2023-04-03 12:02:59,851 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 12:03:14,916 INFO [train.py:903] (3/4) Epoch 28, batch 4050, loss[loss=0.2802, simple_loss=0.3514, pruned_loss=0.1045, over 19496.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2817, pruned_loss=0.05962, over 3808564.14 frames. ], batch size: 64, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:03:40,257 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3731, 3.9667, 2.5407, 3.5144, 0.8113, 3.9908, 3.8047, 3.9298], + device='cuda:3'), covar=tensor([0.0711, 0.1094, 0.2063, 0.0961, 0.4243, 0.0702, 0.0947, 0.1296], + device='cuda:3'), in_proj_covar=tensor([0.0531, 0.0433, 0.0519, 0.0363, 0.0413, 0.0459, 0.0456, 0.0486], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:04:05,447 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5479, 1.5839, 1.6968, 1.7554, 2.3099, 2.1976, 2.3048, 0.8921], + device='cuda:3'), covar=tensor([0.2458, 0.4408, 0.2844, 0.1989, 0.1550, 0.2216, 0.1472, 0.4993], + device='cuda:3'), in_proj_covar=tensor([0.0554, 0.0670, 0.0755, 0.0508, 0.0635, 0.0547, 0.0670, 0.0571], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:04:17,049 INFO [train.py:903] (3/4) Epoch 28, batch 4100, loss[loss=0.2115, simple_loss=0.3023, pruned_loss=0.0604, over 19745.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2814, pruned_loss=0.05922, over 3821300.78 frames. ], batch size: 63, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:04:43,426 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.476e+02 5.116e+02 6.170e+02 8.694e+02 1.686e+03, threshold=1.234e+03, percent-clipped=5.0 +2023-04-03 12:04:53,754 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 12:05:11,248 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=188499.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:05:19,164 INFO [train.py:903] (3/4) Epoch 28, batch 4150, loss[loss=0.1975, simple_loss=0.285, pruned_loss=0.055, over 19658.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2824, pruned_loss=0.05973, over 3818116.49 frames. ], batch size: 55, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:05:38,322 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6475, 1.8358, 2.1628, 1.9926, 3.2142, 2.8272, 3.5021, 1.6929], + device='cuda:3'), covar=tensor([0.2612, 0.4532, 0.2987, 0.1957, 0.1538, 0.2108, 0.1627, 0.4675], + device='cuda:3'), in_proj_covar=tensor([0.0553, 0.0668, 0.0754, 0.0508, 0.0633, 0.0546, 0.0669, 0.0570], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 12:06:02,309 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1246, 3.3818, 1.9624, 2.0767, 3.0979, 1.8082, 1.5818, 2.3725], + device='cuda:3'), covar=tensor([0.1367, 0.0691, 0.1109, 0.0912, 0.0527, 0.1198, 0.0956, 0.0622], + device='cuda:3'), in_proj_covar=tensor([0.0302, 0.0317, 0.0341, 0.0271, 0.0251, 0.0344, 0.0292, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:06:03,686 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.28 vs. limit=5.0 +2023-04-03 12:06:21,496 INFO [train.py:903] (3/4) Epoch 28, batch 4200, loss[loss=0.21, simple_loss=0.2833, pruned_loss=0.06833, over 19849.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2826, pruned_loss=0.05966, over 3835194.13 frames. ], batch size: 52, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:06:24,953 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 12:06:35,212 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9913, 1.9359, 1.6434, 2.0831, 1.7975, 1.6899, 1.6331, 1.8901], + device='cuda:3'), covar=tensor([0.1098, 0.1429, 0.1594, 0.1056, 0.1437, 0.0602, 0.1584, 0.0794], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0359, 0.0318, 0.0256, 0.0304, 0.0256, 0.0320, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:06:46,549 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.847e+02 4.775e+02 5.885e+02 7.540e+02 1.202e+03, threshold=1.177e+03, percent-clipped=0.0 +2023-04-03 12:07:22,472 INFO [train.py:903] (3/4) Epoch 28, batch 4250, loss[loss=0.21, simple_loss=0.2897, pruned_loss=0.0651, over 19177.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2824, pruned_loss=0.06003, over 3844235.76 frames. ], batch size: 69, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 12:07:33,738 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 12:07:44,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 12:08:10,317 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=188645.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:08:12,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4813, 1.5329, 1.8943, 1.7635, 2.6445, 2.3563, 2.8524, 1.2217], + device='cuda:3'), covar=tensor([0.2656, 0.4582, 0.2918, 0.2073, 0.1729, 0.2301, 0.1608, 0.4928], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0672, 0.0757, 0.0509, 0.0637, 0.0549, 0.0671, 0.0573], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:08:24,249 INFO [train.py:903] (3/4) Epoch 28, batch 4300, loss[loss=0.2011, simple_loss=0.2718, pruned_loss=0.06524, over 19482.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2816, pruned_loss=0.05971, over 3832345.72 frames. ], batch size: 49, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 12:08:43,058 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7961, 1.5767, 1.4558, 1.7908, 1.4614, 1.5411, 1.4517, 1.6725], + device='cuda:3'), covar=tensor([0.1165, 0.1343, 0.1660, 0.1115, 0.1406, 0.0637, 0.1648, 0.0890], + device='cuda:3'), in_proj_covar=tensor([0.0276, 0.0358, 0.0317, 0.0255, 0.0304, 0.0255, 0.0319, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:08:45,380 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-03 12:08:51,756 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.988e+02 4.883e+02 5.889e+02 8.258e+02 1.553e+03, threshold=1.178e+03, percent-clipped=6.0 +2023-04-03 12:09:15,642 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 12:09:26,307 INFO [train.py:903] (3/4) Epoch 28, batch 4350, loss[loss=0.2044, simple_loss=0.2887, pruned_loss=0.06008, over 19598.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2824, pruned_loss=0.06024, over 3819093.71 frames. ], batch size: 57, lr: 2.92e-03, grad_scale: 4.0 +2023-04-03 12:10:30,035 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 12:10:30,423 INFO [train.py:903] (3/4) Epoch 28, batch 4400, loss[loss=0.2317, simple_loss=0.3074, pruned_loss=0.07797, over 13845.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2821, pruned_loss=0.05996, over 3816593.66 frames. ], batch size: 136, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:10:35,412 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=188760.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:10:45,485 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7489, 4.3574, 2.9392, 3.8709, 1.6600, 4.3385, 4.1910, 4.3553], + device='cuda:3'), covar=tensor([0.0563, 0.0814, 0.1772, 0.0791, 0.3131, 0.0659, 0.0839, 0.1191], + device='cuda:3'), in_proj_covar=tensor([0.0525, 0.0428, 0.0513, 0.0361, 0.0409, 0.0455, 0.0451, 0.0481], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:10:45,704 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2000, 2.1471, 2.0368, 1.8833, 1.6966, 1.8931, 0.8245, 1.4171], + device='cuda:3'), covar=tensor([0.0631, 0.0664, 0.0489, 0.0865, 0.1068, 0.1004, 0.1392, 0.1081], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0368, 0.0375, 0.0397, 0.0477, 0.0401, 0.0349, 0.0353], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 12:10:52,235 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 12:10:57,775 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.415e+02 4.797e+02 5.749e+02 7.341e+02 1.454e+03, threshold=1.150e+03, percent-clipped=2.0 +2023-04-03 12:11:02,386 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 12:11:08,439 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0944, 1.2936, 1.8097, 1.3174, 2.7309, 3.6689, 3.3649, 3.8179], + device='cuda:3'), covar=tensor([0.1784, 0.4021, 0.3494, 0.2644, 0.0630, 0.0197, 0.0207, 0.0280], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0332, 0.0365, 0.0272, 0.0256, 0.0198, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 12:11:31,836 INFO [train.py:903] (3/4) Epoch 28, batch 4450, loss[loss=0.1984, simple_loss=0.2791, pruned_loss=0.05883, over 19687.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2814, pruned_loss=0.05915, over 3828430.39 frames. ], batch size: 53, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:12:19,636 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=188843.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:12:36,107 INFO [train.py:903] (3/4) Epoch 28, batch 4500, loss[loss=0.1974, simple_loss=0.2833, pruned_loss=0.05577, over 19627.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2808, pruned_loss=0.05876, over 3834170.75 frames. ], batch size: 57, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:13:04,411 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.382e+02 4.536e+02 5.549e+02 7.562e+02 1.666e+03, threshold=1.110e+03, percent-clipped=5.0 +2023-04-03 12:13:38,222 INFO [train.py:903] (3/4) Epoch 28, batch 4550, loss[loss=0.2363, simple_loss=0.3134, pruned_loss=0.07963, over 13798.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.281, pruned_loss=0.05895, over 3835151.75 frames. ], batch size: 136, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:13:46,946 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 12:14:12,230 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 12:14:29,480 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=188947.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:14:31,867 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=188949.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:14:39,519 INFO [train.py:903] (3/4) Epoch 28, batch 4600, loss[loss=0.2064, simple_loss=0.2904, pruned_loss=0.06119, over 19314.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2822, pruned_loss=0.05941, over 3820339.91 frames. ], batch size: 66, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:14:43,965 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=188958.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:15:07,936 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.099e+02 4.624e+02 5.779e+02 7.629e+02 1.899e+03, threshold=1.156e+03, percent-clipped=6.0 +2023-04-03 12:15:43,021 INFO [train.py:903] (3/4) Epoch 28, batch 4650, loss[loss=0.2473, simple_loss=0.3133, pruned_loss=0.09062, over 12796.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06023, over 3821932.86 frames. ], batch size: 136, lr: 2.92e-03, grad_scale: 8.0 +2023-04-03 12:15:55,988 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189016.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:16:01,293 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 12:16:12,603 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 12:16:26,493 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189041.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:16:34,757 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189048.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:16:44,449 INFO [train.py:903] (3/4) Epoch 28, batch 4700, loss[loss=0.269, simple_loss=0.3418, pruned_loss=0.0981, over 19681.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2836, pruned_loss=0.06019, over 3839143.82 frames. ], batch size: 59, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:16:47,419 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.77 vs. limit=5.0 +2023-04-03 12:17:07,720 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 12:17:10,868 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.010e+02 4.848e+02 5.842e+02 7.352e+02 2.015e+03, threshold=1.168e+03, percent-clipped=3.0 +2023-04-03 12:17:17,571 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189082.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:17:36,200 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189097.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:17:43,645 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1222, 1.3609, 1.7033, 1.0423, 2.3294, 3.0903, 2.7830, 3.2289], + device='cuda:3'), covar=tensor([0.1665, 0.3847, 0.3463, 0.2752, 0.0688, 0.0224, 0.0273, 0.0362], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0333, 0.0365, 0.0272, 0.0256, 0.0198, 0.0221, 0.0280], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 12:17:46,802 INFO [train.py:903] (3/4) Epoch 28, batch 4750, loss[loss=0.1918, simple_loss=0.2816, pruned_loss=0.051, over 19575.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2824, pruned_loss=0.05933, over 3828001.67 frames. ], batch size: 61, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:17:56,578 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-04-03 12:18:17,602 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.2255, 2.3356, 2.3106, 2.0605, 4.7611, 1.6141, 3.0093, 5.2817], + device='cuda:3'), covar=tensor([0.0397, 0.2323, 0.2625, 0.1977, 0.0678, 0.2551, 0.1314, 0.0135], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0381, 0.0400, 0.0354, 0.0383, 0.0360, 0.0399, 0.0421], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:18:22,692 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.44 vs. limit=2.0 +2023-04-03 12:18:47,826 INFO [train.py:903] (3/4) Epoch 28, batch 4800, loss[loss=0.194, simple_loss=0.2752, pruned_loss=0.05647, over 19791.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2829, pruned_loss=0.05962, over 3835913.75 frames. ], batch size: 48, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:19:16,003 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.214e+02 5.029e+02 6.156e+02 7.557e+02 1.439e+03, threshold=1.231e+03, percent-clipped=4.0 +2023-04-03 12:19:50,824 INFO [train.py:903] (3/4) Epoch 28, batch 4850, loss[loss=0.2107, simple_loss=0.2976, pruned_loss=0.06188, over 19159.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2843, pruned_loss=0.06039, over 3826864.11 frames. ], batch size: 75, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:20:01,566 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189214.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:20:12,580 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 12:20:32,918 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189239.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:20:33,799 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 12:20:39,520 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 12:20:40,696 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 12:20:51,042 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 12:20:53,316 INFO [train.py:903] (3/4) Epoch 28, batch 4900, loss[loss=0.1689, simple_loss=0.2517, pruned_loss=0.04304, over 19416.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2834, pruned_loss=0.05969, over 3841132.90 frames. ], batch size: 48, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:21:05,668 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189266.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:21:09,132 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3095, 1.2930, 1.4762, 1.4646, 1.8424, 1.8503, 1.8679, 0.6615], + device='cuda:3'), covar=tensor([0.2634, 0.4558, 0.2801, 0.2109, 0.1663, 0.2448, 0.1428, 0.5177], + device='cuda:3'), in_proj_covar=tensor([0.0558, 0.0675, 0.0760, 0.0511, 0.0639, 0.0551, 0.0675, 0.0577], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:21:10,805 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 12:21:19,662 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.182e+02 4.742e+02 6.083e+02 7.635e+02 1.565e+03, threshold=1.217e+03, percent-clipped=2.0 +2023-04-03 12:21:36,051 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189291.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:21:38,173 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189293.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:21:53,516 INFO [train.py:903] (3/4) Epoch 28, batch 4950, loss[loss=0.2054, simple_loss=0.2923, pruned_loss=0.05928, over 19768.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2838, pruned_loss=0.06031, over 3842180.25 frames. ], batch size: 56, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:22:10,588 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 12:22:34,588 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 12:22:49,563 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5126, 1.5245, 1.7995, 1.7267, 2.5783, 2.3116, 2.7839, 1.2101], + device='cuda:3'), covar=tensor([0.2645, 0.4574, 0.2980, 0.2038, 0.1749, 0.2280, 0.1672, 0.4927], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0674, 0.0758, 0.0509, 0.0637, 0.0548, 0.0673, 0.0575], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:22:56,069 INFO [train.py:903] (3/4) Epoch 28, batch 5000, loss[loss=0.2415, simple_loss=0.3138, pruned_loss=0.08461, over 13389.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2836, pruned_loss=0.06038, over 3829747.03 frames. ], batch size: 136, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:23:04,849 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 12:23:08,787 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189365.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:23:11,198 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4487, 1.1727, 1.3512, 1.3020, 2.2094, 1.1051, 1.9033, 2.3983], + device='cuda:3'), covar=tensor([0.0534, 0.2203, 0.2261, 0.1506, 0.0594, 0.1938, 0.1719, 0.0435], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0380, 0.0399, 0.0353, 0.0382, 0.0358, 0.0398, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:23:15,677 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 12:23:24,840 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.397e+02 5.003e+02 6.019e+02 8.012e+02 1.544e+03, threshold=1.204e+03, percent-clipped=7.0 +2023-04-03 12:23:42,173 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189392.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:23:59,713 INFO [train.py:903] (3/4) Epoch 28, batch 5050, loss[loss=0.1958, simple_loss=0.2778, pruned_loss=0.05687, over 19336.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2844, pruned_loss=0.06123, over 3813322.80 frames. ], batch size: 66, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:24:00,182 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189406.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:24:02,612 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189408.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:24:24,216 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189426.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:24:26,723 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7389, 2.7850, 2.4435, 2.8729, 2.6920, 2.3477, 2.2340, 2.7332], + device='cuda:3'), covar=tensor([0.0923, 0.1321, 0.1366, 0.0959, 0.1289, 0.0513, 0.1464, 0.0631], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0359, 0.0318, 0.0256, 0.0304, 0.0255, 0.0321, 0.0262], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:24:33,170 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 12:24:44,047 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189441.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:25:01,203 INFO [train.py:903] (3/4) Epoch 28, batch 5100, loss[loss=0.1686, simple_loss=0.2576, pruned_loss=0.03983, over 19819.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2839, pruned_loss=0.06088, over 3814522.40 frames. ], batch size: 52, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:25:10,105 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 12:25:12,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 12:25:19,115 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 12:25:27,910 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.508e+02 5.191e+02 6.540e+02 8.236e+02 1.634e+03, threshold=1.308e+03, percent-clipped=9.0 +2023-04-03 12:26:01,392 INFO [train.py:903] (3/4) Epoch 28, batch 5150, loss[loss=0.2153, simple_loss=0.2999, pruned_loss=0.06539, over 19622.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2821, pruned_loss=0.05983, over 3826162.90 frames. ], batch size: 57, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:26:02,925 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189507.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:26:11,046 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-04-03 12:26:11,419 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 12:26:18,172 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189519.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:26:44,896 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189541.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:26:46,846 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 12:27:02,654 INFO [train.py:903] (3/4) Epoch 28, batch 5200, loss[loss=0.2061, simple_loss=0.2788, pruned_loss=0.06672, over 19493.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2812, pruned_loss=0.05941, over 3822819.13 frames. ], batch size: 49, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:27:03,050 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189556.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:27:12,489 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-03 12:27:17,234 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 12:27:27,472 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0190, 1.8524, 1.6609, 2.0173, 1.7830, 1.7175, 1.7370, 1.8898], + device='cuda:3'), covar=tensor([0.1088, 0.1510, 0.1569, 0.1048, 0.1356, 0.0601, 0.1425, 0.0779], + device='cuda:3'), in_proj_covar=tensor([0.0277, 0.0358, 0.0317, 0.0255, 0.0303, 0.0254, 0.0320, 0.0261], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:27:30,434 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.217e+02 4.850e+02 5.941e+02 7.550e+02 1.552e+03, threshold=1.188e+03, percent-clipped=2.0 +2023-04-03 12:27:42,769 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-04-03 12:28:01,601 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 12:28:04,859 INFO [train.py:903] (3/4) Epoch 28, batch 5250, loss[loss=0.2092, simple_loss=0.2849, pruned_loss=0.06674, over 16010.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.282, pruned_loss=0.05964, over 3819324.96 frames. ], batch size: 35, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:28:09,622 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189610.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:28:20,325 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.07 vs. limit=5.0 +2023-04-03 12:28:53,302 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2770, 5.7260, 3.3249, 5.0478, 1.1952, 5.8659, 5.7183, 5.9146], + device='cuda:3'), covar=tensor([0.0357, 0.0951, 0.1752, 0.0777, 0.4146, 0.0619, 0.0781, 0.1030], + device='cuda:3'), in_proj_covar=tensor([0.0529, 0.0430, 0.0516, 0.0361, 0.0409, 0.0456, 0.0451, 0.0482], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:29:01,918 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5798, 4.1118, 4.2873, 4.2921, 1.7272, 4.0815, 3.5332, 4.0627], + device='cuda:3'), covar=tensor([0.1831, 0.0897, 0.0648, 0.0766, 0.6206, 0.0968, 0.0769, 0.1124], + device='cuda:3'), in_proj_covar=tensor([0.0816, 0.0783, 0.0992, 0.0871, 0.0862, 0.0757, 0.0584, 0.0923], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:29:04,829 INFO [train.py:903] (3/4) Epoch 28, batch 5300, loss[loss=0.2144, simple_loss=0.295, pruned_loss=0.06692, over 17234.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2827, pruned_loss=0.05981, over 3822255.50 frames. ], batch size: 101, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:29:08,403 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189658.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:29:12,992 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189662.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:29:15,036 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189664.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:29:21,498 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 12:29:31,786 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.136e+02 4.811e+02 6.188e+02 7.440e+02 1.460e+03, threshold=1.238e+03, percent-clipped=2.0 +2023-04-03 12:29:44,175 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189687.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:29:46,199 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189689.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:30:05,527 INFO [train.py:903] (3/4) Epoch 28, batch 5350, loss[loss=0.1799, simple_loss=0.2677, pruned_loss=0.04601, over 19658.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2828, pruned_loss=0.06009, over 3820558.09 frames. ], batch size: 55, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:30:09,144 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189709.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:30:26,406 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189722.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:30:29,830 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189725.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:30:38,356 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 12:30:42,327 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9549, 2.0040, 2.2565, 2.5662, 1.9361, 2.4872, 2.1989, 2.0293], + device='cuda:3'), covar=tensor([0.4489, 0.4366, 0.2127, 0.2621, 0.4460, 0.2428, 0.5620, 0.3877], + device='cuda:3'), in_proj_covar=tensor([0.0941, 0.1020, 0.0746, 0.0955, 0.0918, 0.0860, 0.0865, 0.0813], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:30:54,232 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=189746.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:31:04,989 INFO [train.py:903] (3/4) Epoch 28, batch 5400, loss[loss=0.2065, simple_loss=0.2932, pruned_loss=0.05992, over 19659.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.283, pruned_loss=0.06039, over 3817925.98 frames. ], batch size: 55, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:31:16,009 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189763.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:31:33,364 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.324e+02 4.839e+02 5.999e+02 7.745e+02 2.007e+03, threshold=1.200e+03, percent-clipped=4.0 +2023-04-03 12:31:44,821 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8473, 3.2945, 3.3424, 3.3713, 1.5400, 3.2258, 2.8158, 3.1407], + device='cuda:3'), covar=tensor([0.1859, 0.1185, 0.0878, 0.1059, 0.5416, 0.1332, 0.0936, 0.1285], + device='cuda:3'), in_proj_covar=tensor([0.0815, 0.0783, 0.0992, 0.0871, 0.0862, 0.0757, 0.0585, 0.0923], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:31:44,944 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189788.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:31:56,152 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189797.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:32:07,655 INFO [train.py:903] (3/4) Epoch 28, batch 5450, loss[loss=0.2058, simple_loss=0.2947, pruned_loss=0.05852, over 18684.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2838, pruned_loss=0.06103, over 3815586.75 frames. ], batch size: 74, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:32:15,168 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189812.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:32:18,549 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2783, 3.8119, 3.9237, 3.9382, 1.6964, 3.7187, 3.2115, 3.6737], + device='cuda:3'), covar=tensor([0.1856, 0.1058, 0.0732, 0.0850, 0.5995, 0.1209, 0.0840, 0.1277], + device='cuda:3'), in_proj_covar=tensor([0.0816, 0.0784, 0.0994, 0.0874, 0.0863, 0.0759, 0.0586, 0.0925], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:32:27,681 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189822.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:32:30,050 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189824.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:32:30,069 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6103, 1.4600, 1.5874, 2.1707, 1.6099, 1.8268, 1.8708, 1.6266], + device='cuda:3'), covar=tensor([0.0884, 0.0976, 0.1001, 0.0745, 0.0869, 0.0850, 0.0885, 0.0758], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0224, 0.0228, 0.0241, 0.0226, 0.0215, 0.0189, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 12:32:45,841 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=189837.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:33:03,746 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.7316, 2.5877, 2.1478, 2.0922, 1.9120, 2.2550, 1.3434, 1.9319], + device='cuda:3'), covar=tensor([0.0775, 0.0751, 0.0771, 0.1200, 0.1215, 0.1332, 0.1424, 0.1131], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0367, 0.0370, 0.0395, 0.0474, 0.0399, 0.0347, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 12:33:08,052 INFO [train.py:903] (3/4) Epoch 28, batch 5500, loss[loss=0.2576, simple_loss=0.326, pruned_loss=0.09455, over 17550.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2838, pruned_loss=0.06094, over 3803300.36 frames. ], batch size: 101, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:33:17,139 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=189863.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:33:30,489 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 12:33:35,034 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.121e+02 4.755e+02 5.876e+02 8.325e+02 1.733e+03, threshold=1.175e+03, percent-clipped=4.0 +2023-04-03 12:34:10,042 INFO [train.py:903] (3/4) Epoch 28, batch 5550, loss[loss=0.1873, simple_loss=0.2777, pruned_loss=0.04843, over 19788.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2822, pruned_loss=0.05996, over 3812533.55 frames. ], batch size: 54, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:34:17,012 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 12:35:06,753 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 12:35:11,471 INFO [train.py:903] (3/4) Epoch 28, batch 5600, loss[loss=0.1788, simple_loss=0.2565, pruned_loss=0.05054, over 19838.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2823, pruned_loss=0.06003, over 3821111.29 frames. ], batch size: 52, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:35:40,030 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.214e+02 4.597e+02 5.831e+02 7.652e+02 2.230e+03, threshold=1.166e+03, percent-clipped=4.0 +2023-04-03 12:35:40,319 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=189978.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:35:44,621 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=189981.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:36:11,950 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190002.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:36:17,387 INFO [train.py:903] (3/4) Epoch 28, batch 5650, loss[loss=0.1856, simple_loss=0.2728, pruned_loss=0.04922, over 19537.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2821, pruned_loss=0.05979, over 3817504.44 frames. ], batch size: 54, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:36:17,848 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190006.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:37:04,773 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 12:37:18,828 INFO [train.py:903] (3/4) Epoch 28, batch 5700, loss[loss=0.2052, simple_loss=0.2903, pruned_loss=0.06004, over 19708.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2823, pruned_loss=0.06003, over 3822248.22 frames. ], batch size: 59, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:37:28,354 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4010, 2.0997, 1.6665, 1.4583, 1.8892, 1.3882, 1.2731, 1.9151], + device='cuda:3'), covar=tensor([0.1008, 0.0884, 0.1177, 0.0912, 0.0655, 0.1389, 0.0743, 0.0447], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0322, 0.0345, 0.0275, 0.0254, 0.0348, 0.0297, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:37:31,552 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190066.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:37:45,568 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.971e+02 5.283e+02 6.430e+02 7.892e+02 1.365e+03, threshold=1.286e+03, percent-clipped=3.0 +2023-04-03 12:37:49,361 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190080.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:38:02,455 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190090.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:38:21,036 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190105.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:38:21,824 INFO [train.py:903] (3/4) Epoch 28, batch 5750, loss[loss=0.2138, simple_loss=0.2979, pruned_loss=0.06484, over 19513.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2822, pruned_loss=0.05941, over 3829530.51 frames. ], batch size: 64, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:38:24,243 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 12:38:28,167 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5742, 1.4146, 1.4974, 1.8220, 3.1635, 1.3526, 2.5344, 3.6030], + device='cuda:3'), covar=tensor([0.0524, 0.2911, 0.3104, 0.1683, 0.0676, 0.2399, 0.1349, 0.0268], + device='cuda:3'), in_proj_covar=tensor([0.0424, 0.0378, 0.0398, 0.0353, 0.0382, 0.0357, 0.0397, 0.0419], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:38:32,471 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 12:38:35,228 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190117.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:38:37,064 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 12:39:23,778 INFO [train.py:903] (3/4) Epoch 28, batch 5800, loss[loss=0.2265, simple_loss=0.3018, pruned_loss=0.07558, over 19668.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2827, pruned_loss=0.06, over 3821255.35 frames. ], batch size: 53, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:39:25,215 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190157.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:39:41,886 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190169.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:39:52,008 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.261e+02 4.681e+02 5.738e+02 7.022e+02 1.341e+03, threshold=1.148e+03, percent-clipped=1.0 +2023-04-03 12:39:56,749 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190181.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:40:26,978 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190205.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:40:27,627 INFO [train.py:903] (3/4) Epoch 28, batch 5850, loss[loss=0.3347, simple_loss=0.3661, pruned_loss=0.1516, over 13577.