2023-11-12 20:52:32,875 INFO [train.py:699] (3/4) Training started
2023-11-12 20:52:32,876 INFO [train.py:709] (3/4) Device: cuda:3
2023-11-12 20:52:34,024 INFO [train.py:716] (3/4) {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': -1, 'log_interval': 50, 'valid_interval': 200, 'env_info': {'k2-version': '1.24.4', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '44a9d5682af9fd3ef77074777e15278ec6d390eb', 'k2-git-date': 'Wed Sep 27 11:22:55 2023', 'lhotse-version': '0.0.0+unknown.version', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'vits', 'icefall-git-sha1': 'f55e80a7-dirty', 'icefall-git-date': 'Mon Nov 6 15:05:49 2023', 'icefall-path': '/star-zw/workspace/tts/icefall_tts', 'k2-path': '/star-zw/workspace/k2/k2/k2/python/k2/__init__.py', 'lhotse-path': '/star-zw/workspace/lhotse/lhotse_dev/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-6-0423201309-7c68fd68fb-qfn6b', 'IP address': '10.177.58.19'}, 'sampling_rate': 22050, 'frame_shift': 256, 'frame_length': 1024, 'feature_dim': 513, 'n_mels': 80, 'lambda_adv': 1.0, 'lambda_mel': 45.0, 'lambda_feat_match': 2.0, 'lambda_dur': 1.0, 'lambda_kl': 1.0, 'world_size': 4, 'master_port': 12354, 'tensorboard': True, 'num_epochs': 1000, 'start_epoch': 1, 'exp_dir': PosixPath('vits/exp-g2p-conformer-text-encoder-new'), 'tokens': 'data/tokens.txt', 'lr': 0.0002, 'seed': 42, 'print_diagnostics': False, 'inf_check': False, 'save_every_n': 20, 'use_fp16': True, 'manifest_dir': PosixPath('data/spectrogram'), 'max_duration': 500, 'bucketing_sampler': True, 'num_buckets': 30, 'on_the_fly_feats': False, 'shuffle': True, 'drop_last': True, 'return_cuts': False, 'num_workers': 2, 'input_strategy': 'PrecomputedFeatures', 'blank_id': 0, 'oov_id': 2, 'vocab_size': 79}
2023-11-12 20:52:34,025 INFO [train.py:718] (3/4) About to create model
2023-11-12 20:52:37,019 INFO [train.py:724] (3/4) Number of parameters in generator: 35621746
2023-11-12 20:52:37,020 INFO [train.py:726] (3/4) Number of parameters in discriminator: 50974956
2023-11-12 20:52:37,020 INFO [train.py:727] (3/4) Total number of parameters: 86596702
2023-11-12 20:52:43,921 INFO [train.py:734] (3/4) Using DDP
2023-11-12 20:52:44,569 INFO [tts_datamodule.py:314] (3/4) About to get train cuts
2023-11-12 20:52:44,858 INFO [tts_datamodule.py:169] (3/4) About to create train dataset
2023-11-12 20:52:44,858 INFO [tts_datamodule.py:195] (3/4) Using DynamicBucketingSampler.
2023-11-12 20:52:46,996 INFO [tts_datamodule.py:210] (3/4) About to create train dataloader
2023-11-12 20:52:46,997 INFO [tts_datamodule.py:321] (3/4) About to get validation cuts
2023-11-12 20:52:46,998 INFO [tts_datamodule.py:233] (3/4) About to create dev dataset
2023-11-12 20:52:47,011 INFO [tts_datamodule.py:262] (3/4) About to create valid dataloader
2023-11-12 20:52:47,011 INFO [train.py:628] (3/4) Sanity check -- see if any of the batches in epoch 1 would cause OOM.
2023-11-12 20:52:56,724 INFO [train.py:674] (3/4) Maximum memory allocated so far is 14555MB
2023-11-12 20:53:00,892 INFO [train.py:674] (3/4) Maximum memory allocated so far is 14920MB
2023-11-12 20:53:05,400 INFO [train.py:674] (3/4) Maximum memory allocated so far is 15056MB
2023-11-12 20:53:10,064 INFO [train.py:674] (3/4) Maximum memory allocated so far is 15056MB
2023-11-12 20:53:18,679 INFO [train.py:674] (3/4) Maximum memory allocated so far is 25945MB
2023-11-12 20:53:27,171 INFO [train.py:674] (3/4) Maximum memory allocated so far is 25947MB
2023-11-12 20:53:27,201 INFO [train.py:811] (3/4) Start epoch 1
2023-11-12 20:53:41,221 INFO [train.py:467] (3/4) Epoch 1, batch 0, global_batch_idx: 0, batch size: 50, loss[discriminator_loss=5.934, discriminator_real_loss=5.934, discriminator_fake_loss=0.001226, generator_loss=959.2, generator_mel_loss=77.12, generator_kl_loss=875.1