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2829, pruned_loss=0.06008, over 3820295.36 frames. ], batch size: 135, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:40:44,021 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5719, 4.1410, 4.2764, 4.2869, 1.8130, 4.0554, 3.4903, 4.0692], + device='cuda:3'), covar=tensor([0.1855, 0.0820, 0.0670, 0.0804, 0.5749, 0.1010, 0.0773, 0.1107], + device='cuda:3'), in_proj_covar=tensor([0.0822, 0.0788, 0.1000, 0.0878, 0.0865, 0.0763, 0.0588, 0.0927], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:41:02,516 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190234.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:41:03,605 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5440, 4.0755, 4.2323, 4.2263, 1.6816, 4.0052, 3.4484, 4.0134], + device='cuda:3'), covar=tensor([0.1800, 0.1012, 0.0697, 0.0838, 0.6434, 0.1194, 0.0782, 0.1152], + device='cuda:3'), in_proj_covar=tensor([0.0823, 0.0789, 0.1002, 0.0880, 0.0867, 0.0765, 0.0589, 0.0928], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:41:29,989 INFO [train.py:903] (3/4) Epoch 28, batch 5900, loss[loss=0.2063, simple_loss=0.2951, pruned_loss=0.05874, over 19619.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2836, pruned_loss=0.06015, over 3819470.56 frames. ], batch size: 57, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:41:32,298 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 12:41:33,980 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190259.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:41:54,166 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 12:41:56,486 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.705e+02 4.583e+02 5.674e+02 7.520e+02 1.844e+03, threshold=1.135e+03, percent-clipped=9.0 +2023-04-03 12:42:32,647 INFO [train.py:903] (3/4) Epoch 28, batch 5950, loss[loss=0.1904, simple_loss=0.2755, pruned_loss=0.05261, over 19692.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2831, pruned_loss=0.05995, over 3818335.67 frames. ], batch size: 60, lr: 2.91e-03, grad_scale: 8.0 +2023-04-03 12:43:08,000 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190334.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:43:34,396 INFO [train.py:903] (3/4) Epoch 28, batch 6000, loss[loss=0.1949, simple_loss=0.2673, pruned_loss=0.06127, over 19462.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2821, pruned_loss=0.05942, over 3821401.04 frames. ], batch size: 49, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:43:34,396 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 12:43:48,493 INFO [train.py:937] (3/4) Epoch 28, validation: loss=0.1668, simple_loss=0.2663, pruned_loss=0.03368, over 944034.00 frames. +2023-04-03 12:43:48,494 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 12:44:08,756 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190373.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:44:11,929 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9890, 1.4324, 1.7641, 1.1248, 2.5800, 3.5268, 3.2430, 3.7810], + device='cuda:3'), covar=tensor([0.1760, 0.3785, 0.3449, 0.2714, 0.0660, 0.0216, 0.0252, 0.0303], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0333, 0.0366, 0.0272, 0.0256, 0.0198, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 12:44:15,159 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.216e+02 4.800e+02 5.909e+02 8.119e+02 1.607e+03, threshold=1.182e+03, percent-clipped=4.0 +2023-04-03 12:44:41,370 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190398.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:44:50,437 INFO [train.py:903] (3/4) Epoch 28, batch 6050, loss[loss=0.21, simple_loss=0.2971, pruned_loss=0.06139, over 19740.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2813, pruned_loss=0.05907, over 3828998.50 frames. ], batch size: 51, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:45:04,537 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190417.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:45:13,171 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1737, 1.3273, 1.7118, 1.0422, 2.2553, 3.0091, 2.6990, 3.1319], + device='cuda:3'), covar=tensor([0.1657, 0.3975, 0.3452, 0.2843, 0.0736, 0.0225, 0.0275, 0.0357], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0333, 0.0366, 0.0272, 0.0255, 0.0197, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 12:45:28,537 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190437.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:45:50,732 INFO [train.py:903] (3/4) Epoch 28, batch 6100, loss[loss=0.2216, simple_loss=0.3033, pruned_loss=0.06996, over 19374.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2813, pruned_loss=0.05941, over 3816888.42 frames. ], batch size: 66, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:45:55,935 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.55 vs. limit=5.0 +2023-04-03 12:45:57,940 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190461.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:45:59,837 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190462.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:46:17,944 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.255e+02 4.836e+02 5.950e+02 7.277e+02 1.439e+03, threshold=1.190e+03, percent-clipped=1.0 +2023-04-03 12:46:28,840 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190486.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:46:46,971 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190501.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:46:53,520 INFO [train.py:903] (3/4) Epoch 28, batch 6150, loss[loss=0.1921, simple_loss=0.2792, pruned_loss=0.05248, over 19679.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2809, pruned_loss=0.05887, over 3829191.48 frames. ], batch size: 60, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:47:02,045 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190513.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:47:22,391 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2421, 2.2681, 2.4970, 2.9576, 2.2561, 2.7945, 2.4304, 2.2837], + device='cuda:3'), covar=tensor([0.4434, 0.4303, 0.2038, 0.2733, 0.4643, 0.2465, 0.5190, 0.3601], + device='cuda:3'), in_proj_covar=tensor([0.0940, 0.1020, 0.0748, 0.0957, 0.0917, 0.0858, 0.0862, 0.0812], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:47:23,097 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 12:47:23,495 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3788, 2.0667, 1.5924, 1.3891, 1.9519, 1.2598, 1.3298, 1.8960], + device='cuda:3'), covar=tensor([0.1140, 0.0885, 0.1129, 0.0971, 0.0618, 0.1417, 0.0809, 0.0506], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0320, 0.0344, 0.0274, 0.0253, 0.0345, 0.0295, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:47:56,364 INFO [train.py:903] (3/4) Epoch 28, batch 6200, loss[loss=0.171, simple_loss=0.253, pruned_loss=0.04447, over 19751.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2796, pruned_loss=0.05797, over 3843887.42 frames. ], batch size: 46, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:48:23,170 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.156e+02 4.897e+02 5.800e+02 7.277e+02 1.783e+03, threshold=1.160e+03, percent-clipped=5.0 +2023-04-03 12:48:45,240 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1596, 2.2112, 2.4248, 2.2763, 3.1589, 2.7283, 3.2563, 2.2363], + device='cuda:3'), covar=tensor([0.2043, 0.3347, 0.2282, 0.1661, 0.1356, 0.1905, 0.1399, 0.3712], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0674, 0.0758, 0.0511, 0.0637, 0.0549, 0.0672, 0.0577], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 12:48:59,515 INFO [train.py:903] (3/4) Epoch 28, batch 6250, loss[loss=0.2124, simple_loss=0.2974, pruned_loss=0.06369, over 18209.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2798, pruned_loss=0.05793, over 3826143.21 frames. ], batch size: 83, lr: 2.90e-03, grad_scale: 16.0 +2023-04-03 12:49:03,317 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190609.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:49:11,021 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190616.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 12:49:13,816 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.41 vs. limit=2.0 +2023-04-03 12:49:22,711 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.3950, 4.0047, 2.5642, 3.4570, 0.7854, 4.0006, 3.7754, 3.9083], + device='cuda:3'), covar=tensor([0.0681, 0.0990, 0.2110, 0.0975, 0.4221, 0.0690, 0.1045, 0.1293], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0434, 0.0521, 0.0362, 0.0412, 0.0459, 0.0453, 0.0485], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:49:26,270 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190627.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:49:27,621 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190628.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:49:31,970 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 12:50:00,482 INFO [train.py:903] (3/4) Epoch 28, batch 6300, loss[loss=0.1761, simple_loss=0.2501, pruned_loss=0.051, over 19758.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2805, pruned_loss=0.05864, over 3819453.22 frames. ], batch size: 46, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:50:28,872 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190678.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:50:29,804 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.019e+02 4.803e+02 6.276e+02 7.910e+02 2.564e+03, threshold=1.255e+03, percent-clipped=8.0 +2023-04-03 12:50:47,639 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=190693.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:51:04,102 INFO [train.py:903] (3/4) Epoch 28, batch 6350, loss[loss=0.1999, simple_loss=0.2811, pruned_loss=0.05941, over 19853.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2812, pruned_loss=0.0592, over 3808443.53 frames. ], batch size: 52, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:52:06,527 INFO [train.py:903] (3/4) Epoch 28, batch 6400, loss[loss=0.1835, simple_loss=0.2553, pruned_loss=0.05588, over 19031.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2817, pruned_loss=0.05948, over 3801564.00 frames. ], batch size: 42, lr: 2.90e-03, grad_scale: 8.0 +2023-04-03 12:52:13,580 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190761.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:52:36,097 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.096e+02 4.755e+02 6.032e+02 8.092e+02 1.400e+03, threshold=1.206e+03, percent-clipped=2.0 +2023-04-03 12:52:52,755 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190793.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:53:04,359 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7558, 4.1118, 4.5815, 4.6120, 1.8572, 4.2896, 3.6553, 4.0362], + device='cuda:3'), covar=tensor([0.2487, 0.1435, 0.0849, 0.1174, 0.7386, 0.2109, 0.1173, 0.1836], + device='cuda:3'), in_proj_covar=tensor([0.0828, 0.0793, 0.1004, 0.0882, 0.0867, 0.0768, 0.0590, 0.0932], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 12:53:08,684 INFO [train.py:903] (3/4) Epoch 28, batch 6450, loss[loss=0.1846, simple_loss=0.2677, pruned_loss=0.05077, over 19857.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2828, pruned_loss=0.0598, over 3803165.54 frames. ], batch size: 52, lr: 2.90e-03, grad_scale: 4.0 +2023-04-03 12:53:55,122 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 12:54:10,283 INFO [train.py:903] (3/4) Epoch 28, batch 6500, loss[loss=0.259, simple_loss=0.3209, pruned_loss=0.09852, over 13322.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.283, pruned_loss=0.05978, over 3796570.92 frames. ], batch size: 135, lr: 2.90e-03, grad_scale: 4.0 +2023-04-03 12:54:17,913 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 12:54:32,004 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190872.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:54:36,607 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=190876.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:54:40,957 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.047e+02 4.758e+02 5.991e+02 7.680e+02 1.999e+03, threshold=1.198e+03, percent-clipped=5.0 +2023-04-03 12:54:47,169 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=190884.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:55:01,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190897.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 12:55:13,651 INFO [train.py:903] (3/4) Epoch 28, batch 6550, loss[loss=0.1991, simple_loss=0.2871, pruned_loss=0.0555, over 19793.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2836, pruned_loss=0.06007, over 3797001.98 frames. ], batch size: 56, lr: 2.90e-03, grad_scale: 4.0 +2023-04-03 12:55:18,346 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=190909.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:56:00,418 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4604, 1.3974, 1.5834, 1.4226, 3.0529, 1.0546, 2.3838, 3.4892], + device='cuda:3'), covar=tensor([0.0542, 0.2926, 0.3014, 0.2044, 0.0754, 0.2715, 0.1267, 0.0268], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0380, 0.0399, 0.0355, 0.0385, 0.0360, 0.0400, 0.0419], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:56:14,190 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190953.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:56:17,106 INFO [train.py:903] (3/4) Epoch 28, batch 6600, loss[loss=0.2086, simple_loss=0.2966, pruned_loss=0.06032, over 19657.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2834, pruned_loss=0.05973, over 3807305.08 frames. ], batch size: 60, lr: 2.90e-03, grad_scale: 4.0 +2023-04-03 12:56:35,885 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=190971.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:56:48,224 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.060e+02 4.870e+02 6.050e+02 8.150e+02 2.542e+03, threshold=1.210e+03, percent-clipped=13.0 +2023-04-03 12:57:19,904 INFO [train.py:903] (3/4) Epoch 28, batch 6650, loss[loss=0.2689, simple_loss=0.3297, pruned_loss=0.104, over 13627.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2833, pruned_loss=0.06019, over 3802478.45 frames. ], batch size: 136, lr: 2.90e-03, grad_scale: 2.0 +2023-04-03 12:57:31,516 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.51 vs. limit=5.0 +2023-04-03 12:57:59,362 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=191037.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:58:14,561 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191049.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:58:17,968 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0434, 1.3127, 1.7502, 1.2804, 2.6954, 3.7404, 3.4529, 4.0164], + device='cuda:3'), covar=tensor([0.1830, 0.3978, 0.3603, 0.2694, 0.0681, 0.0220, 0.0229, 0.0290], + device='cuda:3'), in_proj_covar=tensor([0.0283, 0.0336, 0.0367, 0.0274, 0.0258, 0.0199, 0.0222, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 12:58:20,366 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1388, 3.3709, 1.9383, 2.1163, 3.0101, 1.7181, 1.6972, 2.3876], + device='cuda:3'), covar=tensor([0.1281, 0.0608, 0.1169, 0.0915, 0.0596, 0.1337, 0.0929, 0.0623], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0317, 0.0342, 0.0272, 0.0252, 0.0342, 0.0293, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 12:58:22,342 INFO [train.py:903] (3/4) Epoch 28, batch 6700, loss[loss=0.2537, simple_loss=0.3146, pruned_loss=0.09637, over 13123.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2829, pruned_loss=0.05992, over 3803095.96 frames. ], batch size: 138, lr: 2.90e-03, grad_scale: 2.0 +2023-04-03 12:58:38,439 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191068.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 12:58:45,550 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191074.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:58:52,986 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.114e+02 5.245e+02 6.198e+02 8.145e+02 2.088e+03, threshold=1.240e+03, percent-clipped=7.0 +2023-04-03 12:58:58,990 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=191086.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:58:59,053 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191086.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 12:59:11,381 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2182, 1.1664, 1.2172, 1.2935, 1.0033, 1.2996, 1.2864, 1.2695], + device='cuda:3'), covar=tensor([0.0923, 0.1000, 0.1061, 0.0698, 0.0965, 0.0919, 0.0878, 0.0810], + device='cuda:3'), in_proj_covar=tensor([0.0215, 0.0224, 0.0230, 0.0241, 0.0228, 0.0216, 0.0190, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 12:59:21,826 INFO [train.py:903] (3/4) Epoch 28, batch 6750, loss[loss=0.23, simple_loss=0.3107, pruned_loss=0.07467, over 19314.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2828, pruned_loss=0.05992, over 3816769.93 frames. ], batch size: 66, lr: 2.90e-03, grad_scale: 2.0 +2023-04-03 12:59:51,523 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191132.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:00:04,804 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7926, 1.6221, 1.6950, 2.4129, 1.6835, 2.0816, 2.0439, 1.8990], + device='cuda:3'), covar=tensor([0.0846, 0.0886, 0.0967, 0.0660, 0.0895, 0.0782, 0.0897, 0.0667], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0223, 0.0229, 0.0241, 0.0227, 0.0216, 0.0189, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 13:00:11,326 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=191150.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:00:13,736 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191152.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:00:17,984 INFO [train.py:903] (3/4) Epoch 28, batch 6800, loss[loss=0.2338, simple_loss=0.3074, pruned_loss=0.0801, over 19730.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2839, pruned_loss=0.06062, over 3806689.21 frames. ], batch size: 63, lr: 2.90e-03, grad_scale: 4.0 +2023-04-03 13:00:19,437 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191157.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:00:37,661 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0778, 1.0801, 1.1759, 1.1796, 1.3544, 1.4227, 1.3462, 0.6135], + device='cuda:3'), covar=tensor([0.1873, 0.3246, 0.1986, 0.1557, 0.1352, 0.1822, 0.1192, 0.4476], + device='cuda:3'), in_proj_covar=tensor([0.0557, 0.0673, 0.0760, 0.0511, 0.0639, 0.0548, 0.0673, 0.0578], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 13:00:45,260 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.070e+02 5.099e+02 6.292e+02 8.518e+02 1.501e+03, threshold=1.258e+03, percent-clipped=3.0 +2023-04-03 13:01:04,263 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 13:01:05,298 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 13:01:08,826 INFO [train.py:903] (3/4) Epoch 29, batch 0, loss[loss=0.1998, simple_loss=0.281, pruned_loss=0.05936, over 19765.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.281, pruned_loss=0.05936, over 19765.00 frames. ], batch size: 54, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:01:08,827 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 13:01:20,492 INFO [train.py:937] (3/4) Epoch 29, validation: loss=0.1669, simple_loss=0.2669, pruned_loss=0.03339, over 944034.00 frames. +2023-04-03 13:01:20,493 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 13:01:31,763 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 13:01:43,669 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=191203.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:02:01,247 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-04-03 13:02:19,661 INFO [train.py:903] (3/4) Epoch 29, batch 50, loss[loss=0.1816, simple_loss=0.275, pruned_loss=0.04406, over 19669.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2818, pruned_loss=0.05934, over 866215.70 frames. ], batch size: 58, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:02:55,009 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 13:03:01,004 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0881, 2.0513, 2.0075, 1.8553, 1.5732, 1.7807, 0.5552, 1.1321], + device='cuda:3'), covar=tensor([0.0731, 0.0654, 0.0469, 0.0801, 0.1330, 0.0877, 0.1417, 0.1197], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0364, 0.0365, 0.0391, 0.0469, 0.0397, 0.0345, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 13:03:15,130 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.166e+02 5.210e+02 6.215e+02 8.429e+02 1.754e+03, threshold=1.243e+03, percent-clipped=6.0 +2023-04-03 13:03:18,589 INFO [train.py:903] (3/4) Epoch 29, batch 100, loss[loss=0.2145, simple_loss=0.2987, pruned_loss=0.06513, over 19521.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2826, pruned_loss=0.05843, over 1530661.05 frames. ], batch size: 64, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:03:33,066 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 13:04:07,326 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191324.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 13:04:19,665 INFO [train.py:903] (3/4) Epoch 29, batch 150, loss[loss=0.1759, simple_loss=0.2571, pruned_loss=0.04733, over 19754.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2814, pruned_loss=0.05881, over 2044766.13 frames. ], batch size: 47, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:04:29,083 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191342.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:04:36,947 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191349.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 13:04:59,835 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191367.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:05:05,619 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.0943, 2.8261, 2.3788, 2.2568, 2.0517, 2.4633, 1.2287, 2.1051], + device='cuda:3'), covar=tensor([0.0797, 0.0690, 0.0714, 0.1305, 0.1226, 0.1231, 0.1488, 0.1146], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0365, 0.0366, 0.0392, 0.0470, 0.0398, 0.0345, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 13:05:07,807 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8113, 4.3818, 2.9896, 3.7504, 0.9564, 4.3906, 4.2002, 4.3142], + device='cuda:3'), covar=tensor([0.0603, 0.0950, 0.1719, 0.0968, 0.4060, 0.0617, 0.0984, 0.1016], + device='cuda:3'), in_proj_covar=tensor([0.0528, 0.0429, 0.0514, 0.0358, 0.0407, 0.0454, 0.0448, 0.0479], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:05:15,358 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.397e+02 5.272e+02 6.303e+02 7.742e+02 1.475e+03, threshold=1.261e+03, percent-clipped=3.0 +2023-04-03 13:05:15,420 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 13:05:18,079 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=191383.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:05:18,891 INFO [train.py:903] (3/4) Epoch 29, batch 200, loss[loss=0.1705, simple_loss=0.2501, pruned_loss=0.04541, over 19616.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2826, pruned_loss=0.05926, over 2431209.11 frames. ], batch size: 50, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:05:48,172 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191408.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:06:13,737 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=191430.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:06:18,511 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191433.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:06:20,064 INFO [train.py:903] (3/4) Epoch 29, batch 250, loss[loss=0.1972, simple_loss=0.2824, pruned_loss=0.05604, over 19606.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2819, pruned_loss=0.05906, over 2747939.67 frames. ], batch size: 61, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:07:11,569 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6967, 1.4950, 1.6037, 2.1566, 1.6075, 1.8996, 1.8931, 1.7210], + device='cuda:3'), covar=tensor([0.0876, 0.0991, 0.1029, 0.0662, 0.0857, 0.0805, 0.0923, 0.0728], + device='cuda:3'), in_proj_covar=tensor([0.0214, 0.0225, 0.0230, 0.0241, 0.0228, 0.0217, 0.0190, 0.0210], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 13:07:16,619 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.382e+02 5.311e+02 6.503e+02 8.385e+02 2.446e+03, threshold=1.301e+03, percent-clipped=7.0 +2023-04-03 13:07:20,115 INFO [train.py:903] (3/4) Epoch 29, batch 300, loss[loss=0.1771, simple_loss=0.2619, pruned_loss=0.04618, over 19719.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2816, pruned_loss=0.05896, over 2988612.23 frames. ], batch size: 51, lr: 2.85e-03, grad_scale: 8.0 +2023-04-03 13:07:33,157 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=191494.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:07:57,193 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.57 vs. limit=5.0 +2023-04-03 13:08:19,686 INFO [train.py:903] (3/4) Epoch 29, batch 350, loss[loss=0.2192, simple_loss=0.2971, pruned_loss=0.07065, over 19428.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2802, pruned_loss=0.05853, over 3183226.84 frames. ], batch size: 64, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:08:27,202 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 13:08:33,102 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191545.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:08:35,146 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=191547.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:09:02,265 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.59 vs. limit=5.0 +2023-04-03 13:09:16,624 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.400e+02 4.772e+02 5.697e+02 6.938e+02 1.378e+03, threshold=1.139e+03, percent-clipped=1.0 +2023-04-03 13:09:19,980 INFO [train.py:903] (3/4) Epoch 29, batch 400, loss[loss=0.2, simple_loss=0.2808, pruned_loss=0.05965, over 19576.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2797, pruned_loss=0.05826, over 3337003.81 frames. ], batch size: 52, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:09:26,934 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9505, 2.0206, 2.2812, 2.6185, 1.9244, 2.4462, 2.2350, 2.0722], + device='cuda:3'), covar=tensor([0.4300, 0.4084, 0.2048, 0.2479, 0.4306, 0.2296, 0.5101, 0.3529], + device='cuda:3'), in_proj_covar=tensor([0.0941, 0.1020, 0.0746, 0.0955, 0.0918, 0.0859, 0.0863, 0.0810], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 13:09:50,423 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191609.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:10:19,938 INFO [train.py:903] (3/4) Epoch 29, batch 450, loss[loss=0.22, simple_loss=0.2985, pruned_loss=0.07078, over 19769.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.28, pruned_loss=0.05843, over 3458972.61 frames. ], batch size: 54, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:10:55,239 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191662.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:10:57,252 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 13:10:58,392 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 13:11:17,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.061e+02 4.702e+02 5.973e+02 7.083e+02 1.398e+03, threshold=1.195e+03, percent-clipped=2.0 +2023-04-03 13:11:20,347 INFO [train.py:903] (3/4) Epoch 29, batch 500, loss[loss=0.1811, simple_loss=0.2669, pruned_loss=0.04763, over 19779.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.28, pruned_loss=0.05836, over 3551681.65 frames. ], batch size: 56, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:11:27,827 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.08 vs. limit=2.0 +2023-04-03 13:12:13,383 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=191727.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:12:21,065 INFO [train.py:903] (3/4) Epoch 29, batch 550, loss[loss=0.2331, simple_loss=0.3052, pruned_loss=0.08048, over 14042.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2796, pruned_loss=0.05831, over 3603269.93 frames. ], batch size: 136, lr: 2.84e-03, grad_scale: 4.0 +2023-04-03 13:13:19,227 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.607e+02 5.419e+02 6.471e+02 8.968e+02 1.855e+03, threshold=1.294e+03, percent-clipped=10.0 +2023-04-03 13:13:22,383 INFO [train.py:903] (3/4) Epoch 29, batch 600, loss[loss=0.2308, simple_loss=0.3086, pruned_loss=0.07647, over 19758.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2804, pruned_loss=0.05863, over 3659399.26 frames. ], batch size: 63, lr: 2.84e-03, grad_scale: 4.0 +2023-04-03 13:13:42,719 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191801.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:14:02,346 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 13:14:13,358 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191826.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:14:22,145 INFO [train.py:903] (3/4) Epoch 29, batch 650, loss[loss=0.1821, simple_loss=0.2652, pruned_loss=0.04957, over 19617.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2815, pruned_loss=0.05911, over 3701290.07 frames. ], batch size: 50, lr: 2.84e-03, grad_scale: 4.0 +2023-04-03 13:14:31,535 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=191842.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:15:01,025 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191865.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:15:18,379 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.813e+02 4.883e+02 6.286e+02 7.850e+02 2.365e+03, threshold=1.257e+03, percent-clipped=4.0 +2023-04-03 13:15:21,594 INFO [train.py:903] (3/4) Epoch 29, batch 700, loss[loss=0.2049, simple_loss=0.297, pruned_loss=0.05637, over 19477.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2823, pruned_loss=0.05944, over 3732470.09 frames. ], batch size: 64, lr: 2.84e-03, grad_scale: 4.0 +2023-04-03 13:15:29,601 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191890.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:15:36,392 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8165, 1.3323, 1.0681, 0.9417, 1.1636, 1.0062, 0.8758, 1.2084], + device='cuda:3'), covar=tensor([0.0731, 0.1003, 0.1248, 0.0879, 0.0679, 0.1422, 0.0724, 0.0585], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0318, 0.0342, 0.0272, 0.0252, 0.0345, 0.0293, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:16:03,823 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=191918.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:16:23,389 INFO [train.py:903] (3/4) Epoch 29, batch 750, loss[loss=0.1562, simple_loss=0.2451, pruned_loss=0.03368, over 19619.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2811, pruned_loss=0.05885, over 3763498.37 frames. ], batch size: 50, lr: 2.84e-03, grad_scale: 4.0 +2023-04-03 13:16:29,942 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3502, 2.3583, 2.6836, 3.1759, 2.3916, 2.9333, 2.6724, 2.5024], + device='cuda:3'), covar=tensor([0.4294, 0.4389, 0.1939, 0.2631, 0.4696, 0.2391, 0.5049, 0.3390], + device='cuda:3'), in_proj_covar=tensor([0.0944, 0.1021, 0.0748, 0.0956, 0.0920, 0.0860, 0.0867, 0.0811], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 13:16:34,302 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=191943.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:17:21,356 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.970e+02 4.754e+02 5.591e+02 7.301e+02 1.189e+03, threshold=1.118e+03, percent-clipped=0.0 +2023-04-03 13:17:23,708 INFO [train.py:903] (3/4) Epoch 29, batch 800, loss[loss=0.1957, simple_loss=0.2824, pruned_loss=0.05452, over 19758.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.281, pruned_loss=0.05863, over 3790209.47 frames. ], batch size: 54, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:17:36,669 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 13:17:38,139 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8543, 4.4571, 2.6595, 3.8723, 0.7876, 4.3946, 4.2847, 4.3317], + device='cuda:3'), covar=tensor([0.0520, 0.0826, 0.1940, 0.0842, 0.4060, 0.0653, 0.0897, 0.1055], + device='cuda:3'), in_proj_covar=tensor([0.0530, 0.0431, 0.0518, 0.0360, 0.0410, 0.0457, 0.0454, 0.0483], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:17:51,917 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.1727, 1.8122, 2.1090, 1.9500, 4.6254, 1.2863, 2.6972, 5.0904], + device='cuda:3'), covar=tensor([0.0446, 0.2767, 0.2615, 0.1947, 0.0730, 0.2704, 0.1366, 0.0161], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0380, 0.0400, 0.0357, 0.0385, 0.0360, 0.0400, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:18:26,431 INFO [train.py:903] (3/4) Epoch 29, batch 850, loss[loss=0.1818, simple_loss=0.2686, pruned_loss=0.04754, over 19688.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2811, pruned_loss=0.0584, over 3800055.20 frames. ], batch size: 53, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:19:17,579 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 13:19:23,133 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.899e+02 4.771e+02 5.573e+02 7.809e+02 2.111e+03, threshold=1.115e+03, percent-clipped=7.0 +2023-04-03 13:19:25,389 INFO [train.py:903] (3/4) Epoch 29, batch 900, loss[loss=0.1929, simple_loss=0.2772, pruned_loss=0.05432, over 18708.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2812, pruned_loss=0.05856, over 3808381.02 frames. ], batch size: 74, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:19:43,240 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=192098.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:19:53,200 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-04-03 13:20:09,373 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192120.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:20:10,714 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6482, 1.7359, 2.0154, 2.0190, 1.5473, 1.9298, 1.9783, 1.8671], + device='cuda:3'), covar=tensor([0.4351, 0.4067, 0.2061, 0.2546, 0.4020, 0.2376, 0.5440, 0.3548], + device='cuda:3'), in_proj_covar=tensor([0.0949, 0.1027, 0.0753, 0.0962, 0.0923, 0.0864, 0.0870, 0.0815], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 13:20:12,931 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=192123.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:20:26,584 INFO [train.py:903] (3/4) Epoch 29, batch 950, loss[loss=0.1904, simple_loss=0.2806, pruned_loss=0.0501, over 19602.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.281, pruned_loss=0.05828, over 3824112.90 frames. ], batch size: 57, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:20:30,191 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 13:21:13,719 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192173.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:21:24,649 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.322e+02 5.272e+02 6.130e+02 7.632e+02 1.213e+03, threshold=1.226e+03, percent-clipped=2.0 +2023-04-03 13:21:27,723 INFO [train.py:903] (3/4) Epoch 29, batch 1000, loss[loss=0.1975, simple_loss=0.2808, pruned_loss=0.05708, over 19588.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2807, pruned_loss=0.05791, over 3828846.08 frames. ], batch size: 52, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:22:21,150 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2468, 1.9914, 1.5280, 1.3500, 1.8467, 1.1732, 1.3284, 1.7471], + device='cuda:3'), covar=tensor([0.1004, 0.0810, 0.1162, 0.0853, 0.0597, 0.1382, 0.0707, 0.0515], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0318, 0.0342, 0.0273, 0.0252, 0.0345, 0.0293, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:22:23,179 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 13:22:27,656 INFO [train.py:903] (3/4) Epoch 29, batch 1050, loss[loss=0.1704, simple_loss=0.2525, pruned_loss=0.04412, over 19424.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2815, pruned_loss=0.05841, over 3835611.63 frames. ], batch size: 48, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:22:31,101 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192237.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:22:34,342 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192239.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:22:55,092 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.68 vs. limit=2.0 +2023-04-03 13:23:02,363 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 13:23:25,657 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.003e+02 4.755e+02 5.709e+02 7.263e+02 1.610e+03, threshold=1.142e+03, percent-clipped=3.0 +2023-04-03 13:23:28,075 INFO [train.py:903] (3/4) Epoch 29, batch 1100, loss[loss=0.2182, simple_loss=0.299, pruned_loss=0.06863, over 17435.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2811, pruned_loss=0.05823, over 3824789.86 frames. ], batch size: 101, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:24:29,302 INFO [train.py:903] (3/4) Epoch 29, batch 1150, loss[loss=0.1727, simple_loss=0.2497, pruned_loss=0.0479, over 19384.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.281, pruned_loss=0.0583, over 3817021.56 frames. ], batch size: 47, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:24:36,022 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6179, 1.4727, 1.5277, 2.0049, 1.5043, 1.8370, 1.8442, 1.6848], + device='cuda:3'), covar=tensor([0.0852, 0.0943, 0.1007, 0.0681, 0.0950, 0.0793, 0.0887, 0.0713], + device='cuda:3'), in_proj_covar=tensor([0.0215, 0.0224, 0.0230, 0.0241, 0.0228, 0.0216, 0.0189, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 13:25:27,744 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.011e+02 4.839e+02 5.803e+02 7.271e+02 1.538e+03, threshold=1.161e+03, percent-clipped=1.0 +2023-04-03 13:25:30,849 INFO [train.py:903] (3/4) Epoch 29, batch 1200, loss[loss=0.2032, simple_loss=0.2952, pruned_loss=0.05555, over 19640.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2806, pruned_loss=0.05818, over 3825693.50 frames. ], batch size: 57, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:26:01,113 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192410.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:26:01,919 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 13:26:10,314 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192417.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:26:30,220 INFO [train.py:903] (3/4) Epoch 29, batch 1250, loss[loss=0.2356, simple_loss=0.3066, pruned_loss=0.08232, over 19584.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2815, pruned_loss=0.05882, over 3815994.97 frames. ], batch size: 61, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:26:50,091 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192451.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:27:07,166 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192464.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:27:28,115 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.043e+02 5.075e+02 6.202e+02 7.494e+02 1.680e+03, threshold=1.240e+03, percent-clipped=4.0 +2023-04-03 13:27:30,143 INFO [train.py:903] (3/4) Epoch 29, batch 1300, loss[loss=0.274, simple_loss=0.3288, pruned_loss=0.1097, over 12666.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2829, pruned_loss=0.06004, over 3802406.26 frames. ], batch size: 136, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:28:07,574 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-04-03 13:28:10,139 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192517.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:28:30,907 INFO [train.py:903] (3/4) Epoch 29, batch 1350, loss[loss=0.2175, simple_loss=0.3045, pruned_loss=0.0653, over 19581.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2836, pruned_loss=0.06055, over 3804369.18 frames. ], batch size: 61, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:29:25,906 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192579.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:29:27,889 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192581.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:29:28,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.238e+02 5.533e+02 7.774e+02 1.028e+03 2.542e+03, threshold=1.555e+03, percent-clipped=13.0 +2023-04-03 13:29:31,164 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192583.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:29:32,199 INFO [train.py:903] (3/4) Epoch 29, batch 1400, loss[loss=0.1983, simple_loss=0.2826, pruned_loss=0.05704, over 19744.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2833, pruned_loss=0.06044, over 3814235.18 frames. ], batch size: 63, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:29:40,690 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6511, 1.2346, 1.2600, 1.5230, 1.0719, 1.4065, 1.2561, 1.4706], + device='cuda:3'), covar=tensor([0.1192, 0.1302, 0.1729, 0.1086, 0.1424, 0.0663, 0.1653, 0.0890], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0361, 0.0320, 0.0259, 0.0307, 0.0257, 0.0323, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 13:30:30,407 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192632.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:30:32,442 INFO [train.py:903] (3/4) Epoch 29, batch 1450, loss[loss=0.2133, simple_loss=0.2968, pruned_loss=0.06485, over 19601.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2838, pruned_loss=0.06062, over 3819094.29 frames. ], batch size: 57, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:30:32,473 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 13:30:32,745 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192634.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:30:48,187 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7344, 1.7454, 1.6846, 1.5236, 1.3869, 1.3987, 0.3384, 0.7019], + device='cuda:3'), covar=tensor([0.0693, 0.0668, 0.0452, 0.0668, 0.1369, 0.0817, 0.1371, 0.1176], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0365, 0.0368, 0.0392, 0.0472, 0.0399, 0.0346, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 13:30:49,186 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5831, 1.1449, 1.3123, 1.1878, 2.2244, 1.0988, 2.1017, 2.5721], + device='cuda:3'), covar=tensor([0.0730, 0.2916, 0.3182, 0.1925, 0.0869, 0.2177, 0.1228, 0.0435], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0379, 0.0400, 0.0355, 0.0385, 0.0360, 0.0399, 0.0421], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:31:29,875 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.888e+02 4.834e+02 5.734e+02 7.377e+02 1.406e+03, threshold=1.147e+03, percent-clipped=0.0 +2023-04-03 13:31:32,942 INFO [train.py:903] (3/4) Epoch 29, batch 1500, loss[loss=0.1729, simple_loss=0.2606, pruned_loss=0.04254, over 19624.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2836, pruned_loss=0.06047, over 3808572.82 frames. ], batch size: 50, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:31:41,392 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192691.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:31:47,436 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192696.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:31:50,561 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192698.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:32:32,341 INFO [train.py:903] (3/4) Epoch 29, batch 1550, loss[loss=0.223, simple_loss=0.29, pruned_loss=0.07801, over 19615.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.283, pruned_loss=0.06057, over 3818980.77 frames. ], batch size: 50, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:32:56,577 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192753.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:32:57,543 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192754.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:33:05,616 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192761.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:33:31,831 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.013e+02 5.130e+02 6.287e+02 7.695e+02 1.691e+03, threshold=1.257e+03, percent-clipped=3.0 +2023-04-03 13:33:34,857 INFO [train.py:903] (3/4) Epoch 29, batch 1600, loss[loss=0.2272, simple_loss=0.308, pruned_loss=0.07319, over 18678.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06028, over 3810662.07 frames. ], batch size: 74, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:33:47,559 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192795.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:33:57,596 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 13:34:34,143 INFO [train.py:903] (3/4) Epoch 29, batch 1650, loss[loss=0.1755, simple_loss=0.2617, pruned_loss=0.04466, over 19696.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2835, pruned_loss=0.06023, over 3825637.21 frames. ], batch size: 53, lr: 2.84e-03, grad_scale: 8.0 +2023-04-03 13:34:35,741 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=192835.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:34:51,801 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=192849.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 13:35:04,865 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=192860.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:35:15,872 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192869.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:35:23,843 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192876.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:35:25,041 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0311, 1.8127, 1.6781, 2.1149, 1.7818, 1.7200, 1.7156, 1.9192], + device='cuda:3'), covar=tensor([0.1107, 0.1622, 0.1644, 0.1117, 0.1498, 0.0635, 0.1576, 0.0846], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0362, 0.0321, 0.0259, 0.0308, 0.0258, 0.0324, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 13:35:30,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.635e+02 4.944e+02 5.872e+02 7.824e+02 1.796e+03, threshold=1.174e+03, percent-clipped=2.0 +2023-04-03 13:35:32,715 INFO [train.py:903] (3/4) Epoch 29, batch 1700, loss[loss=0.228, simple_loss=0.305, pruned_loss=0.0755, over 19315.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2836, pruned_loss=0.06041, over 3818710.20 frames. ], batch size: 66, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:35:38,395 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=192888.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:35:52,009 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.42 vs. limit=2.0 +2023-04-03 13:36:04,838 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=192910.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:36:08,206 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=192913.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:36:13,393 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 13:36:32,300 INFO [train.py:903] (3/4) Epoch 29, batch 1750, loss[loss=0.2137, simple_loss=0.2923, pruned_loss=0.06752, over 19594.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.283, pruned_loss=0.06035, over 3822591.92 frames. ], batch size: 52, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:36:55,847 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=192952.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:36:58,027 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=192954.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:37:24,866 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=192977.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:37:25,783 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=192978.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:37:27,198 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=192979.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:37:31,717 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.864e+02 4.741e+02 6.117e+02 7.797e+02 1.758e+03, threshold=1.223e+03, percent-clipped=7.0 +2023-04-03 13:37:34,999 INFO [train.py:903] (3/4) Epoch 29, batch 1800, loss[loss=0.2206, simple_loss=0.308, pruned_loss=0.06661, over 19676.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.283, pruned_loss=0.06011, over 3830694.82 frames. ], batch size: 60, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:38:29,372 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 13:38:33,981 INFO [train.py:903] (3/4) Epoch 29, batch 1850, loss[loss=0.2068, simple_loss=0.2887, pruned_loss=0.06246, over 19729.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.283, pruned_loss=0.06064, over 3816161.70 frames. ], batch size: 63, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:38:35,285 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193035.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:39:04,664 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193059.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:39:06,412 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 13:39:32,451 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.498e+02 4.794e+02 5.817e+02 7.567e+02 2.266e+03, threshold=1.163e+03, percent-clipped=4.0 +2023-04-03 13:39:34,764 INFO [train.py:903] (3/4) Epoch 29, batch 1900, loss[loss=0.2381, simple_loss=0.312, pruned_loss=0.08215, over 17480.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2825, pruned_loss=0.05998, over 3819316.17 frames. ], batch size: 101, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:39:45,997 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193093.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:39:49,045 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 13:39:51,204 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193097.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:39:55,327 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 13:40:18,690 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 13:40:25,438 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193125.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:40:33,605 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193132.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:40:35,437 INFO [train.py:903] (3/4) Epoch 29, batch 1950, loss[loss=0.1967, simple_loss=0.2787, pruned_loss=0.05735, over 19677.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2821, pruned_loss=0.05961, over 3826402.78 frames. ], batch size: 53, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:40:53,402 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-04-03 13:40:56,452 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193150.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:40:56,492 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193150.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:41:04,228 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193157.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:41:08,740 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8990, 0.8199, 0.8656, 1.0477, 0.8341, 0.9263, 0.9374, 0.8959], + device='cuda:3'), covar=tensor([0.0810, 0.0923, 0.0948, 0.0571, 0.0939, 0.0799, 0.0861, 0.0755], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0230, 0.0240, 0.0227, 0.0215, 0.0189, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 13:41:15,432 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193166.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:41:34,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.104e+02 4.788e+02 5.811e+02 7.006e+02 1.429e+03, threshold=1.162e+03, percent-clipped=3.0 +2023-04-03 13:41:37,852 INFO [train.py:903] (3/4) Epoch 29, batch 2000, loss[loss=0.1823, simple_loss=0.2725, pruned_loss=0.04608, over 19665.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2823, pruned_loss=0.05953, over 3827725.74 frames. ], batch size: 58, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:41:46,097 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193191.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:41:48,034 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193193.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 13:41:59,702 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.34 vs. limit=5.0 +2023-04-03 13:42:10,516 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193212.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:42:31,349 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 13:42:31,599 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193229.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:42:37,084 INFO [train.py:903] (3/4) Epoch 29, batch 2050, loss[loss=0.1782, simple_loss=0.2513, pruned_loss=0.05252, over 19312.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2827, pruned_loss=0.05982, over 3842619.07 frames. ], batch size: 44, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:42:50,732 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 13:42:50,762 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 13:43:13,504 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 13:43:34,587 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.211e+02 4.912e+02 6.403e+02 8.614e+02 2.001e+03, threshold=1.281e+03, percent-clipped=8.0 +2023-04-03 13:43:36,910 INFO [train.py:903] (3/4) Epoch 29, batch 2100, loss[loss=0.1972, simple_loss=0.2835, pruned_loss=0.05539, over 19774.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2831, pruned_loss=0.06034, over 3838333.02 frames. ], batch size: 56, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:44:00,083 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193302.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:44:07,583 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 13:44:07,963 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193308.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 13:44:25,618 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 13:44:37,257 INFO [train.py:903] (3/4) Epoch 29, batch 2150, loss[loss=0.1794, simple_loss=0.2601, pruned_loss=0.04938, over 19468.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.283, pruned_loss=0.06019, over 3828803.78 frames. ], batch size: 49, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:44:56,808 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193349.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:45:25,847 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193374.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:45:35,169 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.248e+02 5.242e+02 6.391e+02 8.913e+02 2.261e+03, threshold=1.278e+03, percent-clipped=9.0 +2023-04-03 13:45:37,407 INFO [train.py:903] (3/4) Epoch 29, batch 2200, loss[loss=0.1795, simple_loss=0.2665, pruned_loss=0.04627, over 19503.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2842, pruned_loss=0.06077, over 3822527.44 frames. ], batch size: 64, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:46:00,721 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193403.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:46:04,436 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193406.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:46:30,803 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2053, 1.5417, 2.0253, 1.5108, 2.9929, 4.5792, 4.4259, 4.9941], + device='cuda:3'), covar=tensor([0.1743, 0.3887, 0.3416, 0.2543, 0.0713, 0.0210, 0.0176, 0.0211], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0334, 0.0367, 0.0273, 0.0257, 0.0198, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 13:46:36,066 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193431.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:46:39,105 INFO [train.py:903] (3/4) Epoch 29, batch 2250, loss[loss=0.2387, simple_loss=0.31, pruned_loss=0.08369, over 19577.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2835, pruned_loss=0.06046, over 3828535.51 frames. ], batch size: 61, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:46:45,262 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193439.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:47:20,663 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193468.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:47:25,803 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193472.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 13:47:26,835 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2428, 1.3193, 1.7864, 1.2856, 2.5079, 3.2981, 3.0302, 3.4715], + device='cuda:3'), covar=tensor([0.1704, 0.4116, 0.3549, 0.2725, 0.0712, 0.0232, 0.0281, 0.0384], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0335, 0.0367, 0.0274, 0.0257, 0.0198, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 13:47:36,425 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.451e+02 4.884e+02 5.894e+02 7.303e+02 1.661e+03, threshold=1.179e+03, percent-clipped=4.0 +2023-04-03 13:47:38,742 INFO [train.py:903] (3/4) Epoch 29, batch 2300, loss[loss=0.189, simple_loss=0.2759, pruned_loss=0.05102, over 18045.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06023, over 3821448.49 frames. ], batch size: 83, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:47:50,478 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193493.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:47:53,671 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 13:48:21,005 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193518.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:48:39,106 INFO [train.py:903] (3/4) Epoch 29, batch 2350, loss[loss=0.1705, simple_loss=0.2571, pruned_loss=0.04194, over 19849.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.283, pruned_loss=0.05982, over 3829116.09 frames. ], batch size: 52, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:49:16,327 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193564.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 13:49:20,453 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 13:49:27,329 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193573.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:49:36,503 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 13:49:37,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.068e+02 4.546e+02 5.901e+02 7.828e+02 1.716e+03, threshold=1.180e+03, percent-clipped=11.0 +2023-04-03 13:49:40,761 INFO [train.py:903] (3/4) Epoch 29, batch 2400, loss[loss=0.2074, simple_loss=0.305, pruned_loss=0.05487, over 19684.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2829, pruned_loss=0.05963, over 3824000.58 frames. ], batch size: 55, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:49:47,424 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193589.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 13:50:40,462 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193633.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:50:41,253 INFO [train.py:903] (3/4) Epoch 29, batch 2450, loss[loss=0.1993, simple_loss=0.2802, pruned_loss=0.05917, over 19624.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.283, pruned_loss=0.0599, over 3812408.27 frames. ], batch size: 50, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:50:54,660 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193646.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:51:39,135 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.050e+02 4.882e+02 5.957e+02 8.043e+02 1.393e+03, threshold=1.191e+03, percent-clipped=5.0 +2023-04-03 13:51:39,619 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2347, 2.1688, 2.1032, 1.9887, 1.7565, 1.8713, 0.6729, 1.2070], + device='cuda:3'), covar=tensor([0.0707, 0.0655, 0.0481, 0.0775, 0.1208, 0.0927, 0.1435, 0.1190], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0368, 0.0371, 0.0395, 0.0475, 0.0402, 0.0348, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 13:51:41,321 INFO [train.py:903] (3/4) Epoch 29, batch 2500, loss[loss=0.2114, simple_loss=0.2795, pruned_loss=0.07164, over 19801.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2824, pruned_loss=0.05946, over 3805665.17 frames. ], batch size: 49, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:51:45,911 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193688.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:52:40,704 INFO [train.py:903] (3/4) Epoch 29, batch 2550, loss[loss=0.1714, simple_loss=0.2514, pruned_loss=0.04568, over 17750.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2831, pruned_loss=0.06014, over 3807881.24 frames. ], batch size: 39, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:53:15,362 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193761.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:53:30,078 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193774.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:53:35,078 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 13:53:41,002 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.098e+02 5.186e+02 6.141e+02 8.140e+02 1.584e+03, threshold=1.228e+03, percent-clipped=10.0 +2023-04-03 13:53:41,172 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193783.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:53:42,128 INFO [train.py:903] (3/4) Epoch 29, batch 2600, loss[loss=0.2027, simple_loss=0.2741, pruned_loss=0.06568, over 19401.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2835, pruned_loss=0.06013, over 3815412.72 frames. ], batch size: 47, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:54:02,298 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193799.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:54:22,070 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193816.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 13:54:39,253 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8020, 4.3936, 2.7365, 3.8500, 0.9167, 4.3898, 4.2238, 4.3323], + device='cuda:3'), covar=tensor([0.0604, 0.0975, 0.2006, 0.0881, 0.4188, 0.0625, 0.0967, 0.1178], + device='cuda:3'), in_proj_covar=tensor([0.0532, 0.0430, 0.0520, 0.0358, 0.0412, 0.0457, 0.0453, 0.0485], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:54:44,187 INFO [train.py:903] (3/4) Epoch 29, batch 2650, loss[loss=0.2467, simple_loss=0.3177, pruned_loss=0.08785, over 17676.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2844, pruned_loss=0.06043, over 3798157.18 frames. ], batch size: 101, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:54:57,015 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7478, 1.7684, 1.7681, 1.5428, 1.4953, 1.5286, 0.2981, 0.7114], + device='cuda:3'), covar=tensor([0.0768, 0.0727, 0.0452, 0.0689, 0.1357, 0.0823, 0.1434, 0.1250], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0366, 0.0370, 0.0395, 0.0474, 0.0400, 0.0348, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 13:54:59,846 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 13:55:43,537 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.739e+02 4.600e+02 5.538e+02 7.187e+02 1.315e+03, threshold=1.108e+03, percent-clipped=1.0 +2023-04-03 13:55:44,725 INFO [train.py:903] (3/4) Epoch 29, batch 2700, loss[loss=0.2178, simple_loss=0.2957, pruned_loss=0.06996, over 12911.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2835, pruned_loss=0.05972, over 3793540.50 frames. ], batch size: 135, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:56:00,585 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193898.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:56:05,931 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=193902.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:56:42,165 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=193931.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 13:56:44,218 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.4810, 5.0741, 3.1529, 4.4256, 1.5858, 5.0883, 4.9108, 5.0439], + device='cuda:3'), covar=tensor([0.0401, 0.0760, 0.1718, 0.0737, 0.3451, 0.0568, 0.0848, 0.1137], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0431, 0.0520, 0.0358, 0.0412, 0.0457, 0.0453, 0.0485], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:56:45,191 INFO [train.py:903] (3/4) Epoch 29, batch 2750, loss[loss=0.1845, simple_loss=0.2716, pruned_loss=0.04869, over 17362.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2834, pruned_loss=0.05966, over 3799693.55 frames. ], batch size: 101, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:56:58,025 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=193944.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:57:27,813 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=193969.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:57:36,672 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=193977.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:57:44,123 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.356e+02 4.680e+02 5.858e+02 7.532e+02 1.982e+03, threshold=1.172e+03, percent-clipped=4.0 +2023-04-03 13:57:45,310 INFO [train.py:903] (3/4) Epoch 29, batch 2800, loss[loss=0.2025, simple_loss=0.2838, pruned_loss=0.06066, over 19340.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2838, pruned_loss=0.06035, over 3790889.86 frames. ], batch size: 66, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:57:55,488 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8440, 1.3472, 1.0814, 1.0196, 1.1698, 1.0286, 0.9063, 1.2411], + device='cuda:3'), covar=tensor([0.0759, 0.0913, 0.1224, 0.0886, 0.0624, 0.1409, 0.0685, 0.0563], + device='cuda:3'), in_proj_covar=tensor([0.0301, 0.0318, 0.0341, 0.0272, 0.0252, 0.0344, 0.0291, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 13:58:26,695 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194017.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:58:48,020 INFO [train.py:903] (3/4) Epoch 29, batch 2850, loss[loss=0.2074, simple_loss=0.2762, pruned_loss=0.06931, over 17749.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2836, pruned_loss=0.06022, over 3792520.81 frames. ], batch size: 39, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:58:58,385 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194042.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:59:45,136 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 13:59:47,410 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.156e+02 4.652e+02 5.661e+02 7.187e+02 2.087e+03, threshold=1.132e+03, percent-clipped=4.0 +2023-04-03 13:59:48,552 INFO [train.py:903] (3/4) Epoch 29, batch 2900, loss[loss=0.196, simple_loss=0.2907, pruned_loss=0.0506, over 19684.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.283, pruned_loss=0.06029, over 3793692.67 frames. ], batch size: 59, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 13:59:55,306 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194090.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 13:59:57,653 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194092.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:00:47,881 INFO [train.py:903] (3/4) Epoch 29, batch 2950, loss[loss=0.1966, simple_loss=0.288, pruned_loss=0.05263, over 19543.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2832, pruned_loss=0.06071, over 3799059.89 frames. ], batch size: 56, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 14:01:13,365 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194154.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:01:42,969 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194179.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:01:46,911 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.793e+02 4.745e+02 5.839e+02 7.587e+02 1.918e+03, threshold=1.168e+03, percent-clipped=4.0 +2023-04-03 14:01:48,097 INFO [train.py:903] (3/4) Epoch 29, batch 3000, loss[loss=0.196, simple_loss=0.2832, pruned_loss=0.05438, over 19794.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2815, pruned_loss=0.05971, over 3802800.96 frames. ], batch size: 56, lr: 2.83e-03, grad_scale: 8.0 +2023-04-03 14:01:48,097 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 14:02:01,076 INFO [train.py:937] (3/4) Epoch 29, validation: loss=0.1668, simple_loss=0.2661, pruned_loss=0.03375, over 944034.00 frames. +2023-04-03 14:02:01,077 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 14:02:02,348 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 14:02:05,062 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194187.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 14:02:36,777 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194212.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:03:01,745 INFO [train.py:903] (3/4) Epoch 29, batch 3050, loss[loss=0.2147, simple_loss=0.2998, pruned_loss=0.0648, over 18781.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2816, pruned_loss=0.05901, over 3814451.79 frames. ], batch size: 74, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:03:14,228 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.78 vs. limit=5.0 +2023-04-03 14:03:17,761 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194246.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:03:45,897 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.2833, 2.9813, 2.5205, 2.5738, 2.2203, 2.5694, 1.0884, 2.1859], + device='cuda:3'), covar=tensor([0.0711, 0.0598, 0.0622, 0.1020, 0.1126, 0.1172, 0.1466, 0.1116], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0366, 0.0370, 0.0394, 0.0475, 0.0401, 0.0349, 0.0350], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 14:03:54,724 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2812, 2.3215, 2.6135, 2.9812, 2.2919, 2.8407, 2.6022, 2.4516], + device='cuda:3'), covar=tensor([0.4475, 0.4558, 0.2015, 0.2849, 0.4945, 0.2512, 0.4969, 0.3391], + device='cuda:3'), in_proj_covar=tensor([0.0943, 0.1021, 0.0747, 0.0955, 0.0919, 0.0858, 0.0863, 0.0811], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:04:02,892 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.301e+02 5.546e+02 6.449e+02 7.988e+02 2.570e+03, threshold=1.290e+03, percent-clipped=8.0 +2023-04-03 14:04:03,997 INFO [train.py:903] (3/4) Epoch 29, batch 3100, loss[loss=0.2319, simple_loss=0.3014, pruned_loss=0.08124, over 13186.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.282, pruned_loss=0.05931, over 3797384.10 frames. ], batch size: 136, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:05:04,157 INFO [train.py:903] (3/4) Epoch 29, batch 3150, loss[loss=0.2868, simple_loss=0.355, pruned_loss=0.1093, over 19364.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2838, pruned_loss=0.06034, over 3796579.38 frames. ], batch size: 66, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:05:06,674 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194336.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:05:19,984 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194348.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:05:27,478 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 14:05:36,271 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194361.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:05:41,708 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1939, 1.3331, 1.7448, 0.9715, 2.4166, 3.1185, 2.8032, 3.2625], + device='cuda:3'), covar=tensor([0.1546, 0.3833, 0.3354, 0.2793, 0.0607, 0.0227, 0.0254, 0.0335], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0333, 0.0366, 0.0273, 0.0256, 0.0198, 0.0220, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 14:05:49,289 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194371.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:05:51,756 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194373.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:05:59,649 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194380.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:06:02,722 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.204e+02 4.902e+02 5.790e+02 7.151e+02 3.080e+03, threshold=1.158e+03, percent-clipped=4.0 +2023-04-03 14:06:03,888 INFO [train.py:903] (3/4) Epoch 29, batch 3200, loss[loss=0.2284, simple_loss=0.3045, pruned_loss=0.07613, over 18047.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2835, pruned_loss=0.06052, over 3801283.43 frames. ], batch size: 83, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:06:49,303 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4178, 2.1638, 1.6596, 1.4759, 2.0222, 1.3979, 1.3392, 1.9215], + device='cuda:3'), covar=tensor([0.1047, 0.0788, 0.1041, 0.0929, 0.0575, 0.1322, 0.0726, 0.0472], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0320, 0.0343, 0.0273, 0.0253, 0.0346, 0.0291, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:07:04,918 INFO [train.py:903] (3/4) Epoch 29, batch 3250, loss[loss=0.1783, simple_loss=0.2659, pruned_loss=0.04532, over 19600.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2831, pruned_loss=0.06003, over 3806350.62 frames. ], batch size: 50, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:07:05,062 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194434.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:07:23,495 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3720, 1.4932, 1.7962, 1.6062, 2.5054, 2.0348, 2.7041, 1.1822], + device='cuda:3'), covar=tensor([0.2688, 0.4252, 0.2655, 0.2162, 0.1497, 0.2424, 0.1275, 0.4901], + device='cuda:3'), in_proj_covar=tensor([0.0557, 0.0675, 0.0763, 0.0511, 0.0638, 0.0549, 0.0672, 0.0576], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:08:04,290 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.869e+02 4.916e+02 6.594e+02 9.076e+02 2.390e+03, threshold=1.319e+03, percent-clipped=9.0 +2023-04-03 14:08:05,462 INFO [train.py:903] (3/4) Epoch 29, batch 3300, loss[loss=0.1475, simple_loss=0.2207, pruned_loss=0.03717, over 19040.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.283, pruned_loss=0.06003, over 3820164.43 frames. ], batch size: 42, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:08:09,431 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 14:09:07,373 INFO [train.py:903] (3/4) Epoch 29, batch 3350, loss[loss=0.2125, simple_loss=0.294, pruned_loss=0.06546, over 19661.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2832, pruned_loss=0.06039, over 3805816.33 frames. ], batch size: 55, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:09:24,564 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194549.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:10:06,127 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.313e+02 5.166e+02 6.518e+02 8.363e+02 1.379e+03, threshold=1.304e+03, percent-clipped=1.0 +2023-04-03 14:10:07,296 INFO [train.py:903] (3/4) Epoch 29, batch 3400, loss[loss=0.2064, simple_loss=0.2957, pruned_loss=0.0586, over 19304.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2836, pruned_loss=0.06027, over 3816145.54 frames. ], batch size: 66, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:10:48,330 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194617.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:11:07,050 INFO [train.py:903] (3/4) Epoch 29, batch 3450, loss[loss=0.1972, simple_loss=0.2868, pruned_loss=0.05386, over 19535.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2836, pruned_loss=0.06029, over 3809642.23 frames. ], batch size: 56, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:11:09,272 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 14:11:17,047 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194642.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:11:29,993 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194652.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:12:02,512 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194680.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:12:05,594 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.890e+02 4.627e+02 5.712e+02 7.363e+02 1.612e+03, threshold=1.142e+03, percent-clipped=3.0 +2023-04-03 14:12:06,669 INFO [train.py:903] (3/4) Epoch 29, batch 3500, loss[loss=0.1958, simple_loss=0.2766, pruned_loss=0.05754, over 19724.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2824, pruned_loss=0.06001, over 3815046.21 frames. ], batch size: 51, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:12:43,655 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194715.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:12:54,394 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194724.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:13:06,886 INFO [train.py:903] (3/4) Epoch 29, batch 3550, loss[loss=0.2501, simple_loss=0.3236, pruned_loss=0.08825, over 18141.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2819, pruned_loss=0.05936, over 3819831.40 frames. ], batch size: 83, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:14:06,358 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.972e+02 4.806e+02 5.826e+02 6.989e+02 1.690e+03, threshold=1.165e+03, percent-clipped=2.0 +2023-04-03 14:14:07,571 INFO [train.py:903] (3/4) Epoch 29, batch 3600, loss[loss=0.2148, simple_loss=0.2964, pruned_loss=0.06662, over 18388.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2813, pruned_loss=0.05887, over 3813850.18 frames. ], batch size: 84, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:14:20,128 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194795.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:14:20,699 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.70 vs. limit=2.0 +2023-04-03 14:14:30,671 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2225, 1.1797, 1.2224, 1.4300, 1.0703, 1.2250, 1.2894, 1.2369], + device='cuda:3'), covar=tensor([0.0967, 0.1001, 0.1123, 0.0655, 0.0875, 0.0918, 0.0838, 0.0852], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0229, 0.0240, 0.0225, 0.0214, 0.0188, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 14:14:33,723 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=194805.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:15:02,762 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194830.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:15:02,804 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=194830.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:15:06,687 INFO [train.py:903] (3/4) Epoch 29, batch 3650, loss[loss=0.2401, simple_loss=0.3136, pruned_loss=0.08326, over 18796.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2822, pruned_loss=0.05941, over 3818071.05 frames. ], batch size: 74, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:15:12,779 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=194839.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:15:45,396 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=194865.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:16:05,568 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.663e+02 5.135e+02 6.261e+02 7.520e+02 2.080e+03, threshold=1.252e+03, percent-clipped=4.0 +2023-04-03 14:16:06,486 INFO [train.py:903] (3/4) Epoch 29, batch 3700, loss[loss=0.2306, simple_loss=0.3095, pruned_loss=0.07582, over 19661.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2827, pruned_loss=0.05986, over 3823831.09 frames. ], batch size: 58, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:16:20,683 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5835, 1.2615, 1.4957, 1.2589, 2.2740, 1.0908, 2.1192, 2.5545], + device='cuda:3'), covar=tensor([0.0722, 0.2933, 0.2880, 0.1818, 0.0835, 0.2201, 0.1155, 0.0435], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0382, 0.0402, 0.0356, 0.0387, 0.0361, 0.0400, 0.0421], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:17:08,165 INFO [train.py:903] (3/4) Epoch 29, batch 3750, loss[loss=0.1857, simple_loss=0.274, pruned_loss=0.04871, over 18800.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2817, pruned_loss=0.05947, over 3822300.09 frames. ], batch size: 74, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:18:06,318 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.993e+02 4.741e+02 6.225e+02 7.602e+02 2.236e+03, threshold=1.245e+03, percent-clipped=10.0 +2023-04-03 14:18:07,441 INFO [train.py:903] (3/4) Epoch 29, batch 3800, loss[loss=0.2268, simple_loss=0.3075, pruned_loss=0.07311, over 17447.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2832, pruned_loss=0.06022, over 3823892.80 frames. ], batch size: 101, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:18:21,207 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=194996.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:18:34,674 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 14:19:01,234 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=195029.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:19:06,523 INFO [train.py:903] (3/4) Epoch 29, batch 3850, loss[loss=0.1996, simple_loss=0.2635, pruned_loss=0.06781, over 19778.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.284, pruned_loss=0.0604, over 3819533.96 frames. ], batch size: 47, lr: 2.82e-03, grad_scale: 4.0 +2023-04-03 14:19:21,441 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-04-03 14:19:27,256 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195051.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:19:41,455 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=195062.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:19:45,042 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2575, 1.3117, 1.2972, 1.1236, 1.0841, 1.1333, 0.1052, 0.4173], + device='cuda:3'), covar=tensor([0.0826, 0.0754, 0.0514, 0.0637, 0.1403, 0.0726, 0.1441, 0.1215], + device='cuda:3'), in_proj_covar=tensor([0.0371, 0.0369, 0.0374, 0.0398, 0.0478, 0.0403, 0.0350, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 14:19:56,964 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195076.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:20:06,401 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.076e+02 4.743e+02 5.822e+02 7.676e+02 1.767e+03, threshold=1.164e+03, percent-clipped=5.0 +2023-04-03 14:20:06,418 INFO [train.py:903] (3/4) Epoch 29, batch 3900, loss[loss=0.2027, simple_loss=0.2959, pruned_loss=0.0547, over 19711.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2832, pruned_loss=0.06011, over 3824762.76 frames. ], batch size: 59, lr: 2.82e-03, grad_scale: 4.0 +2023-04-03 14:20:08,924 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195086.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:20:13,202 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=195089.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:20:21,710 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195095.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 14:20:39,828 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195111.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:20:39,865 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195111.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:20:50,176 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195120.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:21:07,592 INFO [train.py:903] (3/4) Epoch 29, batch 3950, loss[loss=0.2177, simple_loss=0.2983, pruned_loss=0.06849, over 19525.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.282, pruned_loss=0.05938, over 3824424.37 frames. ], batch size: 64, lr: 2.82e-03, grad_scale: 4.0 +2023-04-03 14:21:12,683 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 14:21:20,032 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-04-03 14:22:08,284 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.119e+02 4.632e+02 5.828e+02 7.591e+02 1.503e+03, threshold=1.166e+03, percent-clipped=4.0 +2023-04-03 14:22:08,302 INFO [train.py:903] (3/4) Epoch 29, batch 4000, loss[loss=0.1667, simple_loss=0.2533, pruned_loss=0.04009, over 19664.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2828, pruned_loss=0.05953, over 3820431.35 frames. ], batch size: 53, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:22:10,980 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9071, 1.1014, 1.4935, 0.8148, 2.0671, 2.4153, 2.2322, 2.7516], + device='cuda:3'), covar=tensor([0.1826, 0.5245, 0.4538, 0.2863, 0.0742, 0.0360, 0.0406, 0.0424], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0334, 0.0367, 0.0273, 0.0258, 0.0198, 0.0221, 0.0278], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 14:22:38,726 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=195209.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:22:38,979 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9648, 1.7115, 1.6414, 1.9464, 1.7014, 1.6390, 1.5668, 1.8390], + device='cuda:3'), covar=tensor([0.1133, 0.1533, 0.1592, 0.1050, 0.1352, 0.0625, 0.1583, 0.0817], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0359, 0.0318, 0.0257, 0.0307, 0.0258, 0.0322, 0.0260], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 14:22:52,716 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 14:23:07,865 INFO [train.py:903] (3/4) Epoch 29, batch 4050, loss[loss=0.2009, simple_loss=0.2808, pruned_loss=0.06043, over 19622.00 frames. ], tot_loss[loss=0.2026, simple_loss=0.2841, pruned_loss=0.0606, over 3815011.63 frames. ], batch size: 50, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:24:08,397 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.197e+02 5.118e+02 6.147e+02 8.377e+02 1.783e+03, threshold=1.229e+03, percent-clipped=5.0 +2023-04-03 14:24:08,416 INFO [train.py:903] (3/4) Epoch 29, batch 4100, loss[loss=0.2265, simple_loss=0.3043, pruned_loss=0.07437, over 19805.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.284, pruned_loss=0.06021, over 3811679.76 frames. ], batch size: 56, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:24:41,892 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 14:24:45,299 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3702, 1.3984, 1.7269, 1.6824, 2.3441, 2.0367, 2.3374, 0.9915], + device='cuda:3'), covar=tensor([0.2986, 0.5076, 0.3186, 0.2328, 0.1847, 0.2860, 0.2041, 0.5526], + device='cuda:3'), in_proj_covar=tensor([0.0560, 0.0678, 0.0769, 0.0515, 0.0641, 0.0552, 0.0677, 0.0579], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:24:57,114 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195324.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:25:08,843 INFO [train.py:903] (3/4) Epoch 29, batch 4150, loss[loss=0.1916, simple_loss=0.2841, pruned_loss=0.04958, over 19349.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2837, pruned_loss=0.06002, over 3821802.78 frames. ], batch size: 70, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:25:31,184 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0708, 1.9843, 1.7816, 2.1493, 1.7874, 1.7626, 1.7518, 2.0093], + device='cuda:3'), covar=tensor([0.1150, 0.1476, 0.1560, 0.1079, 0.1457, 0.0615, 0.1476, 0.0771], + device='cuda:3'), in_proj_covar=tensor([0.0278, 0.0358, 0.0317, 0.0256, 0.0307, 0.0257, 0.0322, 0.0259], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:25:49,118 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195367.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:25:56,332 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=195373.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:26:10,052 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.652e+02 4.925e+02 6.547e+02 8.865e+02 3.200e+03, threshold=1.309e+03, percent-clipped=9.0 +2023-04-03 14:26:10,070 INFO [train.py:903] (3/4) Epoch 29, batch 4200, loss[loss=0.1998, simple_loss=0.2841, pruned_loss=0.05774, over 17218.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2837, pruned_loss=0.0605, over 3826263.12 frames. ], batch size: 101, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:26:13,437 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 14:26:19,382 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195392.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:26:32,833 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0763, 2.0553, 2.2527, 2.1663, 2.9435, 2.5988, 2.9902, 2.0726], + device='cuda:3'), covar=tensor([0.2043, 0.3575, 0.2296, 0.1645, 0.1330, 0.1939, 0.1353, 0.4024], + device='cuda:3'), in_proj_covar=tensor([0.0559, 0.0677, 0.0766, 0.0514, 0.0639, 0.0551, 0.0675, 0.0578], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:26:35,645 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=195406.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 14:26:40,932 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=195410.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:27:07,662 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=195433.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:27:08,636 INFO [train.py:903] (3/4) Epoch 29, batch 4250, loss[loss=0.1676, simple_loss=0.2432, pruned_loss=0.04607, over 19376.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2851, pruned_loss=0.06141, over 3810085.33 frames. ], batch size: 47, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:27:22,566 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 14:27:32,633 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 14:28:07,588 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5271, 2.2963, 1.6567, 1.5438, 2.0956, 1.3613, 1.3712, 1.9878], + device='cuda:3'), covar=tensor([0.1293, 0.0913, 0.1197, 0.0969, 0.0660, 0.1438, 0.0926, 0.0605], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0323, 0.0345, 0.0274, 0.0255, 0.0349, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:28:08,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.783e+02 4.660e+02 5.852e+02 7.750e+02 1.648e+03, threshold=1.170e+03, percent-clipped=3.0 +2023-04-03 14:28:08,347 INFO [train.py:903] (3/4) Epoch 29, batch 4300, loss[loss=0.1845, simple_loss=0.2787, pruned_loss=0.04518, over 19524.00 frames. ], tot_loss[loss=0.2022, simple_loss=0.2837, pruned_loss=0.06034, over 3817288.56 frames. ], batch size: 56, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:28:14,173 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195488.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:28:32,657 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3702, 3.8728, 3.9752, 3.9868, 1.6256, 3.7893, 3.3203, 3.7486], + device='cuda:3'), covar=tensor([0.1686, 0.0907, 0.0709, 0.0780, 0.5882, 0.1141, 0.0735, 0.1158], + device='cuda:3'), in_proj_covar=tensor([0.0829, 0.0796, 0.1007, 0.0881, 0.0874, 0.0772, 0.0596, 0.0936], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 14:28:54,514 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195521.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:29:01,491 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 14:29:09,910 INFO [train.py:903] (3/4) Epoch 29, batch 4350, loss[loss=0.1709, simple_loss=0.2578, pruned_loss=0.04204, over 19783.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2833, pruned_loss=0.06015, over 3817697.76 frames. ], batch size: 49, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:29:28,193 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195548.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:30:05,596 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195580.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:30:10,384 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.562e+02 4.994e+02 6.103e+02 7.630e+02 1.747e+03, threshold=1.221e+03, percent-clipped=1.0 +2023-04-03 14:30:10,402 INFO [train.py:903] (3/4) Epoch 29, batch 4400, loss[loss=0.26, simple_loss=0.3317, pruned_loss=0.09418, over 19321.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2826, pruned_loss=0.06025, over 3819801.44 frames. ], batch size: 66, lr: 2.82e-03, grad_scale: 8.0 +2023-04-03 14:30:24,505 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5709, 2.3536, 1.7117, 1.5541, 2.1760, 1.4437, 1.4014, 2.0219], + device='cuda:3'), covar=tensor([0.1155, 0.0790, 0.1152, 0.0940, 0.0608, 0.1375, 0.0905, 0.0547], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0321, 0.0343, 0.0273, 0.0254, 0.0347, 0.0291, 0.0275], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:30:29,900 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 14:30:35,588 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195605.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:30:38,553 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 14:31:09,646 INFO [train.py:903] (3/4) Epoch 29, batch 4450, loss[loss=0.1852, simple_loss=0.2593, pruned_loss=0.05558, over 19376.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2836, pruned_loss=0.06057, over 3833300.46 frames. ], batch size: 47, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:32:08,773 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.510e+02 5.083e+02 5.988e+02 7.824e+02 1.664e+03, threshold=1.198e+03, percent-clipped=3.0 +2023-04-03 14:32:08,796 INFO [train.py:903] (3/4) Epoch 29, batch 4500, loss[loss=0.241, simple_loss=0.3126, pruned_loss=0.08469, over 19748.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2833, pruned_loss=0.06072, over 3826540.58 frames. ], batch size: 63, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:33:10,071 INFO [train.py:903] (3/4) Epoch 29, batch 4550, loss[loss=0.219, simple_loss=0.3096, pruned_loss=0.06415, over 19528.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2843, pruned_loss=0.06127, over 3821372.05 frames. ], batch size: 56, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:33:15,624 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 14:33:21,610 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195744.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:33:33,977 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=195754.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:33:39,344 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 14:33:51,778 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195769.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:34:01,505 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195777.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:34:08,707 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.333e+02 5.047e+02 5.905e+02 8.200e+02 1.793e+03, threshold=1.181e+03, percent-clipped=7.0 +2023-04-03 14:34:08,728 INFO [train.py:903] (3/4) Epoch 29, batch 4600, loss[loss=0.2065, simple_loss=0.2948, pruned_loss=0.0591, over 19368.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2844, pruned_loss=0.06173, over 3826262.45 frames. ], batch size: 66, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:34:31,284 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195802.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 14:34:33,627 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=195804.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:35:05,200 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=195829.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:35:10,606 INFO [train.py:903] (3/4) Epoch 29, batch 4650, loss[loss=0.1853, simple_loss=0.2729, pruned_loss=0.04885, over 19534.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2852, pruned_loss=0.06202, over 3811433.03 frames. ], batch size: 56, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:35:22,177 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 14:35:33,379 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 14:35:38,583 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6526, 1.4570, 1.5641, 2.2105, 1.6576, 1.8224, 1.7961, 1.6240], + device='cuda:3'), covar=tensor([0.0973, 0.1114, 0.1099, 0.0735, 0.1026, 0.0915, 0.1003, 0.0878], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0225, 0.0229, 0.0241, 0.0227, 0.0214, 0.0189, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 14:35:54,064 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=195869.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:36:10,282 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.073e+02 5.115e+02 6.253e+02 8.443e+02 2.371e+03, threshold=1.251e+03, percent-clipped=7.0 +2023-04-03 14:36:10,300 INFO [train.py:903] (3/4) Epoch 29, batch 4700, loss[loss=0.2269, simple_loss=0.3013, pruned_loss=0.0763, over 13102.00 frames. ], tot_loss[loss=0.2043, simple_loss=0.2854, pruned_loss=0.06161, over 3811455.01 frames. ], batch size: 136, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:36:29,177 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 14:36:33,172 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.08 vs. limit=5.0 +2023-04-03 14:36:48,437 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-04-03 14:37:03,934 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.30 vs. limit=5.0 +2023-04-03 14:37:11,423 INFO [train.py:903] (3/4) Epoch 29, batch 4750, loss[loss=0.1624, simple_loss=0.2361, pruned_loss=0.04437, over 19746.00 frames. ], tot_loss[loss=0.2025, simple_loss=0.2837, pruned_loss=0.06064, over 3814499.31 frames. ], batch size: 46, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:37:53,318 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3217, 1.8634, 1.5579, 1.2232, 1.7572, 1.2108, 1.2941, 1.7453], + device='cuda:3'), covar=tensor([0.1006, 0.0940, 0.0981, 0.0999, 0.0614, 0.1303, 0.0766, 0.0484], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0321, 0.0343, 0.0273, 0.0254, 0.0347, 0.0291, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:38:11,673 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.651e+02 4.980e+02 6.350e+02 8.555e+02 2.103e+03, threshold=1.270e+03, percent-clipped=3.0 +2023-04-03 14:38:11,690 INFO [train.py:903] (3/4) Epoch 29, batch 4800, loss[loss=0.2005, simple_loss=0.2913, pruned_loss=0.05483, over 19355.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2826, pruned_loss=0.05985, over 3826185.57 frames. ], batch size: 66, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:39:14,512 INFO [train.py:903] (3/4) Epoch 29, batch 4850, loss[loss=0.2174, simple_loss=0.2983, pruned_loss=0.06818, over 19337.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.283, pruned_loss=0.0599, over 3836633.54 frames. ], batch size: 66, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:39:37,049 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 14:39:55,387 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 14:40:01,753 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 14:40:02,681 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 14:40:11,426 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 14:40:13,829 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.943e+02 4.641e+02 5.767e+02 7.663e+02 1.512e+03, threshold=1.153e+03, percent-clipped=2.0 +2023-04-03 14:40:13,852 INFO [train.py:903] (3/4) Epoch 29, batch 4900, loss[loss=0.1683, simple_loss=0.2537, pruned_loss=0.04143, over 19569.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2835, pruned_loss=0.06032, over 3820142.91 frames. ], batch size: 52, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:40:32,816 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 14:41:03,530 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=196125.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:41:14,060 INFO [train.py:903] (3/4) Epoch 29, batch 4950, loss[loss=0.2234, simple_loss=0.307, pruned_loss=0.06985, over 19664.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.2841, pruned_loss=0.06023, over 3824788.72 frames. ], batch size: 60, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:41:18,075 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.4360, 2.5273, 2.7507, 3.2259, 2.5259, 3.1612, 2.6887, 2.4324], + device='cuda:3'), covar=tensor([0.4287, 0.4249, 0.1976, 0.2609, 0.4237, 0.2192, 0.5254, 0.3596], + device='cuda:3'), in_proj_covar=tensor([0.0947, 0.1029, 0.0751, 0.0962, 0.0927, 0.0865, 0.0866, 0.0815], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:41:31,258 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 14:41:33,821 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=196150.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:41:54,186 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 14:42:14,323 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.427e+02 4.864e+02 6.197e+02 7.725e+02 1.739e+03, threshold=1.239e+03, percent-clipped=10.0 +2023-04-03 14:42:14,341 INFO [train.py:903] (3/4) Epoch 29, batch 5000, loss[loss=0.2247, simple_loss=0.2941, pruned_loss=0.07761, over 19676.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2833, pruned_loss=0.06007, over 3825081.52 frames. ], batch size: 53, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:42:23,520 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 14:42:33,258 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9334, 2.0243, 2.2471, 2.4216, 1.8537, 2.3743, 2.1960, 2.1348], + device='cuda:3'), covar=tensor([0.4264, 0.4127, 0.2102, 0.2659, 0.4334, 0.2364, 0.5265, 0.3478], + device='cuda:3'), in_proj_covar=tensor([0.0945, 0.1026, 0.0749, 0.0959, 0.0924, 0.0863, 0.0864, 0.0813], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:42:35,050 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 14:43:14,851 INFO [train.py:903] (3/4) Epoch 29, batch 5050, loss[loss=0.2251, simple_loss=0.3003, pruned_loss=0.07499, over 13645.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2829, pruned_loss=0.05987, over 3824598.20 frames. ], batch size: 138, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:43:49,695 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 14:44:15,610 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.659e+02 4.774e+02 5.782e+02 7.176e+02 1.455e+03, threshold=1.156e+03, percent-clipped=5.0 +2023-04-03 14:44:15,632 INFO [train.py:903] (3/4) Epoch 29, batch 5100, loss[loss=0.1819, simple_loss=0.2682, pruned_loss=0.04775, over 19675.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2826, pruned_loss=0.05935, over 3829392.26 frames. ], batch size: 53, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:44:24,668 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 14:44:27,980 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 14:44:32,507 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 14:45:16,435 INFO [train.py:903] (3/4) Epoch 29, batch 5150, loss[loss=0.1846, simple_loss=0.2687, pruned_loss=0.0503, over 19421.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2833, pruned_loss=0.05959, over 3829034.51 frames. ], batch size: 48, lr: 2.81e-03, grad_scale: 4.0 +2023-04-03 14:45:23,033 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196339.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:45:26,218 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 14:45:59,978 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 14:46:06,892 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4641, 1.5147, 1.7519, 1.6936, 2.4620, 2.1616, 2.6583, 1.1336], + device='cuda:3'), covar=tensor([0.2612, 0.4622, 0.2881, 0.2100, 0.1605, 0.2349, 0.1491, 0.4866], + device='cuda:3'), in_proj_covar=tensor([0.0557, 0.0677, 0.0765, 0.0514, 0.0640, 0.0550, 0.0672, 0.0578], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:46:10,269 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196378.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:46:17,532 INFO [train.py:903] (3/4) Epoch 29, batch 5200, loss[loss=0.1897, simple_loss=0.2539, pruned_loss=0.06272, over 19726.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.282, pruned_loss=0.05937, over 3830081.35 frames. ], batch size: 46, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:46:18,463 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.574e+02 5.187e+02 6.337e+02 8.295e+02 1.863e+03, threshold=1.267e+03, percent-clipped=5.0 +2023-04-03 14:46:28,788 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 14:47:12,670 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 14:47:19,220 INFO [train.py:903] (3/4) Epoch 29, batch 5250, loss[loss=0.1837, simple_loss=0.2698, pruned_loss=0.04881, over 19766.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2816, pruned_loss=0.05881, over 3837517.26 frames. ], batch size: 54, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:48:20,445 INFO [train.py:903] (3/4) Epoch 29, batch 5300, loss[loss=0.2041, simple_loss=0.3022, pruned_loss=0.05303, over 19533.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2829, pruned_loss=0.05922, over 3832252.78 frames. ], batch size: 56, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:48:21,523 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.666e+02 4.792e+02 5.709e+02 7.130e+02 1.478e+03, threshold=1.142e+03, percent-clipped=2.0 +2023-04-03 14:48:35,852 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 14:49:20,281 INFO [train.py:903] (3/4) Epoch 29, batch 5350, loss[loss=0.2021, simple_loss=0.285, pruned_loss=0.0596, over 19666.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2823, pruned_loss=0.05908, over 3837916.98 frames. ], batch size: 53, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:49:51,996 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 14:50:21,629 INFO [train.py:903] (3/4) Epoch 29, batch 5400, loss[loss=0.1832, simple_loss=0.2753, pruned_loss=0.04556, over 19527.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2828, pruned_loss=0.05949, over 3823483.11 frames. ], batch size: 54, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:50:22,761 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.031e+02 4.834e+02 5.946e+02 7.840e+02 1.782e+03, threshold=1.189e+03, percent-clipped=5.0 +2023-04-03 14:50:36,285 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196596.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:51:22,486 INFO [train.py:903] (3/4) Epoch 29, batch 5450, loss[loss=0.1639, simple_loss=0.2379, pruned_loss=0.04496, over 19723.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2819, pruned_loss=0.05874, over 3826557.74 frames. ], batch size: 45, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:51:38,162 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8702, 1.2034, 1.5834, 0.5997, 1.9667, 2.4605, 2.1722, 2.6327], + device='cuda:3'), covar=tensor([0.1730, 0.4098, 0.3581, 0.3098, 0.0687, 0.0294, 0.0348, 0.0383], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0335, 0.0367, 0.0274, 0.0257, 0.0199, 0.0221, 0.0279], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 14:51:45,536 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7825, 4.2366, 4.4672, 4.4881, 1.5792, 4.2119, 3.6146, 4.2016], + device='cuda:3'), covar=tensor([0.1723, 0.0964, 0.0662, 0.0766, 0.6645, 0.1059, 0.0759, 0.1182], + device='cuda:3'), in_proj_covar=tensor([0.0830, 0.0803, 0.1014, 0.0887, 0.0880, 0.0775, 0.0598, 0.0939], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 14:52:21,558 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=196683.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:52:22,556 INFO [train.py:903] (3/4) Epoch 29, batch 5500, loss[loss=0.1909, simple_loss=0.2681, pruned_loss=0.05685, over 19617.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2825, pruned_loss=0.0594, over 3813041.00 frames. ], batch size: 50, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:52:23,692 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.565e+02 4.767e+02 5.823e+02 7.066e+02 1.237e+03, threshold=1.165e+03, percent-clipped=1.0 +2023-04-03 14:52:46,218 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 14:53:08,294 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=196722.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:53:22,918 INFO [train.py:903] (3/4) Epoch 29, batch 5550, loss[loss=0.1729, simple_loss=0.2528, pruned_loss=0.04645, over 19487.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2822, pruned_loss=0.05956, over 3825904.66 frames. ], batch size: 49, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:53:30,040 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 14:54:18,742 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 14:54:22,915 INFO [train.py:903] (3/4) Epoch 29, batch 5600, loss[loss=0.1733, simple_loss=0.2646, pruned_loss=0.04094, over 19528.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2819, pruned_loss=0.05934, over 3839982.98 frames. ], batch size: 54, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:54:24,078 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.131e+02 4.937e+02 5.977e+02 7.468e+02 1.263e+03, threshold=1.195e+03, percent-clipped=3.0 +2023-04-03 14:54:36,515 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3133, 2.1707, 2.0572, 1.9362, 1.7283, 1.9050, 0.7323, 1.3392], + device='cuda:3'), covar=tensor([0.0637, 0.0637, 0.0549, 0.0899, 0.1251, 0.0982, 0.1458, 0.1124], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0366, 0.0371, 0.0395, 0.0476, 0.0403, 0.0350, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 14:54:39,801 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196797.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:54:40,968 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=196798.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:55:24,436 INFO [train.py:903] (3/4) Epoch 29, batch 5650, loss[loss=0.207, simple_loss=0.2672, pruned_loss=0.07336, over 19309.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2811, pruned_loss=0.05898, over 3840770.25 frames. ], batch size: 44, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:55:28,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=196837.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:56:06,799 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9537, 2.0414, 2.2776, 2.5846, 1.9705, 2.4251, 2.2502, 2.1267], + device='cuda:3'), covar=tensor([0.4224, 0.4140, 0.1940, 0.2513, 0.4083, 0.2348, 0.5115, 0.3418], + device='cuda:3'), in_proj_covar=tensor([0.0946, 0.1026, 0.0751, 0.0958, 0.0926, 0.0865, 0.0866, 0.0814], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 14:56:10,126 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196872.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:56:10,956 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 14:56:24,237 INFO [train.py:903] (3/4) Epoch 29, batch 5700, loss[loss=0.1895, simple_loss=0.2793, pruned_loss=0.04981, over 19378.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2806, pruned_loss=0.05846, over 3837154.38 frames. ], batch size: 66, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:56:25,388 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.192e+02 5.159e+02 6.169e+02 7.777e+02 1.148e+03, threshold=1.234e+03, percent-clipped=0.0 +2023-04-03 14:56:28,884 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2527, 2.0261, 1.5731, 1.3614, 1.8225, 1.3005, 1.2400, 1.8577], + device='cuda:3'), covar=tensor([0.0943, 0.0786, 0.1019, 0.0869, 0.0572, 0.1257, 0.0736, 0.0440], + device='cuda:3'), in_proj_covar=tensor([0.0306, 0.0321, 0.0344, 0.0275, 0.0254, 0.0347, 0.0292, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 14:56:41,818 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=196899.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:57:04,568 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3011, 2.1876, 2.1004, 1.9276, 1.7786, 1.9030, 0.7553, 1.3055], + device='cuda:3'), covar=tensor([0.0750, 0.0723, 0.0526, 0.0968, 0.1260, 0.1108, 0.1522, 0.1226], + device='cuda:3'), in_proj_covar=tensor([0.0370, 0.0367, 0.0372, 0.0395, 0.0476, 0.0404, 0.0350, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 14:57:24,055 INFO [train.py:903] (3/4) Epoch 29, batch 5750, loss[loss=0.2273, simple_loss=0.3002, pruned_loss=0.07722, over 19857.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2812, pruned_loss=0.05897, over 3840262.97 frames. ], batch size: 52, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:57:25,244 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 14:57:30,928 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=196940.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:57:31,938 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 14:57:38,015 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 14:57:58,319 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.39 vs. limit=2.0 +2023-04-03 14:58:24,227 INFO [train.py:903] (3/4) Epoch 29, batch 5800, loss[loss=0.186, simple_loss=0.2584, pruned_loss=0.05682, over 19783.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2827, pruned_loss=0.05983, over 3811255.99 frames. ], batch size: 46, lr: 2.81e-03, grad_scale: 8.0 +2023-04-03 14:58:25,407 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.058e+02 4.987e+02 6.141e+02 8.368e+02 1.662e+03, threshold=1.228e+03, percent-clipped=5.0 +2023-04-03 14:59:24,878 INFO [train.py:903] (3/4) Epoch 29, batch 5850, loss[loss=0.1632, simple_loss=0.2424, pruned_loss=0.04195, over 19766.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2825, pruned_loss=0.05989, over 3817481.73 frames. ], batch size: 46, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 14:59:49,587 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197054.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 14:59:50,553 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197055.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:00:19,107 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197079.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:00:24,408 INFO [train.py:903] (3/4) Epoch 29, batch 5900, loss[loss=0.1798, simple_loss=0.2735, pruned_loss=0.04307, over 19665.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2833, pruned_loss=0.05991, over 3815888.11 frames. ], batch size: 60, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:00:25,528 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.208e+02 4.889e+02 5.905e+02 7.501e+02 2.168e+03, threshold=1.181e+03, percent-clipped=6.0 +2023-04-03 15:00:26,682 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 15:00:35,852 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197093.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:00:46,555 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 15:01:06,443 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197118.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:01:24,667 INFO [train.py:903] (3/4) Epoch 29, batch 5950, loss[loss=0.1712, simple_loss=0.2511, pruned_loss=0.0457, over 19625.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2827, pruned_loss=0.05964, over 3832559.79 frames. ], batch size: 50, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:01:32,742 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197141.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:01:37,748 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.61 vs. limit=2.0 +2023-04-03 15:02:06,322 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.2295, 2.3021, 2.5064, 3.0088, 2.3967, 2.8678, 2.4726, 2.3304], + device='cuda:3'), covar=tensor([0.4358, 0.4160, 0.1977, 0.2672, 0.4528, 0.2348, 0.5058, 0.3414], + device='cuda:3'), in_proj_covar=tensor([0.0947, 0.1026, 0.0750, 0.0959, 0.0926, 0.0863, 0.0865, 0.0814], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 15:02:12,822 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1463, 3.5108, 2.1482, 1.8368, 3.2582, 1.7650, 1.5163, 2.6256], + device='cuda:3'), covar=tensor([0.1311, 0.0590, 0.0957, 0.1154, 0.0530, 0.1287, 0.1047, 0.0517], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0321, 0.0345, 0.0275, 0.0254, 0.0348, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:02:24,684 INFO [train.py:903] (3/4) Epoch 29, batch 6000, loss[loss=0.1927, simple_loss=0.2784, pruned_loss=0.05343, over 19774.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2823, pruned_loss=0.05933, over 3830594.50 frames. ], batch size: 56, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:02:24,685 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 15:02:43,144 INFO [train.py:937] (3/4) Epoch 29, validation: loss=0.167, simple_loss=0.2662, pruned_loss=0.03392, over 944034.00 frames. +2023-04-03 15:02:43,145 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 15:02:44,354 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.479e+02 4.704e+02 5.833e+02 7.372e+02 1.660e+03, threshold=1.167e+03, percent-clipped=5.0 +2023-04-03 15:03:12,927 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0356, 1.8610, 1.6884, 2.0296, 1.7539, 1.6930, 1.6933, 1.9448], + device='cuda:3'), covar=tensor([0.1150, 0.1527, 0.1582, 0.1112, 0.1501, 0.0653, 0.1613, 0.0793], + device='cuda:3'), in_proj_covar=tensor([0.0279, 0.0359, 0.0319, 0.0257, 0.0308, 0.0259, 0.0323, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 15:03:23,061 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197216.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:03:44,115 INFO [train.py:903] (3/4) Epoch 29, batch 6050, loss[loss=0.1738, simple_loss=0.2555, pruned_loss=0.04601, over 19739.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2818, pruned_loss=0.0591, over 3822549.43 frames. ], batch size: 51, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:03:44,608 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5202, 1.5981, 1.9284, 1.8148, 2.6218, 2.3549, 2.8378, 1.2184], + device='cuda:3'), covar=tensor([0.2580, 0.4536, 0.2831, 0.1978, 0.1734, 0.2214, 0.1633, 0.5014], + device='cuda:3'), in_proj_covar=tensor([0.0560, 0.0680, 0.0769, 0.0516, 0.0643, 0.0551, 0.0676, 0.0580], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 15:03:56,028 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197243.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:03:56,327 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7696, 1.8883, 2.1398, 2.2855, 1.7684, 2.2244, 2.1149, 1.9522], + device='cuda:3'), covar=tensor([0.4504, 0.4075, 0.2160, 0.2610, 0.4178, 0.2423, 0.5364, 0.3734], + device='cuda:3'), in_proj_covar=tensor([0.0947, 0.1026, 0.0750, 0.0959, 0.0926, 0.0864, 0.0866, 0.0815], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 15:04:11,916 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197256.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:04:45,072 INFO [train.py:903] (3/4) Epoch 29, batch 6100, loss[loss=0.1801, simple_loss=0.2469, pruned_loss=0.05663, over 19734.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2818, pruned_loss=0.05907, over 3822021.64 frames. ], batch size: 46, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:04:46,195 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.008e+02 5.045e+02 6.568e+02 8.070e+02 1.422e+03, threshold=1.314e+03, percent-clipped=5.0 +2023-04-03 15:05:17,481 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197311.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:05:42,397 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197331.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:05:45,467 INFO [train.py:903] (3/4) Epoch 29, batch 6150, loss[loss=0.2175, simple_loss=0.2965, pruned_loss=0.06929, over 17232.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2825, pruned_loss=0.05945, over 3818893.84 frames. ], batch size: 100, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:05:48,251 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197336.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:06:13,324 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 15:06:14,674 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197358.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:06:46,126 INFO [train.py:903] (3/4) Epoch 29, batch 6200, loss[loss=0.2071, simple_loss=0.2865, pruned_loss=0.06383, over 19780.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.2834, pruned_loss=0.06021, over 3812503.57 frames. ], batch size: 54, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:06:47,112 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.451e+02 5.150e+02 5.975e+02 7.505e+02 2.002e+03, threshold=1.195e+03, percent-clipped=6.0 +2023-04-03 15:07:32,512 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8860, 1.8677, 1.8689, 1.6854, 1.5188, 1.6375, 0.5171, 0.9706], + device='cuda:3'), covar=tensor([0.0763, 0.0711, 0.0497, 0.0860, 0.1361, 0.1020, 0.1495, 0.1247], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0364, 0.0369, 0.0394, 0.0472, 0.0401, 0.0347, 0.0349], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 15:07:45,746 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=197433.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:07:46,598 INFO [train.py:903] (3/4) Epoch 29, batch 6250, loss[loss=0.1701, simple_loss=0.2489, pruned_loss=0.04565, over 19473.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2818, pruned_loss=0.05969, over 3806796.06 frames. ], batch size: 49, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:08:18,023 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 15:08:48,228 INFO [train.py:903] (3/4) Epoch 29, batch 6300, loss[loss=0.2219, simple_loss=0.2997, pruned_loss=0.07206, over 18118.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2825, pruned_loss=0.05964, over 3810415.25 frames. ], batch size: 83, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:08:50,595 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.231e+02 4.620e+02 5.575e+02 6.491e+02 1.508e+03, threshold=1.115e+03, percent-clipped=2.0 +2023-04-03 15:09:13,181 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.36 vs. limit=5.0 +2023-04-03 15:09:21,983 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197512.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:09:49,200 INFO [train.py:903] (3/4) Epoch 29, batch 6350, loss[loss=0.1943, simple_loss=0.2824, pruned_loss=0.05308, over 19544.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2814, pruned_loss=0.05846, over 3825537.97 frames. ], batch size: 61, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:09:49,488 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=197534.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:09:52,893 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197537.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:10:23,567 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=197563.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:10:49,102 INFO [train.py:903] (3/4) Epoch 29, batch 6400, loss[loss=0.2045, simple_loss=0.2906, pruned_loss=0.05919, over 19638.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2801, pruned_loss=0.05794, over 3827710.19 frames. ], batch size: 55, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:10:52,231 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.915e+02 4.287e+02 5.708e+02 7.339e+02 1.464e+03, threshold=1.142e+03, percent-clipped=5.0 +2023-04-03 15:10:53,764 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197587.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:11:23,622 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197612.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:11:25,880 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=197614.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:11:36,491 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2268, 5.6782, 3.3242, 4.8906, 1.3354, 5.7778, 5.6172, 5.8116], + device='cuda:3'), covar=tensor([0.0316, 0.0732, 0.1666, 0.0730, 0.3806, 0.0531, 0.0807, 0.1010], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0433, 0.0521, 0.0357, 0.0410, 0.0458, 0.0453, 0.0487], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:11:49,891 INFO [train.py:903] (3/4) Epoch 29, batch 6450, loss[loss=0.2183, simple_loss=0.299, pruned_loss=0.06878, over 18115.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2808, pruned_loss=0.05814, over 3812216.60 frames. ], batch size: 83, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:11:56,924 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=197639.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:12:13,560 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2647, 1.3161, 1.2824, 1.0752, 1.1220, 1.1495, 0.0762, 0.3784], + device='cuda:3'), covar=tensor([0.0758, 0.0696, 0.0505, 0.0660, 0.1344, 0.0706, 0.1474, 0.1231], + device='cuda:3'), in_proj_covar=tensor([0.0366, 0.0362, 0.0367, 0.0393, 0.0470, 0.0397, 0.0345, 0.0347], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 15:12:37,434 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 15:12:49,913 INFO [train.py:903] (3/4) Epoch 29, batch 6500, loss[loss=0.2125, simple_loss=0.2945, pruned_loss=0.06525, over 19524.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2819, pruned_loss=0.05881, over 3810530.21 frames. ], batch size: 54, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:12:52,112 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.446e+02 4.970e+02 6.097e+02 8.194e+02 1.467e+03, threshold=1.219e+03, percent-clipped=8.0 +2023-04-03 15:12:58,634 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 15:13:50,301 INFO [train.py:903] (3/4) Epoch 29, batch 6550, loss[loss=0.1811, simple_loss=0.2565, pruned_loss=0.05286, over 19773.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2821, pruned_loss=0.05932, over 3814235.61 frames. ], batch size: 46, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:13:54,318 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-04-03 15:14:03,100 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.53 vs. limit=5.0 +2023-04-03 15:14:42,389 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197777.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:14:50,635 INFO [train.py:903] (3/4) Epoch 29, batch 6600, loss[loss=0.2202, simple_loss=0.2918, pruned_loss=0.07428, over 13317.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2831, pruned_loss=0.05947, over 3814375.10 frames. ], batch size: 135, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:14:54,078 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.141e+02 4.758e+02 5.997e+02 7.159e+02 2.116e+03, threshold=1.199e+03, percent-clipped=4.0 +2023-04-03 15:15:28,413 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.4630, 1.2618, 1.5143, 1.5824, 3.0820, 1.2159, 2.3214, 3.4462], + device='cuda:3'), covar=tensor([0.0566, 0.2972, 0.2952, 0.1896, 0.0681, 0.2478, 0.1354, 0.0286], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0380, 0.0400, 0.0355, 0.0384, 0.0358, 0.0399, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:15:51,169 INFO [train.py:903] (3/4) Epoch 29, batch 6650, loss[loss=0.213, simple_loss=0.2947, pruned_loss=0.06563, over 19597.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2832, pruned_loss=0.05963, over 3820079.02 frames. ], batch size: 57, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:16:45,534 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197878.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:16:52,896 INFO [train.py:903] (3/4) Epoch 29, batch 6700, loss[loss=0.1811, simple_loss=0.2785, pruned_loss=0.04185, over 19718.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2827, pruned_loss=0.0595, over 3807457.68 frames. ], batch size: 59, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:16:56,373 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.106e+02 5.093e+02 5.944e+02 7.634e+02 1.783e+03, threshold=1.189e+03, percent-clipped=4.0 +2023-04-03 15:17:02,452 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197892.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:17:15,694 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7524, 1.5287, 1.5866, 2.2632, 1.7680, 1.8912, 1.9769, 1.7249], + device='cuda:3'), covar=tensor([0.0838, 0.0937, 0.1020, 0.0698, 0.0825, 0.0772, 0.0874, 0.0753], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0225, 0.0230, 0.0243, 0.0228, 0.0215, 0.0189, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 15:17:19,985 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=197907.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:17:23,552 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=197910.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:17:24,791 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8162, 1.2136, 0.9850, 0.9225, 1.0616, 0.9218, 0.8983, 1.1182], + device='cuda:3'), covar=tensor([0.0714, 0.0857, 0.1100, 0.0772, 0.0617, 0.1325, 0.0641, 0.0559], + device='cuda:3'), in_proj_covar=tensor([0.0306, 0.0320, 0.0344, 0.0276, 0.0254, 0.0346, 0.0291, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:17:50,661 INFO [train.py:903] (3/4) Epoch 29, batch 6750, loss[loss=0.1847, simple_loss=0.2709, pruned_loss=0.04925, over 19718.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2834, pruned_loss=0.05986, over 3818586.43 frames. ], batch size: 51, lr: 2.80e-03, grad_scale: 4.0 +2023-04-03 15:17:51,944 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.2733, 5.2961, 6.1165, 6.1404, 2.3214, 5.7320, 4.8569, 5.8159], + device='cuda:3'), covar=tensor([0.1722, 0.0777, 0.0649, 0.0599, 0.6124, 0.0839, 0.0646, 0.1209], + device='cuda:3'), in_proj_covar=tensor([0.0822, 0.0795, 0.1004, 0.0879, 0.0871, 0.0771, 0.0592, 0.0935], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 15:18:36,377 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3259, 2.0368, 1.6292, 1.4601, 1.8464, 1.4086, 1.3152, 1.8640], + device='cuda:3'), covar=tensor([0.1037, 0.0914, 0.1130, 0.0957, 0.0634, 0.1344, 0.0732, 0.0479], + device='cuda:3'), in_proj_covar=tensor([0.0306, 0.0321, 0.0345, 0.0276, 0.0254, 0.0347, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:18:45,938 INFO [train.py:903] (3/4) Epoch 29, batch 6800, loss[loss=0.1778, simple_loss=0.2587, pruned_loss=0.04849, over 19835.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2832, pruned_loss=0.0597, over 3826205.09 frames. ], batch size: 52, lr: 2.80e-03, grad_scale: 8.0 +2023-04-03 15:18:48,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8693, 4.4954, 2.8640, 3.9530, 1.2748, 4.3852, 4.2772, 4.3345], + device='cuda:3'), covar=tensor([0.0544, 0.0959, 0.1795, 0.0825, 0.3607, 0.0601, 0.0909, 0.0945], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0432, 0.0520, 0.0357, 0.0410, 0.0460, 0.0453, 0.0487], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:18:49,160 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.543e+02 4.853e+02 6.089e+02 7.661e+02 1.249e+03, threshold=1.218e+03, percent-clipped=1.0 +2023-04-03 15:18:55,186 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1135, 1.8850, 1.7879, 1.9816, 1.7981, 1.7518, 1.6277, 1.9435], + device='cuda:3'), covar=tensor([0.1068, 0.1506, 0.1516, 0.1183, 0.1502, 0.0613, 0.1574, 0.0800], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0360, 0.0320, 0.0259, 0.0310, 0.0259, 0.0322, 0.0264], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 15:18:56,346 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=197993.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:19:33,678 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-04-03 15:19:34,755 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-04-03 15:19:37,012 INFO [train.py:903] (3/4) Epoch 30, batch 0, loss[loss=0.178, simple_loss=0.2585, pruned_loss=0.04876, over 19577.00 frames. ], tot_loss[loss=0.178, simple_loss=0.2585, pruned_loss=0.04876, over 19577.00 frames. ], batch size: 52, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:19:37,012 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 15:19:49,305 INFO [train.py:937] (3/4) Epoch 30, validation: loss=0.167, simple_loss=0.2667, pruned_loss=0.03362, over 944034.00 frames. +2023-04-03 15:19:49,306 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 15:20:02,499 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=198022.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:20:03,294 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-04-03 15:20:35,582 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-04-03 15:20:51,414 INFO [train.py:903] (3/4) Epoch 30, batch 50, loss[loss=0.1998, simple_loss=0.2876, pruned_loss=0.05605, over 19538.00 frames. ], tot_loss[loss=0.2035, simple_loss=0.2862, pruned_loss=0.06047, over 866455.61 frames. ], batch size: 54, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:21:13,154 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-04-03 15:21:20,382 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.592e+02 4.959e+02 5.838e+02 7.711e+02 1.808e+03, threshold=1.168e+03, percent-clipped=5.0 +2023-04-03 15:21:26,258 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1_rvb from training. Duration: 27.0318125 +2023-04-03 15:21:32,472 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 15:21:36,598 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198099.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:21:52,087 INFO [train.py:903] (3/4) Epoch 30, batch 100, loss[loss=0.2192, simple_loss=0.3057, pruned_loss=0.06635, over 19355.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2843, pruned_loss=0.05986, over 1522514.60 frames. ], batch size: 70, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:22:04,478 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-04-03 15:22:04,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0660, 5.1398, 5.9317, 5.9549, 2.1655, 5.6045, 4.6953, 5.5923], + device='cuda:3'), covar=tensor([0.1672, 0.0784, 0.0526, 0.0585, 0.6017, 0.0806, 0.0645, 0.1074], + device='cuda:3'), in_proj_covar=tensor([0.0823, 0.0797, 0.1003, 0.0881, 0.0872, 0.0769, 0.0592, 0.0934], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 15:22:31,191 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5180, 1.5430, 1.8246, 1.7841, 2.5922, 2.2101, 2.7081, 1.2726], + device='cuda:3'), covar=tensor([0.2740, 0.4733, 0.2982, 0.2128, 0.1621, 0.2448, 0.1551, 0.4965], + device='cuda:3'), in_proj_covar=tensor([0.0561, 0.0682, 0.0770, 0.0517, 0.0642, 0.0553, 0.0675, 0.0579], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 15:22:37,681 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=198148.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 15:22:54,432 INFO [train.py:903] (3/4) Epoch 30, batch 150, loss[loss=0.2082, simple_loss=0.2908, pruned_loss=0.06275, over 19764.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2821, pruned_loss=0.05847, over 2038248.81 frames. ], batch size: 54, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:23:07,411 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=198173.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:23:10,046 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.76 vs. limit=2.0 +2023-04-03 15:23:25,378 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.519e+02 4.863e+02 5.858e+02 7.546e+02 1.579e+03, threshold=1.172e+03, percent-clipped=2.0 +2023-04-03 15:23:52,150 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-04-03 15:23:55,535 INFO [train.py:903] (3/4) Epoch 30, batch 200, loss[loss=0.1789, simple_loss=0.2523, pruned_loss=0.05276, over 19746.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2804, pruned_loss=0.05816, over 2427740.41 frames. ], batch size: 46, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:24:42,342 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=198249.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:24:47,682 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=198254.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:24:57,533 INFO [train.py:903] (3/4) Epoch 30, batch 250, loss[loss=0.1969, simple_loss=0.2777, pruned_loss=0.05808, over 19599.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2812, pruned_loss=0.05839, over 2736618.97 frames. ], batch size: 52, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:25:13,976 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=198274.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:25:19,537 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=198278.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:25:29,203 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.056e+02 4.793e+02 5.720e+02 7.294e+02 1.524e+03, threshold=1.144e+03, percent-clipped=6.0 +2023-04-03 15:25:49,114 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=198303.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:26:01,274 INFO [train.py:903] (3/4) Epoch 30, batch 300, loss[loss=0.2108, simple_loss=0.2945, pruned_loss=0.0636, over 17485.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2807, pruned_loss=0.05823, over 2980965.20 frames. ], batch size: 101, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:27:03,945 INFO [train.py:903] (3/4) Epoch 30, batch 350, loss[loss=0.1905, simple_loss=0.2801, pruned_loss=0.0504, over 19525.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2806, pruned_loss=0.05843, over 3177127.65 frames. ], batch size: 54, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:27:06,267 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 15:27:12,181 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=198369.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:27:33,419 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.869e+02 4.876e+02 6.987e+02 8.997e+02 2.429e+03, threshold=1.397e+03, percent-clipped=9.0 +2023-04-03 15:28:04,751 INFO [train.py:903] (3/4) Epoch 30, batch 400, loss[loss=0.2077, simple_loss=0.2912, pruned_loss=0.06215, over 19497.00 frames. ], tot_loss[loss=0.198, simple_loss=0.28, pruned_loss=0.05807, over 3330793.63 frames. ], batch size: 64, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:28:44,692 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=198443.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:29:06,534 INFO [train.py:903] (3/4) Epoch 30, batch 450, loss[loss=0.177, simple_loss=0.2593, pruned_loss=0.0474, over 19615.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.279, pruned_loss=0.05712, over 3454687.90 frames. ], batch size: 50, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:29:08,589 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-04-03 15:29:38,786 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.061e+02 5.102e+02 6.085e+02 7.779e+02 1.785e+03, threshold=1.217e+03, percent-clipped=4.0 +2023-04-03 15:29:40,844 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-04-03 15:29:42,029 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-04-03 15:29:54,086 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0476, 2.1449, 2.3810, 2.5862, 2.0188, 2.5020, 2.3297, 2.2199], + device='cuda:3'), covar=tensor([0.4295, 0.4105, 0.1923, 0.2678, 0.4370, 0.2367, 0.5131, 0.3445], + device='cuda:3'), in_proj_covar=tensor([0.0946, 0.1028, 0.0748, 0.0959, 0.0925, 0.0863, 0.0864, 0.0815], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 15:30:08,494 INFO [train.py:903] (3/4) Epoch 30, batch 500, loss[loss=0.1895, simple_loss=0.2661, pruned_loss=0.05647, over 19470.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2792, pruned_loss=0.05742, over 3540584.69 frames. ], batch size: 49, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:30:11,060 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7880, 1.4446, 1.8289, 1.6225, 4.3434, 1.1857, 2.4858, 4.7141], + device='cuda:3'), covar=tensor([0.0482, 0.3061, 0.2809, 0.2069, 0.0737, 0.2838, 0.1692, 0.0183], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0383, 0.0402, 0.0356, 0.0387, 0.0360, 0.0401, 0.0424], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:30:16,803 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198517.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:31:06,315 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=198558.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:31:11,516 INFO [train.py:903] (3/4) Epoch 30, batch 550, loss[loss=0.1862, simple_loss=0.2667, pruned_loss=0.05283, over 19565.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2802, pruned_loss=0.05747, over 3606206.49 frames. ], batch size: 52, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:31:40,933 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.425e+02 4.928e+02 6.417e+02 8.667e+02 2.852e+03, threshold=1.283e+03, percent-clipped=10.0 +2023-04-03 15:31:47,763 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5040, 1.4362, 1.4535, 1.8704, 1.5008, 1.6238, 1.6813, 1.5295], + device='cuda:3'), covar=tensor([0.0832, 0.0893, 0.0993, 0.0616, 0.0903, 0.0824, 0.0860, 0.0726], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0227, 0.0240, 0.0226, 0.0215, 0.0188, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 15:31:56,099 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-04-03 15:32:13,287 INFO [train.py:903] (3/4) Epoch 30, batch 600, loss[loss=0.2021, simple_loss=0.2938, pruned_loss=0.05517, over 19767.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2795, pruned_loss=0.05689, over 3663227.25 frames. ], batch size: 54, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:32:27,174 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7039, 1.5137, 1.7067, 2.2571, 1.8394, 1.7166, 1.7595, 1.7204], + device='cuda:3'), covar=tensor([0.0985, 0.1299, 0.1032, 0.0668, 0.1109, 0.1098, 0.1208, 0.0889], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0228, 0.0240, 0.0226, 0.0215, 0.0188, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 15:32:28,389 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=198625.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:32:38,410 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198633.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:32:52,855 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-04-03 15:33:01,346 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=198650.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:33:14,979 INFO [train.py:903] (3/4) Epoch 30, batch 650, loss[loss=0.1899, simple_loss=0.2699, pruned_loss=0.055, over 19832.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2792, pruned_loss=0.05694, over 3692777.48 frames. ], batch size: 52, lr: 2.75e-03, grad_scale: 8.0 +2023-04-03 15:33:46,600 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.134e+02 5.000e+02 5.780e+02 7.067e+02 2.104e+03, threshold=1.156e+03, percent-clipped=2.0 +2023-04-03 15:34:00,601 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0507, 3.3644, 1.8724, 2.0914, 3.0847, 1.7442, 1.3941, 2.2349], + device='cuda:3'), covar=tensor([0.1495, 0.0715, 0.1188, 0.0944, 0.0596, 0.1366, 0.1127, 0.0766], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0321, 0.0346, 0.0276, 0.0253, 0.0349, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:34:02,337 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-04-03 15:34:16,541 INFO [train.py:903] (3/4) Epoch 30, batch 700, loss[loss=0.201, simple_loss=0.279, pruned_loss=0.06149, over 19843.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2803, pruned_loss=0.0576, over 3724004.61 frames. ], batch size: 52, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:35:18,678 INFO [train.py:903] (3/4) Epoch 30, batch 750, loss[loss=0.2047, simple_loss=0.2763, pruned_loss=0.06654, over 19408.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2812, pruned_loss=0.05786, over 3756637.47 frames. ], batch size: 47, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:35:51,149 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.284e+02 4.908e+02 5.891e+02 7.920e+02 1.978e+03, threshold=1.178e+03, percent-clipped=7.0 +2023-04-03 15:36:23,528 INFO [train.py:903] (3/4) Epoch 30, batch 800, loss[loss=0.1803, simple_loss=0.2462, pruned_loss=0.05719, over 19733.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2798, pruned_loss=0.05741, over 3765342.20 frames. ], batch size: 45, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:36:27,263 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=198814.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:36:39,855 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-04-03 15:36:56,808 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=198839.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:37:25,167 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=198861.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:37:26,181 INFO [train.py:903] (3/4) Epoch 30, batch 850, loss[loss=0.2084, simple_loss=0.2946, pruned_loss=0.06108, over 19541.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2799, pruned_loss=0.05724, over 3778929.70 frames. ], batch size: 64, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:37:35,509 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198870.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:37:55,570 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.084e+02 4.755e+02 5.781e+02 7.125e+02 1.514e+03, threshold=1.156e+03, percent-clipped=6.0 +2023-04-03 15:38:13,675 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198901.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:38:20,315 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9_rvb from training. Duration: 25.061125 +2023-04-03 15:38:26,177 INFO [train.py:903] (3/4) Epoch 30, batch 900, loss[loss=0.2097, simple_loss=0.2908, pruned_loss=0.0643, over 19483.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2797, pruned_loss=0.05774, over 3789158.80 frames. ], batch size: 64, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:39:07,725 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=198945.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:39:27,821 INFO [train.py:903] (3/4) Epoch 30, batch 950, loss[loss=0.1841, simple_loss=0.2665, pruned_loss=0.05088, over 19686.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2799, pruned_loss=0.05792, over 3787747.05 frames. ], batch size: 53, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:39:32,350 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9_rvb from training. Duration: 26.32775 +2023-04-03 15:39:47,116 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=198976.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:39:48,127 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=198977.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:39:59,673 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.142e+02 5.052e+02 6.305e+02 7.870e+02 1.588e+03, threshold=1.261e+03, percent-clipped=4.0 +2023-04-03 15:40:30,262 INFO [train.py:903] (3/4) Epoch 30, batch 1000, loss[loss=0.2292, simple_loss=0.303, pruned_loss=0.07774, over 19414.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2806, pruned_loss=0.05824, over 3798867.19 frames. ], batch size: 70, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:40:34,412 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-03 15:41:23,981 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-04-03 15:41:33,182 INFO [train.py:903] (3/4) Epoch 30, batch 1050, loss[loss=0.2083, simple_loss=0.2876, pruned_loss=0.06446, over 19662.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2811, pruned_loss=0.05861, over 3783150.32 frames. ], batch size: 53, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:42:03,436 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.270e+02 4.631e+02 5.961e+02 7.864e+02 1.953e+03, threshold=1.192e+03, percent-clipped=2.0 +2023-04-03 15:42:04,604 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-04-03 15:42:11,038 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199092.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:42:35,181 INFO [train.py:903] (3/4) Epoch 30, batch 1100, loss[loss=0.1646, simple_loss=0.2469, pruned_loss=0.0412, over 19394.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2804, pruned_loss=0.05813, over 3790774.23 frames. ], batch size: 47, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:43:38,423 INFO [train.py:903] (3/4) Epoch 30, batch 1150, loss[loss=0.1882, simple_loss=0.2764, pruned_loss=0.05002, over 18850.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2806, pruned_loss=0.05859, over 3786094.76 frames. ], batch size: 74, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:44:10,291 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.218e+02 5.110e+02 5.915e+02 7.639e+02 1.372e+03, threshold=1.183e+03, percent-clipped=6.0 +2023-04-03 15:44:18,777 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199194.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:44:28,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.8954, 1.3428, 1.1007, 1.0059, 1.1737, 1.0579, 0.9537, 1.2658], + device='cuda:3'), covar=tensor([0.0698, 0.0921, 0.1223, 0.0827, 0.0652, 0.1361, 0.0666, 0.0537], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0319, 0.0343, 0.0274, 0.0252, 0.0346, 0.0290, 0.0274], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:44:41,424 INFO [train.py:903] (3/4) Epoch 30, batch 1200, loss[loss=0.1894, simple_loss=0.2601, pruned_loss=0.05933, over 19466.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2805, pruned_loss=0.05868, over 3799011.20 frames. ], batch size: 49, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:44:43,962 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199214.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:45:06,899 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199232.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:45:14,390 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-04-03 15:45:22,172 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199245.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:45:38,756 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199257.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:45:44,966 INFO [train.py:903] (3/4) Epoch 30, batch 1250, loss[loss=0.1988, simple_loss=0.2817, pruned_loss=0.05796, over 18761.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2811, pruned_loss=0.05901, over 3814161.88 frames. ], batch size: 74, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:45:54,324 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.5570, 4.1312, 2.8198, 3.6080, 1.0132, 4.1498, 3.9782, 4.0777], + device='cuda:3'), covar=tensor([0.0696, 0.0995, 0.1915, 0.0925, 0.4057, 0.0722, 0.0985, 0.1385], + device='cuda:3'), in_proj_covar=tensor([0.0534, 0.0432, 0.0520, 0.0359, 0.0408, 0.0459, 0.0453, 0.0488], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:46:14,641 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.608e+02 5.004e+02 5.991e+02 7.632e+02 1.398e+03, threshold=1.198e+03, percent-clipped=2.0 +2023-04-03 15:46:17,149 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199289.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:46:46,511 INFO [train.py:903] (3/4) Epoch 30, batch 1300, loss[loss=0.2229, simple_loss=0.2828, pruned_loss=0.08154, over 19053.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2821, pruned_loss=0.05924, over 3823485.37 frames. ], batch size: 42, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:47:07,594 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199329.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:47:32,056 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199348.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:47:46,101 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199360.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:47:47,888 INFO [train.py:903] (3/4) Epoch 30, batch 1350, loss[loss=0.2193, simple_loss=0.2986, pruned_loss=0.07, over 19767.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2815, pruned_loss=0.05878, over 3817540.98 frames. ], batch size: 54, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:48:03,218 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199373.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:48:20,889 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.443e+02 4.672e+02 5.739e+02 7.239e+02 1.592e+03, threshold=1.148e+03, percent-clipped=7.0 +2023-04-03 15:48:40,938 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199404.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:48:50,927 INFO [train.py:903] (3/4) Epoch 30, batch 1400, loss[loss=0.1848, simple_loss=0.2776, pruned_loss=0.04603, over 18249.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2818, pruned_loss=0.05894, over 3806955.76 frames. ], batch size: 83, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:48:58,990 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199418.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:49:03,142 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-04-03 15:49:53,480 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.28 vs. limit=5.0 +2023-04-03 15:49:55,030 INFO [train.py:903] (3/4) Epoch 30, batch 1450, loss[loss=0.223, simple_loss=0.3068, pruned_loss=0.06956, over 19536.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2816, pruned_loss=0.059, over 3813936.96 frames. ], batch size: 56, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:49:56,189 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_rvb from training. Duration: 25.85 +2023-04-03 15:50:11,596 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199476.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:50:25,539 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.487e+02 5.040e+02 6.121e+02 7.927e+02 1.972e+03, threshold=1.224e+03, percent-clipped=6.0 +2023-04-03 15:50:56,161 INFO [train.py:903] (3/4) Epoch 30, batch 1500, loss[loss=0.1944, simple_loss=0.2843, pruned_loss=0.0523, over 19656.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2817, pruned_loss=0.05875, over 3827097.79 frames. ], batch size: 55, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:50:56,472 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199512.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:51:28,282 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199538.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:51:46,212 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.5957, 1.3152, 1.4014, 1.4169, 2.3713, 1.1717, 1.9634, 2.5946], + device='cuda:3'), covar=tensor([0.0563, 0.2370, 0.2436, 0.1552, 0.0608, 0.2026, 0.1695, 0.0432], + device='cuda:3'), in_proj_covar=tensor([0.0427, 0.0383, 0.0403, 0.0356, 0.0388, 0.0362, 0.0403, 0.0424], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 15:51:47,420 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199553.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 15:51:58,247 INFO [train.py:903] (3/4) Epoch 30, batch 1550, loss[loss=0.1749, simple_loss=0.2647, pruned_loss=0.04253, over 19664.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2822, pruned_loss=0.05896, over 3819041.20 frames. ], batch size: 58, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:52:28,194 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199585.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:52:31,297 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.284e+02 4.717e+02 5.804e+02 6.903e+02 1.639e+03, threshold=1.161e+03, percent-clipped=1.0 +2023-04-03 15:52:46,522 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1374, 2.0258, 1.9345, 1.7831, 1.6070, 1.7089, 0.5823, 1.1081], + device='cuda:3'), covar=tensor([0.0710, 0.0730, 0.0548, 0.0879, 0.1356, 0.1016, 0.1561, 0.1182], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0364, 0.0370, 0.0395, 0.0472, 0.0398, 0.0347, 0.0348], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 15:52:59,340 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199610.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:53:01,355 INFO [train.py:903] (3/4) Epoch 30, batch 1600, loss[loss=0.2048, simple_loss=0.2941, pruned_loss=0.05773, over 17468.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2824, pruned_loss=0.05872, over 3815560.59 frames. ], batch size: 101, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:53:06,429 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199616.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:53:26,387 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9_rvb from training. Duration: 30.1555625 +2023-04-03 15:53:37,708 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199641.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:53:53,168 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199653.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:54:01,237 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199660.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:54:03,852 INFO [train.py:903] (3/4) Epoch 30, batch 1650, loss[loss=0.2017, simple_loss=0.2873, pruned_loss=0.058, over 18279.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2822, pruned_loss=0.05838, over 3817867.90 frames. ], batch size: 83, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:54:32,734 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199685.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:54:35,820 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.122e+02 4.724e+02 5.858e+02 7.730e+02 3.208e+03, threshold=1.172e+03, percent-clipped=5.0 +2023-04-03 15:55:06,071 INFO [train.py:903] (3/4) Epoch 30, batch 1700, loss[loss=0.2239, simple_loss=0.3056, pruned_loss=0.07114, over 19082.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2814, pruned_loss=0.05816, over 3818556.06 frames. ], batch size: 69, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 15:55:24,049 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0862, 1.9650, 1.9004, 1.7788, 1.5526, 1.6705, 0.6204, 1.0927], + device='cuda:3'), covar=tensor([0.0794, 0.0775, 0.0563, 0.0904, 0.1391, 0.1082, 0.1503, 0.1276], + device='cuda:3'), in_proj_covar=tensor([0.0371, 0.0366, 0.0373, 0.0398, 0.0476, 0.0402, 0.0350, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 15:55:40,547 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.70 vs. limit=2.0 +2023-04-03 15:55:46,467 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-04-03 15:56:09,059 INFO [train.py:903] (3/4) Epoch 30, batch 1750, loss[loss=0.2107, simple_loss=0.2943, pruned_loss=0.06355, over 19782.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2808, pruned_loss=0.05814, over 3818654.36 frames. ], batch size: 56, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:56:09,245 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199762.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:56:42,931 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.076e+02 4.893e+02 6.270e+02 7.375e+02 1.627e+03, threshold=1.254e+03, percent-clipped=1.0 +2023-04-03 15:57:00,469 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8336, 4.2877, 4.5034, 4.4979, 1.7741, 4.2402, 3.6653, 4.2601], + device='cuda:3'), covar=tensor([0.1715, 0.0827, 0.0643, 0.0757, 0.6491, 0.1142, 0.0741, 0.1135], + device='cuda:3'), in_proj_covar=tensor([0.0829, 0.0803, 0.1015, 0.0891, 0.0877, 0.0776, 0.0600, 0.0943], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 15:57:11,415 INFO [train.py:903] (3/4) Epoch 30, batch 1800, loss[loss=0.2282, simple_loss=0.3029, pruned_loss=0.07674, over 12718.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2802, pruned_loss=0.05786, over 3817799.66 frames. ], batch size: 135, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:57:21,125 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199820.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:58:06,473 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199856.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 15:58:08,425 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9_rvb from training. Duration: 25.0944375 +2023-04-03 15:58:13,035 INFO [train.py:903] (3/4) Epoch 30, batch 1850, loss[loss=0.1733, simple_loss=0.2565, pruned_loss=0.04506, over 19610.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2803, pruned_loss=0.05781, over 3818288.18 frames. ], batch size: 50, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:58:15,402 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.7906, 4.2958, 4.5071, 4.4953, 1.6438, 4.2664, 3.6641, 4.2424], + device='cuda:3'), covar=tensor([0.1719, 0.0875, 0.0640, 0.0732, 0.6522, 0.1060, 0.0739, 0.1097], + device='cuda:3'), in_proj_covar=tensor([0.0831, 0.0806, 0.1020, 0.0894, 0.0881, 0.0779, 0.0603, 0.0948], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 15:58:32,560 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199877.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:58:46,821 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.213e+02 5.043e+02 6.338e+02 8.325e+02 2.069e+03, threshold=1.268e+03, percent-clipped=7.0 +2023-04-03 15:58:46,860 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9_rvb from training. Duration: 27.8166875 +2023-04-03 15:58:57,092 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=199897.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 15:59:12,888 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=199909.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:59:15,996 INFO [train.py:903] (3/4) Epoch 30, batch 1900, loss[loss=0.1684, simple_loss=0.2521, pruned_loss=0.04235, over 19668.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2795, pruned_loss=0.05724, over 3834936.24 frames. ], batch size: 53, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 15:59:33,273 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-04-03 15:59:33,566 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=199927.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:59:38,036 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9_rvb from training. Duration: 27.02225 +2023-04-03 15:59:42,795 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=199934.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 15:59:43,930 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199935.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:00:03,955 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-04-03 16:00:17,076 INFO [train.py:903] (3/4) Epoch 30, batch 1950, loss[loss=0.2008, simple_loss=0.2822, pruned_loss=0.05969, over 19575.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.28, pruned_loss=0.05775, over 3830821.48 frames. ], batch size: 61, lr: 2.74e-03, grad_scale: 4.0 +2023-04-03 16:00:29,591 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=199971.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 16:00:51,090 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.452e+02 5.211e+02 6.525e+02 7.685e+02 1.771e+03, threshold=1.305e+03, percent-clipped=2.0 +2023-04-03 16:01:21,187 INFO [train.py:903] (3/4) Epoch 30, batch 2000, loss[loss=0.1972, simple_loss=0.2789, pruned_loss=0.05777, over 19657.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2807, pruned_loss=0.05781, over 3825825.00 frames. ], batch size: 55, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 16:01:22,698 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=200012.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 16:02:21,500 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_rvb from training. Duration: 26.8349375 +2023-04-03 16:02:23,758 INFO [train.py:903] (3/4) Epoch 30, batch 2050, loss[loss=0.1385, simple_loss=0.2197, pruned_loss=0.02862, over 19744.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2809, pruned_loss=0.05806, over 3834844.03 frames. ], batch size: 46, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 16:02:43,237 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-04-03 16:02:43,268 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-04-03 16:02:50,489 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9864, 2.0592, 2.3619, 2.6435, 2.0572, 2.5191, 2.3187, 2.1553], + device='cuda:3'), covar=tensor([0.4253, 0.4021, 0.1992, 0.2471, 0.4109, 0.2250, 0.5044, 0.3356], + device='cuda:3'), in_proj_covar=tensor([0.0947, 0.1032, 0.0750, 0.0959, 0.0926, 0.0866, 0.0867, 0.0815], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:02:57,797 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.348e+02 4.986e+02 6.252e+02 8.206e+02 1.738e+03, threshold=1.250e+03, percent-clipped=4.0 +2023-04-03 16:03:03,561 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-04-03 16:03:11,191 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.36 vs. limit=5.0 +2023-04-03 16:03:26,457 INFO [train.py:903] (3/4) Epoch 30, batch 2100, loss[loss=0.1855, simple_loss=0.2799, pruned_loss=0.04554, over 19621.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2803, pruned_loss=0.05771, over 3836628.89 frames. ], batch size: 57, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 16:03:52,969 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200133.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:03:57,151 WARNING [train.py:1073] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9_rvb from training. Duration: 25.3333125 +2023-04-03 16:04:18,930 WARNING [train.py:1073] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9_rvb from training. Duration: 26.6166875 +2023-04-03 16:04:24,940 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200158.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:04:29,059 INFO [train.py:903] (3/4) Epoch 30, batch 2150, loss[loss=0.2118, simple_loss=0.3, pruned_loss=0.06181, over 19745.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2809, pruned_loss=0.05824, over 3839442.53 frames. ], batch size: 63, lr: 2.74e-03, grad_scale: 8.0 +2023-04-03 16:05:02,778 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.654e+02 4.819e+02 5.957e+02 8.119e+02 2.108e+03, threshold=1.191e+03, percent-clipped=2.0 +2023-04-03 16:05:05,543 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200191.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:05:10,687 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=200195.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:05:30,903 INFO [train.py:903] (3/4) Epoch 30, batch 2200, loss[loss=0.1953, simple_loss=0.2714, pruned_loss=0.05959, over 19605.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2811, pruned_loss=0.05869, over 3844355.36 frames. ], batch size: 50, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:05:37,204 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200216.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:05:39,453 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0051, 4.4243, 4.7137, 4.7042, 1.6716, 4.4079, 3.8167, 4.4455], + device='cuda:3'), covar=tensor([0.1659, 0.0832, 0.0628, 0.0678, 0.6518, 0.0911, 0.0700, 0.1154], + device='cuda:3'), in_proj_covar=tensor([0.0824, 0.0798, 0.1011, 0.0885, 0.0872, 0.0772, 0.0597, 0.0936], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 16:05:50,128 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200227.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 16:06:20,844 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200252.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:06:33,032 INFO [train.py:903] (3/4) Epoch 30, batch 2250, loss[loss=0.1868, simple_loss=0.2742, pruned_loss=0.04969, over 19678.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2817, pruned_loss=0.0594, over 3849574.08 frames. ], batch size: 60, lr: 2.73e-03, grad_scale: 4.0 +2023-04-03 16:06:41,188 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200268.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 16:06:44,368 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=200271.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:07:08,303 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.532e+02 5.298e+02 6.530e+02 8.599e+02 1.543e+03, threshold=1.306e+03, percent-clipped=6.0 +2023-04-03 16:07:12,104 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200293.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:07:17,571 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=200297.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:07:35,295 INFO [train.py:903] (3/4) Epoch 30, batch 2300, loss[loss=0.2221, simple_loss=0.3039, pruned_loss=0.07012, over 17409.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2821, pruned_loss=0.0598, over 3828199.61 frames. ], batch size: 101, lr: 2.73e-03, grad_scale: 4.0 +2023-04-03 16:07:51,106 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_rvb from training. Duration: 26.205 +2023-04-03 16:08:18,925 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-04-03 16:08:37,754 INFO [train.py:903] (3/4) Epoch 30, batch 2350, loss[loss=0.1645, simple_loss=0.2474, pruned_loss=0.04083, over 19776.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2811, pruned_loss=0.05903, over 3833090.47 frames. ], batch size: 47, lr: 2.73e-03, grad_scale: 4.0 +2023-04-03 16:08:47,325 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.65 vs. limit=5.0 +2023-04-03 16:09:07,678 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=200386.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:09:12,874 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.379e+02 4.836e+02 5.833e+02 7.163e+02 1.475e+03, threshold=1.167e+03, percent-clipped=1.0 +2023-04-03 16:09:21,096 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_rvb from training. Duration: 25.775 +2023-04-03 16:09:34,047 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.8488, 1.7008, 1.8679, 1.4587, 4.3738, 1.1614, 2.6100, 4.7241], + device='cuda:3'), covar=tensor([0.0486, 0.2839, 0.2972, 0.2280, 0.0771, 0.2882, 0.1641, 0.0188], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0383, 0.0404, 0.0356, 0.0386, 0.0361, 0.0402, 0.0423], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:09:38,137 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9_rvb from training. Duration: 25.45 +2023-04-03 16:09:40,555 INFO [train.py:903] (3/4) Epoch 30, batch 2400, loss[loss=0.194, simple_loss=0.2745, pruned_loss=0.05678, over 19583.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2812, pruned_loss=0.05922, over 3835038.08 frames. ], batch size: 52, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:10:24,211 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.6004, 1.1337, 1.3545, 1.1723, 2.2262, 1.0529, 2.0342, 2.5229], + device='cuda:3'), covar=tensor([0.0754, 0.3044, 0.3216, 0.1919, 0.0908, 0.2228, 0.1269, 0.0488], + device='cuda:3'), in_proj_covar=tensor([0.0427, 0.0383, 0.0403, 0.0356, 0.0386, 0.0361, 0.0402, 0.0423], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:10:43,361 INFO [train.py:903] (3/4) Epoch 30, batch 2450, loss[loss=0.2125, simple_loss=0.2959, pruned_loss=0.0646, over 18237.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2816, pruned_loss=0.05905, over 3832570.88 frames. ], batch size: 84, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:11:19,107 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.119e+02 5.121e+02 5.946e+02 7.814e+02 2.121e+03, threshold=1.189e+03, percent-clipped=7.0 +2023-04-03 16:11:27,561 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8400, 3.3181, 3.3593, 3.3644, 1.4519, 3.2409, 2.8373, 3.1588], + device='cuda:3'), covar=tensor([0.1804, 0.0956, 0.0860, 0.1023, 0.5599, 0.1130, 0.0896, 0.1307], + device='cuda:3'), in_proj_covar=tensor([0.0819, 0.0795, 0.1006, 0.0883, 0.0868, 0.0770, 0.0593, 0.0935], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 16:11:46,465 INFO [train.py:903] (3/4) Epoch 30, batch 2500, loss[loss=0.2232, simple_loss=0.3063, pruned_loss=0.07006, over 19332.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2828, pruned_loss=0.05971, over 3822741.93 frames. ], batch size: 66, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:12:20,493 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=200539.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:12:48,793 INFO [train.py:903] (3/4) Epoch 30, batch 2550, loss[loss=0.1902, simple_loss=0.2713, pruned_loss=0.05453, over 19744.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2832, pruned_loss=0.06008, over 3818102.77 frames. ], batch size: 54, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:13:23,044 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.619e+02 4.923e+02 6.163e+02 7.973e+02 2.573e+03, threshold=1.233e+03, percent-clipped=12.0 +2023-04-03 16:13:41,769 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.7253, 4.3571, 2.9328, 3.8259, 1.2922, 4.2744, 4.1573, 4.2375], + device='cuda:3'), covar=tensor([0.0606, 0.0870, 0.1729, 0.0793, 0.3407, 0.0663, 0.0958, 0.1190], + device='cuda:3'), in_proj_covar=tensor([0.0533, 0.0431, 0.0517, 0.0358, 0.0409, 0.0458, 0.0449, 0.0487], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:13:47,178 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-04-03 16:13:51,625 INFO [train.py:903] (3/4) Epoch 30, batch 2600, loss[loss=0.2051, simple_loss=0.2746, pruned_loss=0.0678, over 19735.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2828, pruned_loss=0.05965, over 3827021.47 frames. ], batch size: 51, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:14:28,371 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=200641.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:14:29,709 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200642.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:14:44,515 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=200654.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:14:54,163 INFO [train.py:903] (3/4) Epoch 30, batch 2650, loss[loss=0.1884, simple_loss=0.279, pruned_loss=0.04888, over 19595.00 frames. ], tot_loss[loss=0.2023, simple_loss=0.284, pruned_loss=0.06028, over 3816102.40 frames. ], batch size: 52, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:14:55,566 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=200663.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:15:00,429 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200667.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:15:12,708 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.8076, 1.4675, 1.7024, 1.7282, 3.3660, 1.3800, 2.4610, 3.8551], + device='cuda:3'), covar=tensor([0.0498, 0.2893, 0.2831, 0.1782, 0.0680, 0.2374, 0.1379, 0.0222], + device='cuda:3'), in_proj_covar=tensor([0.0425, 0.0382, 0.0402, 0.0355, 0.0386, 0.0361, 0.0402, 0.0422], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:15:15,800 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9_rvb from training. Duration: 27.25 +2023-04-03 16:15:28,264 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.276e+02 4.881e+02 6.412e+02 8.116e+02 1.737e+03, threshold=1.282e+03, percent-clipped=7.0 +2023-04-03 16:15:55,154 INFO [train.py:903] (3/4) Epoch 30, batch 2700, loss[loss=0.1864, simple_loss=0.2669, pruned_loss=0.05293, over 19658.00 frames. ], tot_loss[loss=0.202, simple_loss=0.284, pruned_loss=0.06004, over 3816094.63 frames. ], batch size: 53, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:16:34,929 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-03 16:16:51,756 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=200756.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:16:59,612 INFO [train.py:903] (3/4) Epoch 30, batch 2750, loss[loss=0.2177, simple_loss=0.3008, pruned_loss=0.06726, over 18961.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2829, pruned_loss=0.0593, over 3821878.70 frames. ], batch size: 74, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:17:34,086 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.266e+02 4.594e+02 5.767e+02 6.844e+02 1.269e+03, threshold=1.153e+03, percent-clipped=0.0 +2023-04-03 16:18:02,567 INFO [train.py:903] (3/4) Epoch 30, batch 2800, loss[loss=0.201, simple_loss=0.2882, pruned_loss=0.05692, over 17970.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.282, pruned_loss=0.05893, over 3815892.31 frames. ], batch size: 83, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:19:05,168 INFO [train.py:903] (3/4) Epoch 30, batch 2850, loss[loss=0.1987, simple_loss=0.2818, pruned_loss=0.05777, over 19349.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2819, pruned_loss=0.05869, over 3819406.87 frames. ], batch size: 66, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:19:39,283 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.089e+02 4.865e+02 5.683e+02 7.765e+02 1.857e+03, threshold=1.137e+03, percent-clipped=6.0 +2023-04-03 16:20:04,686 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=200910.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:20:06,539 INFO [train.py:903] (3/4) Epoch 30, batch 2900, loss[loss=0.2136, simple_loss=0.2964, pruned_loss=0.06539, over 19654.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2813, pruned_loss=0.05825, over 3831378.11 frames. ], batch size: 55, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:20:07,768 WARNING [train.py:1073] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-04-03 16:20:36,303 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=200935.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 16:21:00,057 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3384, 1.9731, 1.5155, 1.3647, 1.8147, 1.2556, 1.2946, 1.7916], + device='cuda:3'), covar=tensor([0.1016, 0.0887, 0.1215, 0.0921, 0.0664, 0.1427, 0.0772, 0.0490], + device='cuda:3'), in_proj_covar=tensor([0.0305, 0.0321, 0.0345, 0.0277, 0.0255, 0.0350, 0.0291, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:21:08,763 INFO [train.py:903] (3/4) Epoch 30, batch 2950, loss[loss=0.1907, simple_loss=0.2697, pruned_loss=0.05582, over 19740.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2813, pruned_loss=0.05851, over 3822043.01 frames. ], batch size: 51, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:21:44,102 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.449e+02 4.844e+02 6.179e+02 7.399e+02 1.416e+03, threshold=1.236e+03, percent-clipped=4.0 +2023-04-03 16:22:06,393 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=201007.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:22:11,902 INFO [train.py:903] (3/4) Epoch 30, batch 3000, loss[loss=0.191, simple_loss=0.2808, pruned_loss=0.05064, over 19673.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2798, pruned_loss=0.05758, over 3837848.37 frames. ], batch size: 55, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:22:11,902 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 16:22:26,180 INFO [train.py:937] (3/4) Epoch 30, validation: loss=0.1666, simple_loss=0.266, pruned_loss=0.03357, over 944034.00 frames. +2023-04-03 16:22:26,181 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 16:22:26,663 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=201012.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:22:32,366 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_rvb from training. Duration: 29.735 +2023-04-03 16:22:57,372 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=201037.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:23:27,828 INFO [train.py:903] (3/4) Epoch 30, batch 3050, loss[loss=0.2175, simple_loss=0.3045, pruned_loss=0.06528, over 19686.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2795, pruned_loss=0.05742, over 3843065.15 frames. ], batch size: 59, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:23:57,907 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.0561, 1.2793, 1.6736, 0.9869, 2.3106, 3.0236, 2.7193, 3.2226], + device='cuda:3'), covar=tensor([0.1755, 0.4084, 0.3649, 0.2969, 0.0701, 0.0253, 0.0279, 0.0353], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0335, 0.0367, 0.0274, 0.0257, 0.0200, 0.0222, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 16:24:02,339 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201089.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:24:03,230 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.915e+02 4.896e+02 5.871e+02 7.483e+02 2.064e+03, threshold=1.174e+03, percent-clipped=5.0 +2023-04-03 16:24:32,128 INFO [train.py:903] (3/4) Epoch 30, batch 3100, loss[loss=0.1956, simple_loss=0.2833, pruned_loss=0.05396, over 19727.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2794, pruned_loss=0.05762, over 3841449.82 frames. ], batch size: 63, lr: 2.73e-03, grad_scale: 4.0 +2023-04-03 16:24:43,641 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=201122.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:25:03,520 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-04-03 16:25:33,508 INFO [train.py:903] (3/4) Epoch 30, batch 3150, loss[loss=0.1852, simple_loss=0.2533, pruned_loss=0.05856, over 18560.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2788, pruned_loss=0.05747, over 3837917.95 frames. ], batch size: 41, lr: 2.73e-03, grad_scale: 4.0 +2023-04-03 16:26:02,513 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-04-03 16:26:10,329 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.152e+02 4.625e+02 5.752e+02 7.402e+02 1.953e+03, threshold=1.150e+03, percent-clipped=5.0 +2023-04-03 16:26:16,496 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5447, 1.3610, 1.3023, 1.5267, 1.3144, 1.2998, 1.3005, 1.4645], + device='cuda:3'), covar=tensor([0.0875, 0.1156, 0.1217, 0.0879, 0.1136, 0.0539, 0.1256, 0.0710], + device='cuda:3'), in_proj_covar=tensor([0.0280, 0.0358, 0.0319, 0.0258, 0.0309, 0.0260, 0.0324, 0.0263], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:3') +2023-04-03 16:26:35,926 INFO [train.py:903] (3/4) Epoch 30, batch 3200, loss[loss=0.1869, simple_loss=0.2796, pruned_loss=0.04707, over 17562.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2786, pruned_loss=0.05739, over 3826086.50 frames. ], batch size: 101, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:26:38,599 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1382, 1.6790, 1.8426, 2.9985, 2.2675, 2.2547, 2.4223, 2.0023], + device='cuda:3'), covar=tensor([0.0868, 0.1088, 0.1100, 0.0745, 0.0839, 0.0934, 0.0964, 0.0794], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0227, 0.0240, 0.0227, 0.0216, 0.0187, 0.0209], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 16:27:03,294 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-04-03 16:27:26,678 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.5915, 1.6047, 1.7618, 1.7304, 2.3671, 2.0788, 2.4271, 1.5324], + device='cuda:3'), covar=tensor([0.2028, 0.3562, 0.2345, 0.1663, 0.1322, 0.1897, 0.1204, 0.4128], + device='cuda:3'), in_proj_covar=tensor([0.0557, 0.0679, 0.0766, 0.0514, 0.0637, 0.0549, 0.0672, 0.0580], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:27:40,098 INFO [train.py:903] (3/4) Epoch 30, batch 3250, loss[loss=0.21, simple_loss=0.295, pruned_loss=0.06252, over 18032.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2791, pruned_loss=0.05763, over 3815075.58 frames. ], batch size: 83, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:27:51,503 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201271.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:28:14,881 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.466e+02 4.729e+02 5.856e+02 7.119e+02 1.424e+03, threshold=1.171e+03, percent-clipped=4.0 +2023-04-03 16:28:16,398 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201292.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:28:42,757 INFO [train.py:903] (3/4) Epoch 30, batch 3300, loss[loss=0.1694, simple_loss=0.2552, pruned_loss=0.04173, over 19778.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2787, pruned_loss=0.05742, over 3825928.65 frames. ], batch size: 48, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:28:49,690 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-04-03 16:28:51,286 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6904, 2.3206, 1.6965, 1.5381, 2.1736, 1.3694, 1.4477, 2.1325], + device='cuda:3'), covar=tensor([0.1083, 0.0838, 0.1165, 0.0952, 0.0581, 0.1367, 0.0816, 0.0503], + device='cuda:3'), in_proj_covar=tensor([0.0303, 0.0319, 0.0343, 0.0276, 0.0253, 0.0348, 0.0290, 0.0276], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:29:22,254 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-04-03 16:29:44,707 INFO [train.py:903] (3/4) Epoch 30, batch 3350, loss[loss=0.2, simple_loss=0.2736, pruned_loss=0.0632, over 19736.00 frames. ], tot_loss[loss=0.198, simple_loss=0.28, pruned_loss=0.05802, over 3828259.21 frames. ], batch size: 51, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:30:05,863 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=201378.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:30:21,794 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.184e+02 4.772e+02 5.759e+02 7.012e+02 1.305e+03, threshold=1.152e+03, percent-clipped=2.0 +2023-04-03 16:30:37,323 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=201403.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:30:48,360 INFO [train.py:903] (3/4) Epoch 30, batch 3400, loss[loss=0.2107, simple_loss=0.2855, pruned_loss=0.06795, over 19665.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2807, pruned_loss=0.0582, over 3820069.83 frames. ], batch size: 55, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:31:16,058 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=201433.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:31:51,805 INFO [train.py:903] (3/4) Epoch 30, batch 3450, loss[loss=0.2084, simple_loss=0.2971, pruned_loss=0.05985, over 19582.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2811, pruned_loss=0.05878, over 3815987.75 frames. ], batch size: 61, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:31:57,362 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9_rvb from training. Duration: 25.2444375 +2023-04-03 16:32:27,320 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.371e+02 5.269e+02 6.567e+02 8.372e+02 2.121e+03, threshold=1.313e+03, percent-clipped=9.0 +2023-04-03 16:32:29,858 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7601, 1.7598, 1.6676, 1.4719, 1.4402, 1.4910, 0.2574, 0.7039], + device='cuda:3'), covar=tensor([0.0692, 0.0667, 0.0448, 0.0721, 0.1271, 0.0796, 0.1388, 0.1219], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0366, 0.0371, 0.0396, 0.0474, 0.0401, 0.0349, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 16:32:55,651 INFO [train.py:903] (3/4) Epoch 30, batch 3500, loss[loss=0.1769, simple_loss=0.2553, pruned_loss=0.04926, over 19735.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2812, pruned_loss=0.05903, over 3793554.65 frames. ], batch size: 46, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:33:41,237 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=201548.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:33:58,213 INFO [train.py:903] (3/4) Epoch 30, batch 3550, loss[loss=0.1805, simple_loss=0.2632, pruned_loss=0.04885, over 19774.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2811, pruned_loss=0.05839, over 3808226.98 frames. ], batch size: 54, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:34:35,156 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.339e+02 4.645e+02 6.006e+02 7.208e+02 1.141e+03, threshold=1.201e+03, percent-clipped=0.0 +2023-04-03 16:34:37,086 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.45 vs. limit=2.0 +2023-04-03 16:35:01,955 INFO [train.py:903] (3/4) Epoch 30, batch 3600, loss[loss=0.1807, simple_loss=0.2566, pruned_loss=0.05243, over 19308.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2813, pruned_loss=0.0585, over 3808406.95 frames. ], batch size: 44, lr: 2.73e-03, grad_scale: 8.0 +2023-04-03 16:35:06,946 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=201615.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:35:14,966 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4613, 1.4757, 1.5805, 1.6099, 1.8020, 1.9619, 1.8507, 0.6483], + device='cuda:3'), covar=tensor([0.2382, 0.4342, 0.2752, 0.1926, 0.1604, 0.2340, 0.1398, 0.4917], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0677, 0.0764, 0.0514, 0.0637, 0.0551, 0.0670, 0.0579], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 16:35:32,771 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=201636.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:36:04,985 INFO [train.py:903] (3/4) Epoch 30, batch 3650, loss[loss=0.2029, simple_loss=0.2732, pruned_loss=0.06633, over 19141.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2808, pruned_loss=0.05898, over 3789162.58 frames. ], batch size: 42, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:36:05,310 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201662.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:36:20,487 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0817, 4.4505, 4.8185, 4.8375, 1.7797, 4.5784, 3.8952, 4.5518], + device='cuda:3'), covar=tensor([0.1720, 0.0885, 0.0663, 0.0659, 0.6408, 0.0864, 0.0707, 0.1129], + device='cuda:3'), in_proj_covar=tensor([0.0824, 0.0798, 0.1008, 0.0886, 0.0874, 0.0775, 0.0596, 0.0942], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 16:36:27,550 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201679.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:36:42,039 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.871e+02 4.844e+02 5.974e+02 7.556e+02 1.962e+03, threshold=1.195e+03, percent-clipped=4.0 +2023-04-03 16:37:09,545 INFO [train.py:903] (3/4) Epoch 30, batch 3700, loss[loss=0.1982, simple_loss=0.2687, pruned_loss=0.06384, over 19735.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2809, pruned_loss=0.05843, over 3797993.66 frames. ], batch size: 45, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:37:16,772 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.6180, 1.5274, 1.7394, 1.5549, 3.2225, 1.1739, 2.5741, 3.6260], + device='cuda:3'), covar=tensor([0.0512, 0.2817, 0.2668, 0.1870, 0.0683, 0.2565, 0.1239, 0.0248], + device='cuda:3'), in_proj_covar=tensor([0.0431, 0.0386, 0.0404, 0.0358, 0.0390, 0.0364, 0.0405, 0.0427], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:37:20,704 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-04-03 16:37:23,588 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=201724.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:37:31,413 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=201730.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:37:57,747 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=201751.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:38:11,311 INFO [train.py:903] (3/4) Epoch 30, batch 3750, loss[loss=0.1749, simple_loss=0.2498, pruned_loss=0.04999, over 19760.00 frames. ], tot_loss[loss=0.199, simple_loss=0.281, pruned_loss=0.05848, over 3797380.23 frames. ], batch size: 46, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:38:47,452 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.241e+02 4.921e+02 6.534e+02 8.381e+02 2.079e+03, threshold=1.307e+03, percent-clipped=7.0 +2023-04-03 16:39:04,352 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=201804.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:39:14,021 INFO [train.py:903] (3/4) Epoch 30, batch 3800, loss[loss=0.2376, simple_loss=0.3245, pruned_loss=0.0754, over 19536.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2818, pruned_loss=0.05883, over 3819576.36 frames. ], batch size: 54, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:39:35,678 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=201829.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:39:41,263 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9_rvb from training. Duration: 29.1166875 +2023-04-03 16:39:41,670 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.3234, 3.0011, 2.3674, 2.3437, 2.3162, 2.6624, 1.0445, 2.1850], + device='cuda:3'), covar=tensor([0.0712, 0.0643, 0.0780, 0.1324, 0.1103, 0.1167, 0.1582, 0.1230], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0365, 0.0371, 0.0395, 0.0473, 0.0401, 0.0348, 0.0351], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 16:39:46,578 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-04-03 16:40:15,946 INFO [train.py:903] (3/4) Epoch 30, batch 3850, loss[loss=0.2107, simple_loss=0.2955, pruned_loss=0.06296, over 18711.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2829, pruned_loss=0.05961, over 3812800.67 frames. ], batch size: 74, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:40:51,662 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.891e+02 4.977e+02 6.719e+02 8.916e+02 2.147e+03, threshold=1.344e+03, percent-clipped=8.0 +2023-04-03 16:41:00,222 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9749, 2.0339, 2.3095, 2.5680, 1.9236, 2.4640, 2.2759, 2.0025], + device='cuda:3'), covar=tensor([0.4492, 0.4288, 0.2124, 0.2707, 0.4498, 0.2448, 0.5501, 0.3843], + device='cuda:3'), in_proj_covar=tensor([0.0956, 0.1039, 0.0757, 0.0966, 0.0934, 0.0873, 0.0872, 0.0821], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:41:18,287 INFO [train.py:903] (3/4) Epoch 30, batch 3900, loss[loss=0.2094, simple_loss=0.281, pruned_loss=0.06891, over 19692.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.282, pruned_loss=0.05951, over 3801776.75 frames. ], batch size: 53, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:42:10,307 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0900, 2.1837, 2.4925, 2.7384, 2.1166, 2.6377, 2.3695, 2.2380], + device='cuda:3'), covar=tensor([0.4655, 0.4276, 0.2103, 0.2461, 0.4301, 0.2276, 0.5498, 0.3491], + device='cuda:3'), in_proj_covar=tensor([0.0956, 0.1039, 0.0756, 0.0965, 0.0932, 0.0874, 0.0872, 0.0821], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:42:20,922 INFO [train.py:903] (3/4) Epoch 30, batch 3950, loss[loss=0.1875, simple_loss=0.2704, pruned_loss=0.05233, over 19682.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2808, pruned_loss=0.05867, over 3805395.83 frames. ], batch size: 53, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:42:20,946 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-04-03 16:42:51,066 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=201986.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:42:56,914 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.990e+02 4.814e+02 5.725e+02 7.208e+02 1.816e+03, threshold=1.145e+03, percent-clipped=2.0 +2023-04-03 16:43:11,427 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-04-03 16:43:16,600 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=202006.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:43:18,126 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=202007.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:43:19,120 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=202008.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:43:23,682 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202011.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:43:24,452 INFO [train.py:903] (3/4) Epoch 30, batch 4000, loss[loss=0.225, simple_loss=0.3064, pruned_loss=0.07185, over 19389.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2814, pruned_loss=0.05887, over 3808408.10 frames. ], batch size: 70, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:43:38,593 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=202023.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:43:42,204 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1560, 2.8096, 1.7914, 1.7901, 2.6322, 1.6868, 1.4800, 2.3553], + device='cuda:3'), covar=tensor([0.1216, 0.0949, 0.1109, 0.1023, 0.0597, 0.1262, 0.1037, 0.0661], + device='cuda:3'), in_proj_covar=tensor([0.0304, 0.0322, 0.0345, 0.0277, 0.0255, 0.0350, 0.0292, 0.0277], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:43:47,429 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.3451, 1.3392, 1.4772, 1.4929, 1.8936, 1.7589, 1.9090, 1.1524], + device='cuda:3'), covar=tensor([0.2132, 0.3731, 0.2363, 0.1756, 0.1386, 0.2052, 0.1317, 0.4421], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0677, 0.0764, 0.0514, 0.0637, 0.0551, 0.0670, 0.0579], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 16:43:49,619 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202032.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:44:08,192 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1_rvb from training. Duration: 0.7545625 +2023-04-03 16:44:27,670 INFO [train.py:903] (3/4) Epoch 30, batch 4050, loss[loss=0.1593, simple_loss=0.2401, pruned_loss=0.0392, over 19752.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2813, pruned_loss=0.05878, over 3814821.97 frames. ], batch size: 47, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:44:35,678 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=202068.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:45:02,976 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.018e+02 4.674e+02 6.115e+02 7.260e+02 2.667e+03, threshold=1.223e+03, percent-clipped=7.0 +2023-04-03 16:45:30,404 INFO [train.py:903] (3/4) Epoch 30, batch 4100, loss[loss=0.2033, simple_loss=0.2886, pruned_loss=0.059, over 19601.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2813, pruned_loss=0.05858, over 3819029.27 frames. ], batch size: 57, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:45:42,124 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=202121.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:45:47,978 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.9139, 4.3434, 4.6693, 4.6496, 1.8294, 4.4117, 3.8128, 4.4121], + device='cuda:3'), covar=tensor([0.1740, 0.0898, 0.0612, 0.0760, 0.6260, 0.0978, 0.0685, 0.1050], + device='cuda:3'), in_proj_covar=tensor([0.0826, 0.0803, 0.1014, 0.0893, 0.0877, 0.0777, 0.0600, 0.0944], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 16:45:48,423 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 16:46:03,041 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=202138.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:46:04,982 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1_rvb from training. Duration: 0.97725 +2023-04-03 16:46:32,902 INFO [train.py:903] (3/4) Epoch 30, batch 4150, loss[loss=0.1889, simple_loss=0.271, pruned_loss=0.05337, over 19543.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2823, pruned_loss=0.05934, over 3805592.84 frames. ], batch size: 54, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:46:45,118 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-04-03 16:46:59,282 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=202183.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:47:08,854 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.767e+02 4.823e+02 5.653e+02 7.013e+02 1.196e+03, threshold=1.131e+03, percent-clipped=0.0 +2023-04-03 16:47:27,798 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.0633, 5.1184, 5.9280, 5.9188, 1.9889, 5.6142, 4.6869, 5.6033], + device='cuda:3'), covar=tensor([0.1842, 0.0853, 0.0577, 0.0642, 0.6567, 0.0840, 0.0641, 0.1142], + device='cuda:3'), in_proj_covar=tensor([0.0829, 0.0805, 0.1015, 0.0894, 0.0878, 0.0778, 0.0601, 0.0947], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 16:47:34,531 INFO [train.py:903] (3/4) Epoch 30, batch 4200, loss[loss=0.1774, simple_loss=0.2484, pruned_loss=0.05322, over 18724.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.281, pruned_loss=0.05838, over 3819051.83 frames. ], batch size: 41, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:47:37,784 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-04-03 16:48:35,837 INFO [train.py:903] (3/4) Epoch 30, batch 4250, loss[loss=0.1893, simple_loss=0.2729, pruned_loss=0.0528, over 19688.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2824, pruned_loss=0.05927, over 3834586.03 frames. ], batch size: 53, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:48:52,007 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9_rvb from training. Duration: 29.816625 +2023-04-03 16:48:57,127 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7592, 1.6064, 1.6039, 2.4749, 1.7893, 1.9774, 1.9908, 1.8031], + device='cuda:3'), covar=tensor([0.0843, 0.0924, 0.1033, 0.0596, 0.0822, 0.0781, 0.0883, 0.0713], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0223, 0.0228, 0.0240, 0.0227, 0.0216, 0.0187, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 16:49:03,283 WARNING [train.py:1073] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-04-03 16:49:12,501 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.375e+02 5.334e+02 6.771e+02 9.277e+02 2.113e+03, threshold=1.354e+03, percent-clipped=7.0 +2023-04-03 16:49:38,454 INFO [train.py:903] (3/4) Epoch 30, batch 4300, loss[loss=0.1818, simple_loss=0.2649, pruned_loss=0.04933, over 19800.00 frames. ], tot_loss[loss=0.2, simple_loss=0.282, pruned_loss=0.05896, over 3836981.57 frames. ], batch size: 49, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:50:28,702 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=202352.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:50:34,353 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-04-03 16:50:41,242 INFO [train.py:903] (3/4) Epoch 30, batch 4350, loss[loss=0.2155, simple_loss=0.3099, pruned_loss=0.06052, over 19744.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2815, pruned_loss=0.05865, over 3840154.64 frames. ], batch size: 63, lr: 2.72e-03, grad_scale: 4.0 +2023-04-03 16:50:59,687 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=202377.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:51:05,029 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.6579, 4.2399, 2.6774, 3.6989, 0.8636, 4.2320, 4.0607, 4.1710], + device='cuda:3'), covar=tensor([0.0595, 0.0942, 0.2069, 0.0936, 0.4135, 0.0701, 0.1006, 0.1290], + device='cuda:3'), in_proj_covar=tensor([0.0541, 0.0440, 0.0525, 0.0365, 0.0416, 0.0464, 0.0459, 0.0495], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 16:51:18,783 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.299e+02 4.712e+02 5.570e+02 7.585e+02 1.545e+03, threshold=1.114e+03, percent-clipped=3.0 +2023-04-03 16:51:22,541 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=202394.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:51:31,130 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202402.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:51:43,039 INFO [train.py:903] (3/4) Epoch 30, batch 4400, loss[loss=0.1991, simple_loss=0.2709, pruned_loss=0.06369, over 19378.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2817, pruned_loss=0.05926, over 3824035.80 frames. ], batch size: 47, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:51:52,573 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202419.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:52:08,207 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_rvb from training. Duration: 25.285 +2023-04-03 16:52:17,004 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-04-03 16:52:17,410 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=202439.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:52:46,168 INFO [train.py:903] (3/4) Epoch 30, batch 4450, loss[loss=0.1765, simple_loss=0.2534, pruned_loss=0.04978, over 15995.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2801, pruned_loss=0.05821, over 3823572.53 frames. ], batch size: 35, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:52:48,945 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202464.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:52:52,378 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=202467.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:52:58,717 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.50 vs. limit=5.0 +2023-04-03 16:53:23,819 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.801e+02 4.882e+02 5.921e+02 7.342e+02 1.295e+03, threshold=1.184e+03, percent-clipped=2.0 +2023-04-03 16:53:39,216 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.6912, 1.8751, 2.1502, 1.9794, 3.2686, 2.7193, 3.5398, 1.8267], + device='cuda:3'), covar=tensor([0.2621, 0.4456, 0.2953, 0.1956, 0.1523, 0.2232, 0.1643, 0.4426], + device='cuda:3'), in_proj_covar=tensor([0.0561, 0.0682, 0.0769, 0.0519, 0.0640, 0.0555, 0.0673, 0.0583], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:53:48,741 INFO [train.py:903] (3/4) Epoch 30, batch 4500, loss[loss=0.232, simple_loss=0.3188, pruned_loss=0.07255, over 19601.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2809, pruned_loss=0.05844, over 3817755.61 frames. ], batch size: 57, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:54:50,692 INFO [train.py:903] (3/4) Epoch 30, batch 4550, loss[loss=0.2485, simple_loss=0.3314, pruned_loss=0.08285, over 19608.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2806, pruned_loss=0.05822, over 3810260.13 frames. ], batch size: 57, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:55:01,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9_rvb from training. Duration: 28.72225 +2023-04-03 16:55:25,955 WARNING [train.py:1073] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-04-03 16:55:27,094 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.113e+02 5.006e+02 6.254e+02 8.173e+02 1.576e+03, threshold=1.251e+03, percent-clipped=3.0 +2023-04-03 16:55:52,912 INFO [train.py:903] (3/4) Epoch 30, batch 4600, loss[loss=0.2123, simple_loss=0.295, pruned_loss=0.06479, over 19527.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2811, pruned_loss=0.05883, over 3815597.26 frames. ], batch size: 56, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:56:54,622 INFO [train.py:903] (3/4) Epoch 30, batch 4650, loss[loss=0.1953, simple_loss=0.2729, pruned_loss=0.05888, over 19609.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2808, pruned_loss=0.05846, over 3809919.84 frames. ], batch size: 50, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:57:12,731 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_rvb from training. Duration: 0.92 +2023-04-03 16:57:22,314 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8040, 1.9031, 2.1761, 2.2999, 1.7601, 2.2489, 2.1313, 1.9766], + device='cuda:3'), covar=tensor([0.4407, 0.4134, 0.2020, 0.2541, 0.4273, 0.2338, 0.5437, 0.3714], + device='cuda:3'), in_proj_covar=tensor([0.0953, 0.1036, 0.0753, 0.0964, 0.0930, 0.0871, 0.0868, 0.0820], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 16:57:23,006 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_rvb from training. Duration: 0.83 +2023-04-03 16:57:31,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.354e+02 5.029e+02 5.947e+02 7.618e+02 1.686e+03, threshold=1.189e+03, percent-clipped=3.0 +2023-04-03 16:57:55,568 INFO [train.py:903] (3/4) Epoch 30, batch 4700, loss[loss=0.1992, simple_loss=0.2876, pruned_loss=0.05534, over 19546.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2811, pruned_loss=0.05895, over 3803522.50 frames. ], batch size: 56, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:58:10,563 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=202723.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:58:18,351 WARNING [train.py:1073] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9_rvb from training. Duration: 0.92225 +2023-04-03 16:58:41,355 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=202748.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 16:58:59,217 INFO [train.py:903] (3/4) Epoch 30, batch 4750, loss[loss=0.1848, simple_loss=0.2785, pruned_loss=0.04556, over 19536.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2809, pruned_loss=0.05857, over 3798769.62 frames. ], batch size: 56, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 16:59:35,972 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.278e+02 5.055e+02 5.959e+02 7.636e+02 1.705e+03, threshold=1.192e+03, percent-clipped=6.0 +2023-04-03 17:00:01,990 INFO [train.py:903] (3/4) Epoch 30, batch 4800, loss[loss=0.1886, simple_loss=0.2789, pruned_loss=0.04911, over 19661.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2815, pruned_loss=0.05892, over 3808375.94 frames. ], batch size: 55, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:00:08,682 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-04-03 17:00:19,798 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.67 vs. limit=2.0 +2023-04-03 17:00:47,843 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=202848.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 17:01:03,241 INFO [train.py:903] (3/4) Epoch 30, batch 4850, loss[loss=0.1622, simple_loss=0.2399, pruned_loss=0.04221, over 19756.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2815, pruned_loss=0.05889, over 3802219.82 frames. ], batch size: 47, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:01:27,953 WARNING [train.py:1073] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1_rvb from training. Duration: 0.95 +2023-04-03 17:01:34,074 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3579, 1.9648, 2.1415, 2.8575, 1.9849, 2.4558, 2.4683, 2.3422], + device='cuda:3'), covar=tensor([0.0769, 0.0881, 0.0902, 0.0757, 0.0916, 0.0746, 0.0914, 0.0631], + device='cuda:3'), in_proj_covar=tensor([0.0213, 0.0224, 0.0228, 0.0240, 0.0227, 0.0216, 0.0188, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 17:01:41,373 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.446e+02 5.041e+02 6.050e+02 7.335e+02 2.031e+03, threshold=1.210e+03, percent-clipped=4.0 +2023-04-03 17:01:47,279 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1_rvb from training. Duration: 0.9409375 +2023-04-03 17:01:53,230 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-04-03 17:01:55,296 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-04-03 17:02:02,748 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.8566, 1.9647, 2.2939, 2.2944, 1.7748, 2.1946, 2.2466, 2.1486], + device='cuda:3'), covar=tensor([0.4248, 0.3990, 0.2013, 0.2546, 0.4229, 0.2392, 0.5112, 0.3462], + device='cuda:3'), in_proj_covar=tensor([0.0950, 0.1035, 0.0753, 0.0961, 0.0929, 0.0870, 0.0867, 0.0817], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:02:03,502 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1_rvb from training. Duration: 25.3818125 +2023-04-03 17:02:05,914 INFO [train.py:903] (3/4) Epoch 30, batch 4900, loss[loss=0.1668, simple_loss=0.2425, pruned_loss=0.04551, over 19732.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2806, pruned_loss=0.05851, over 3794341.59 frames. ], batch size: 45, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:02:25,014 WARNING [train.py:1073] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-04-03 17:03:07,819 INFO [train.py:903] (3/4) Epoch 30, batch 4950, loss[loss=0.1988, simple_loss=0.276, pruned_loss=0.0608, over 19465.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2816, pruned_loss=0.0595, over 3800538.20 frames. ], batch size: 49, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:03:22,407 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-04-03 17:03:27,587 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-04-03 17:03:43,835 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.897e+02 4.771e+02 6.104e+02 8.250e+02 2.222e+03, threshold=1.221e+03, percent-clipped=2.0 +2023-04-03 17:03:46,198 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-04-03 17:04:09,798 INFO [train.py:903] (3/4) Epoch 30, batch 5000, loss[loss=0.1782, simple_loss=0.2588, pruned_loss=0.04882, over 19477.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2812, pruned_loss=0.05936, over 3795737.56 frames. ], batch size: 49, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:04:16,534 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_rvb from training. Duration: 27.14 +2023-04-03 17:04:28,733 WARNING [train.py:1073] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-04-03 17:05:02,533 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203056.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:05:09,000 INFO [train.py:903] (3/4) Epoch 30, batch 5050, loss[loss=0.2251, simple_loss=0.3086, pruned_loss=0.07079, over 19586.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2807, pruned_loss=0.05871, over 3803562.77 frames. ], batch size: 52, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:05:18,170 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([6.2366, 5.6773, 3.0654, 4.9335, 1.3989, 5.8219, 5.6251, 5.8136], + device='cuda:3'), covar=tensor([0.0346, 0.0748, 0.1887, 0.0728, 0.3534, 0.0401, 0.0816, 0.0834], + device='cuda:3'), in_proj_covar=tensor([0.0538, 0.0437, 0.0520, 0.0361, 0.0413, 0.0461, 0.0454, 0.0491], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 17:05:45,255 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9_rvb from training. Duration: 26.62775 +2023-04-03 17:05:46,418 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.400e+02 4.511e+02 5.411e+02 6.959e+02 2.884e+03, threshold=1.082e+03, percent-clipped=1.0 +2023-04-03 17:05:59,241 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203103.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:05:59,653 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.63 vs. limit=5.0 +2023-04-03 17:06:10,217 INFO [train.py:903] (3/4) Epoch 30, batch 5100, loss[loss=0.1784, simple_loss=0.2602, pruned_loss=0.04829, over 19842.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2808, pruned_loss=0.05867, over 3813550.99 frames. ], batch size: 52, lr: 2.72e-03, grad_scale: 8.0 +2023-04-03 17:06:23,794 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-04-03 17:06:27,358 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0005-9467_rvb from training. Duration: 25.035 +2023-04-03 17:06:28,814 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0915, 1.7316, 1.8700, 2.8942, 2.1186, 2.3648, 2.3747, 2.1284], + device='cuda:3'), covar=tensor([0.0821, 0.0921, 0.1001, 0.0728, 0.0826, 0.0769, 0.0860, 0.0677], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0227, 0.0239, 0.0226, 0.0216, 0.0187, 0.0208], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 17:06:30,778 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_rvb from training. Duration: 27.92 +2023-04-03 17:07:12,614 INFO [train.py:903] (3/4) Epoch 30, batch 5150, loss[loss=0.2033, simple_loss=0.2856, pruned_loss=0.06048, over 19280.00 frames. ], tot_loss[loss=0.198, simple_loss=0.28, pruned_loss=0.058, over 3820608.92 frames. ], batch size: 66, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:07:27,945 WARNING [train.py:1073] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-04-03 17:07:49,458 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.244e+02 5.115e+02 6.435e+02 8.085e+02 2.061e+03, threshold=1.287e+03, percent-clipped=7.0 +2023-04-03 17:07:49,683 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=203192.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:08:02,564 WARNING [train.py:1073] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1_rvb from training. Duration: 0.9681875 +2023-04-03 17:08:14,981 INFO [train.py:903] (3/4) Epoch 30, batch 5200, loss[loss=0.1999, simple_loss=0.2794, pruned_loss=0.06017, over 19467.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2802, pruned_loss=0.05821, over 3816154.97 frames. ], batch size: 49, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:08:30,103 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-04-03 17:09:14,153 WARNING [train.py:1073] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-04-03 17:09:16,379 INFO [train.py:903] (3/4) Epoch 30, batch 5250, loss[loss=0.1951, simple_loss=0.2676, pruned_loss=0.06132, over 18198.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2802, pruned_loss=0.05821, over 3814116.03 frames. ], batch size: 40, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:09:53,738 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.978e+02 4.929e+02 6.276e+02 8.119e+02 2.486e+03, threshold=1.255e+03, percent-clipped=4.0 +2023-04-03 17:10:12,248 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=203307.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 17:10:17,558 INFO [train.py:903] (3/4) Epoch 30, batch 5300, loss[loss=0.1593, simple_loss=0.249, pruned_loss=0.03477, over 19805.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2804, pruned_loss=0.05846, over 3822229.48 frames. ], batch size: 49, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:10:36,829 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-04-03 17:11:02,659 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.9431, 4.4871, 2.8345, 3.9672, 1.1136, 4.5100, 4.3681, 4.4535], + device='cuda:3'), covar=tensor([0.0532, 0.0938, 0.1871, 0.0807, 0.3937, 0.0592, 0.0918, 0.1044], + device='cuda:3'), in_proj_covar=tensor([0.0537, 0.0436, 0.0519, 0.0361, 0.0412, 0.0461, 0.0453, 0.0490], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 17:11:18,772 INFO [train.py:903] (3/4) Epoch 30, batch 5350, loss[loss=0.2123, simple_loss=0.2998, pruned_loss=0.06239, over 19326.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2806, pruned_loss=0.05875, over 3821889.47 frames. ], batch size: 70, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:11:53,869 WARNING [train.py:1073] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9_rvb from training. Duration: 26.438875 +2023-04-03 17:11:56,057 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.690e+02 5.524e+02 6.450e+02 8.325e+02 1.910e+03, threshold=1.290e+03, percent-clipped=5.0 +2023-04-03 17:12:05,339 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=203400.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:12:20,979 INFO [train.py:903] (3/4) Epoch 30, batch 5400, loss[loss=0.2249, simple_loss=0.2874, pruned_loss=0.08123, over 19432.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2814, pruned_loss=0.05928, over 3816229.77 frames. ], batch size: 48, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:13:02,976 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=203447.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:13:20,914 INFO [train.py:903] (3/4) Epoch 30, batch 5450, loss[loss=0.2132, simple_loss=0.3047, pruned_loss=0.06084, over 19616.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2817, pruned_loss=0.05945, over 3822567.83 frames. ], batch size: 57, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:13:21,398 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3065, 2.3230, 2.5538, 2.9793, 2.3040, 2.8160, 2.5081, 2.3310], + device='cuda:3'), covar=tensor([0.4341, 0.4567, 0.2036, 0.2872, 0.4739, 0.2514, 0.5095, 0.3601], + device='cuda:3'), in_proj_covar=tensor([0.0954, 0.1036, 0.0753, 0.0963, 0.0930, 0.0869, 0.0867, 0.0819], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:13:50,543 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203485.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:13:59,460 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.086e+02 4.910e+02 6.531e+02 8.117e+02 1.984e+03, threshold=1.306e+03, percent-clipped=3.0 +2023-04-03 17:14:10,106 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4526, 1.3473, 1.8996, 1.6827, 3.0562, 4.5898, 4.4481, 5.0908], + device='cuda:3'), covar=tensor([0.1640, 0.4149, 0.3628, 0.2466, 0.0676, 0.0204, 0.0187, 0.0197], + device='cuda:3'), in_proj_covar=tensor([0.0282, 0.0336, 0.0368, 0.0275, 0.0258, 0.0200, 0.0221, 0.0283], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 17:14:23,410 INFO [train.py:903] (3/4) Epoch 30, batch 5500, loss[loss=0.2134, simple_loss=0.2973, pruned_loss=0.06476, over 19429.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2823, pruned_loss=0.05965, over 3816903.59 frames. ], batch size: 70, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:14:26,977 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=203515.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:14:48,302 WARNING [train.py:1073] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9_rvb from training. Duration: 27.47775 +2023-04-03 17:15:24,339 INFO [train.py:903] (3/4) Epoch 30, batch 5550, loss[loss=0.1877, simple_loss=0.2673, pruned_loss=0.05406, over 19842.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2819, pruned_loss=0.05918, over 3818337.94 frames. ], batch size: 52, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:15:24,659 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=203562.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:15:26,892 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=203563.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:15:31,867 WARNING [train.py:1073] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-04-03 17:15:56,784 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=203588.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:16:01,699 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.118e+02 4.952e+02 6.311e+02 7.819e+02 2.120e+03, threshold=1.262e+03, percent-clipped=2.0 +2023-04-03 17:16:03,243 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7697, 1.5695, 1.6581, 2.2125, 1.6397, 1.9827, 1.9702, 1.8317], + device='cuda:3'), covar=tensor([0.0808, 0.0895, 0.0921, 0.0713, 0.0887, 0.0768, 0.0858, 0.0695], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0223, 0.0227, 0.0238, 0.0226, 0.0215, 0.0187, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 17:16:20,043 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([5.2841, 5.2293, 6.1039, 6.1064, 1.9994, 5.7507, 4.8251, 5.8009], + device='cuda:3'), covar=tensor([0.1787, 0.0718, 0.0545, 0.0641, 0.6829, 0.0733, 0.0627, 0.1099], + device='cuda:3'), in_proj_covar=tensor([0.0830, 0.0807, 0.1016, 0.0892, 0.0879, 0.0779, 0.0600, 0.0947], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 17:16:22,145 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-04-03 17:16:27,763 INFO [train.py:903] (3/4) Epoch 30, batch 5600, loss[loss=0.2151, simple_loss=0.2948, pruned_loss=0.06769, over 19642.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2827, pruned_loss=0.05971, over 3823195.68 frames. ], batch size: 50, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:16:48,417 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-04-03 17:17:02,907 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.0960, 2.1437, 2.4509, 2.6255, 2.0676, 2.5329, 2.4177, 2.2525], + device='cuda:3'), covar=tensor([0.4437, 0.4299, 0.1998, 0.2716, 0.4369, 0.2468, 0.5141, 0.3533], + device='cuda:3'), in_proj_covar=tensor([0.0955, 0.1039, 0.0755, 0.0966, 0.0933, 0.0871, 0.0869, 0.0821], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:17:29,530 INFO [train.py:903] (3/4) Epoch 30, batch 5650, loss[loss=0.1989, simple_loss=0.2831, pruned_loss=0.05731, over 17451.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2814, pruned_loss=0.0591, over 3824633.56 frames. ], batch size: 101, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:18:06,343 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.203e+02 4.913e+02 5.869e+02 7.597e+02 1.607e+03, threshold=1.174e+03, percent-clipped=2.0 +2023-04-03 17:18:18,649 WARNING [train.py:1073] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9_rvb from training. Duration: 28.0944375 +2023-04-03 17:18:29,969 INFO [train.py:903] (3/4) Epoch 30, batch 5700, loss[loss=0.1897, simple_loss=0.2643, pruned_loss=0.05755, over 19716.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2821, pruned_loss=0.05977, over 3816981.85 frames. ], batch size: 46, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:19:03,862 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4471, 1.4943, 1.8379, 1.7372, 2.8293, 2.2912, 2.9877, 1.2561], + device='cuda:3'), covar=tensor([0.2774, 0.4799, 0.3064, 0.2130, 0.1483, 0.2361, 0.1474, 0.5171], + device='cuda:3'), in_proj_covar=tensor([0.0562, 0.0684, 0.0769, 0.0518, 0.0641, 0.0555, 0.0674, 0.0584], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:19:31,945 INFO [train.py:903] (3/4) Epoch 30, batch 5750, loss[loss=0.1796, simple_loss=0.2764, pruned_loss=0.04139, over 19501.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2818, pruned_loss=0.05942, over 3825492.62 frames. ], batch size: 64, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:19:35,373 WARNING [train.py:1073] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9_rvb from training. Duration: 33.038875 +2023-04-03 17:19:35,667 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203765.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:19:44,462 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=203771.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:19:45,208 WARNING [train.py:1073] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-04-03 17:19:51,013 WARNING [train.py:1073] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-04-03 17:20:00,635 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203784.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:20:09,168 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.085e+02 5.021e+02 5.985e+02 7.897e+02 1.857e+03, threshold=1.197e+03, percent-clipped=7.0 +2023-04-03 17:20:15,341 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=203796.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:20:35,340 INFO [train.py:903] (3/4) Epoch 30, batch 5800, loss[loss=0.1843, simple_loss=0.2732, pruned_loss=0.04766, over 19666.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2814, pruned_loss=0.05934, over 3820998.91 frames. ], batch size: 58, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:20:38,237 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-04-03 17:20:43,827 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=203818.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:20:55,518 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.4825, 2.1333, 1.7428, 1.3108, 1.9711, 1.2654, 1.3547, 1.9379], + device='cuda:3'), covar=tensor([0.1060, 0.0845, 0.1062, 0.1179, 0.0691, 0.1442, 0.0819, 0.0534], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0324, 0.0347, 0.0279, 0.0257, 0.0351, 0.0293, 0.0281], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 17:20:56,429 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=203829.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:21:13,724 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=203843.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:21:37,905 INFO [train.py:903] (3/4) Epoch 30, batch 5850, loss[loss=0.2047, simple_loss=0.2906, pruned_loss=0.05937, over 19752.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2809, pruned_loss=0.05901, over 3824120.59 frames. ], batch size: 54, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:21:41,711 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203865.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:22:15,806 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.367e+02 4.919e+02 5.824e+02 7.977e+02 1.569e+03, threshold=1.165e+03, percent-clipped=4.0 +2023-04-03 17:22:39,479 INFO [train.py:903] (3/4) Epoch 30, batch 5900, loss[loss=0.1612, simple_loss=0.2446, pruned_loss=0.03885, over 19481.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.282, pruned_loss=0.0598, over 3808513.79 frames. ], batch size: 49, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:22:42,668 WARNING [train.py:1073] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-04-03 17:22:47,262 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=203918.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:23:06,099 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9_rvb from training. Duration: 27.511125 +2023-04-03 17:23:19,559 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=203944.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:23:25,701 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.9964, 2.1298, 2.3706, 2.6792, 2.0824, 2.5856, 2.3267, 2.1586], + device='cuda:3'), covar=tensor([0.4460, 0.4326, 0.2118, 0.2808, 0.4708, 0.2475, 0.5266, 0.3634], + device='cuda:3'), in_proj_covar=tensor([0.0952, 0.1036, 0.0753, 0.0965, 0.0931, 0.0870, 0.0865, 0.0818], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:23:40,674 INFO [train.py:903] (3/4) Epoch 30, batch 5950, loss[loss=0.2146, simple_loss=0.2947, pruned_loss=0.06729, over 19681.00 frames. ], tot_loss[loss=0.202, simple_loss=0.283, pruned_loss=0.06055, over 3812094.83 frames. ], batch size: 59, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:24:18,145 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.982e+02 5.092e+02 6.183e+02 7.281e+02 1.501e+03, threshold=1.237e+03, percent-clipped=4.0 +2023-04-03 17:24:45,151 INFO [train.py:903] (3/4) Epoch 30, batch 6000, loss[loss=0.2165, simple_loss=0.3002, pruned_loss=0.06643, over 19776.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2827, pruned_loss=0.05992, over 3818004.35 frames. ], batch size: 56, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:24:45,151 INFO [train.py:928] (3/4) Computing validation loss +2023-04-03 17:24:58,727 INFO [train.py:937] (3/4) Epoch 30, validation: loss=0.167, simple_loss=0.2658, pruned_loss=0.03407, over 944034.00 frames. +2023-04-03 17:24:58,728 INFO [train.py:938] (3/4) Maximum memory allocated so far is 18741MB +2023-04-03 17:26:02,080 INFO [train.py:903] (3/4) Epoch 30, batch 6050, loss[loss=0.2038, simple_loss=0.2899, pruned_loss=0.05887, over 19349.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2816, pruned_loss=0.05934, over 3812667.12 frames. ], batch size: 70, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:26:38,219 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.289e+02 5.054e+02 6.352e+02 7.854e+02 1.582e+03, threshold=1.270e+03, percent-clipped=1.0 +2023-04-03 17:26:40,404 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-04-03 17:26:42,110 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=204095.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:27:00,737 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204109.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:27:04,111 INFO [train.py:903] (3/4) Epoch 30, batch 6100, loss[loss=0.1891, simple_loss=0.2591, pruned_loss=0.05957, over 19787.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2814, pruned_loss=0.05963, over 3812553.66 frames. ], batch size: 48, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:27:23,325 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204128.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:28:04,387 INFO [train.py:903] (3/4) Epoch 30, batch 6150, loss[loss=0.1897, simple_loss=0.2787, pruned_loss=0.05035, over 19781.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2824, pruned_loss=0.05969, over 3814580.81 frames. ], batch size: 54, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:28:04,790 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([0.9855, 1.2707, 1.5921, 0.6764, 2.0216, 2.5058, 2.2039, 2.6873], + device='cuda:3'), covar=tensor([0.1546, 0.3922, 0.3501, 0.2940, 0.0667, 0.0278, 0.0346, 0.0376], + device='cuda:3'), in_proj_covar=tensor([0.0281, 0.0336, 0.0368, 0.0275, 0.0258, 0.0200, 0.0221, 0.0282], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0004, 0.0004, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003], + device='cuda:3') +2023-04-03 17:28:29,840 INFO [zipformer.py:1188] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=204181.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:28:34,976 WARNING [train.py:1073] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9_rvb from training. Duration: 31.02225 +2023-04-03 17:28:42,856 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.206e+02 5.061e+02 6.485e+02 7.945e+02 1.594e+03, threshold=1.297e+03, percent-clipped=2.0 +2023-04-03 17:28:52,863 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204200.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:29:03,794 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204209.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:29:07,249 INFO [train.py:903] (3/4) Epoch 30, batch 6200, loss[loss=0.1621, simple_loss=0.2472, pruned_loss=0.03847, over 19577.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2825, pruned_loss=0.05999, over 3794078.86 frames. ], batch size: 52, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:29:24,454 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204224.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:29:25,562 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204225.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:29:29,110 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7231, 1.7574, 1.6204, 1.4333, 1.4636, 1.4418, 0.3032, 0.6721], + device='cuda:3'), covar=tensor([0.0722, 0.0721, 0.0498, 0.0808, 0.1299, 0.0850, 0.1374, 0.1272], + device='cuda:3'), in_proj_covar=tensor([0.0369, 0.0369, 0.0372, 0.0397, 0.0476, 0.0402, 0.0349, 0.0352], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 17:29:45,851 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204243.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:30:09,328 INFO [train.py:903] (3/4) Epoch 30, batch 6250, loss[loss=0.2457, simple_loss=0.3135, pruned_loss=0.08895, over 12665.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2826, pruned_loss=0.06, over 3787707.71 frames. ], batch size: 136, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:30:10,640 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204262.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:30:18,766 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.87 vs. limit=5.0 +2023-04-03 17:30:38,604 WARNING [train.py:1073] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9_rvb from training. Duration: 25.988875 +2023-04-03 17:30:46,053 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.091e+02 5.004e+02 5.938e+02 7.424e+02 1.100e+03, threshold=1.188e+03, percent-clipped=0.0 +2023-04-03 17:31:10,306 INFO [train.py:903] (3/4) Epoch 30, batch 6300, loss[loss=0.198, simple_loss=0.2746, pruned_loss=0.06071, over 19628.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2836, pruned_loss=0.06032, over 3795528.39 frames. ], batch size: 50, lr: 2.71e-03, grad_scale: 4.0 +2023-04-03 17:31:24,177 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204324.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:32:11,321 INFO [train.py:903] (3/4) Epoch 30, batch 6350, loss[loss=0.1818, simple_loss=0.2554, pruned_loss=0.05407, over 19409.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2822, pruned_loss=0.0598, over 3799189.71 frames. ], batch size: 48, lr: 2.71e-03, grad_scale: 4.0 +2023-04-03 17:32:30,587 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204377.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:32:50,298 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.055e+02 4.728e+02 6.049e+02 7.442e+02 1.533e+03, threshold=1.210e+03, percent-clipped=4.0 +2023-04-03 17:33:07,776 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.1106, 2.0517, 1.9392, 1.7681, 1.6680, 1.7716, 0.5899, 1.0532], + device='cuda:3'), covar=tensor([0.0734, 0.0703, 0.0507, 0.0813, 0.1241, 0.0925, 0.1448, 0.1227], + device='cuda:3'), in_proj_covar=tensor([0.0367, 0.0366, 0.0370, 0.0394, 0.0472, 0.0398, 0.0347, 0.0350], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 17:33:12,664 INFO [train.py:903] (3/4) Epoch 30, batch 6400, loss[loss=0.1831, simple_loss=0.2566, pruned_loss=0.05486, over 19387.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2821, pruned_loss=0.05974, over 3810009.92 frames. ], batch size: 48, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:33:45,718 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204439.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:33:45,889 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.8155, 3.3157, 3.3436, 3.3431, 1.4351, 3.2119, 2.7960, 3.1476], + device='cuda:3'), covar=tensor([0.1855, 0.1133, 0.0846, 0.0994, 0.5936, 0.1227, 0.0896, 0.1332], + device='cuda:3'), in_proj_covar=tensor([0.0825, 0.0803, 0.1009, 0.0883, 0.0872, 0.0774, 0.0597, 0.0940], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-04-03 17:34:00,722 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.2392, 1.2937, 1.2834, 1.1097, 1.0980, 1.1492, 0.1023, 0.3863], + device='cuda:3'), covar=tensor([0.0830, 0.0744, 0.0535, 0.0698, 0.1434, 0.0778, 0.1517, 0.1300], + device='cuda:3'), in_proj_covar=tensor([0.0368, 0.0367, 0.0370, 0.0395, 0.0474, 0.0399, 0.0349, 0.0350], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 17:34:13,529 INFO [train.py:903] (3/4) Epoch 30, batch 6450, loss[loss=0.1527, simple_loss=0.231, pruned_loss=0.03725, over 19773.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2824, pruned_loss=0.05956, over 3811708.86 frames. ], batch size: 47, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:34:35,941 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204480.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:34:51,302 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.580e+02 4.981e+02 6.276e+02 8.010e+02 2.155e+03, threshold=1.255e+03, percent-clipped=8.0 +2023-04-03 17:34:54,949 WARNING [train.py:1073] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9_rvb from training. Duration: 28.638875 +2023-04-03 17:34:59,509 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204499.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:35:08,056 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204505.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:35:16,837 INFO [train.py:903] (3/4) Epoch 30, batch 6500, loss[loss=0.2262, simple_loss=0.3081, pruned_loss=0.07219, over 18137.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.282, pruned_loss=0.05925, over 3817193.02 frames. ], batch size: 83, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:35:19,109 WARNING [train.py:1073] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1_rvb from training. Duration: 0.836375 +2023-04-03 17:35:30,878 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204524.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:35:31,809 INFO [zipformer.py:1188] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=204525.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:36:09,153 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204554.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:36:11,508 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.7493, 1.6150, 1.6823, 2.1671, 1.7343, 2.0238, 1.9138, 1.8442], + device='cuda:3'), covar=tensor([0.0776, 0.0844, 0.0888, 0.0614, 0.0811, 0.0688, 0.0853, 0.0635], + device='cuda:3'), in_proj_covar=tensor([0.0212, 0.0221, 0.0227, 0.0238, 0.0226, 0.0215, 0.0186, 0.0207], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0004, 0.0005], + device='cuda:3') +2023-04-03 17:36:18,054 INFO [train.py:903] (3/4) Epoch 30, batch 6550, loss[loss=0.2193, simple_loss=0.3006, pruned_loss=0.06899, over 19589.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2811, pruned_loss=0.05876, over 3808205.09 frames. ], batch size: 61, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:36:28,054 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([4.0398, 3.6995, 2.6111, 3.3084, 0.8821, 3.7079, 3.5274, 3.6624], + device='cuda:3'), covar=tensor([0.0769, 0.1042, 0.1839, 0.0946, 0.3801, 0.0696, 0.1062, 0.1113], + device='cuda:3'), in_proj_covar=tensor([0.0535, 0.0434, 0.0517, 0.0358, 0.0407, 0.0457, 0.0451, 0.0486], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 17:36:40,518 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204580.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:36:56,886 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.390e+02 5.288e+02 6.916e+02 9.470e+02 2.608e+03, threshold=1.383e+03, percent-clipped=11.0 +2023-04-03 17:37:11,407 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204605.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:37:20,120 INFO [train.py:903] (3/4) Epoch 30, batch 6600, loss[loss=0.2133, simple_loss=0.2955, pruned_loss=0.06552, over 19661.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2812, pruned_loss=0.05847, over 3812034.07 frames. ], batch size: 58, lr: 2.71e-03, grad_scale: 8.0 +2023-04-03 17:37:46,930 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204633.0, num_to_drop=1, layers_to_drop={0} +2023-04-03 17:37:54,607 INFO [zipformer.py:1188] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=204640.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:38:17,215 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204658.0, num_to_drop=1, layers_to_drop={1} +2023-04-03 17:38:21,420 INFO [train.py:903] (3/4) Epoch 30, batch 6650, loss[loss=0.211, simple_loss=0.283, pruned_loss=0.06956, over 19691.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2816, pruned_loss=0.05876, over 3806524.67 frames. ], batch size: 53, lr: 2.70e-03, grad_scale: 8.0 +2023-04-03 17:38:59,516 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 3.417e+02 4.923e+02 6.497e+02 8.594e+02 3.232e+03, threshold=1.299e+03, percent-clipped=5.0 +2023-04-03 17:39:24,423 INFO [train.py:903] (3/4) Epoch 30, batch 6700, loss[loss=0.2168, simple_loss=0.2929, pruned_loss=0.07033, over 19539.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2808, pruned_loss=0.05838, over 3820267.64 frames. ], batch size: 56, lr: 2.70e-03, grad_scale: 4.0 +2023-04-03 17:39:42,300 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([3.5049, 1.6760, 1.8634, 1.7408, 3.2019, 1.4079, 2.5955, 3.5614], + device='cuda:3'), covar=tensor([0.0554, 0.2539, 0.2529, 0.1820, 0.0599, 0.2338, 0.1399, 0.0271], + device='cuda:3'), in_proj_covar=tensor([0.0426, 0.0381, 0.0402, 0.0357, 0.0386, 0.0361, 0.0401, 0.0423], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:3') +2023-04-03 17:40:21,757 INFO [train.py:903] (3/4) Epoch 30, batch 6750, loss[loss=0.2542, simple_loss=0.3245, pruned_loss=0.09194, over 19402.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2799, pruned_loss=0.05818, over 3807544.18 frames. ], batch size: 70, lr: 2.70e-03, grad_scale: 4.0 +2023-04-03 17:40:41,843 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([1.1832, 1.2505, 1.2278, 1.0594, 0.9381, 1.0865, 0.1674, 0.3582], + device='cuda:3'), covar=tensor([0.0923, 0.0881, 0.0552, 0.0744, 0.1725, 0.0921, 0.1614, 0.1491], + device='cuda:3'), in_proj_covar=tensor([0.0371, 0.0370, 0.0374, 0.0399, 0.0477, 0.0402, 0.0351, 0.0353], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-04-03 17:40:46,244 INFO [zipformer.py:2441] (3/4) attn_weights_entropy = tensor([2.3354, 2.3725, 2.5778, 2.9495, 2.3536, 2.8009, 2.5046, 2.3580], + device='cuda:3'), covar=tensor([0.4180, 0.4070, 0.1916, 0.2693, 0.4353, 0.2355, 0.4796, 0.3344], + device='cuda:3'), in_proj_covar=tensor([0.0955, 0.1039, 0.0754, 0.0967, 0.0935, 0.0871, 0.0869, 0.0821], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:3') +2023-04-03 17:40:58,539 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 2.758e+02 4.849e+02 6.262e+02 9.905e+02 2.863e+03, threshold=1.252e+03, percent-clipped=11.0 +2023-04-03 17:41:17,946 INFO [zipformer.py:1188] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=204810.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:41:19,756 INFO [train.py:903] (3/4) Epoch 30, batch 6800, loss[loss=0.2164, simple_loss=0.2907, pruned_loss=0.07102, over 19695.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2796, pruned_loss=0.05788, over 3811641.84 frames. ], batch size: 53, lr: 2.70e-03, grad_scale: 8.0 +2023-04-03 17:41:45,193 INFO [zipformer.py:1188] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=204835.0, num_to_drop=0, layers_to_drop=set() +2023-04-03 17:41:49,905 INFO [train.py:1171] (3/4) Done!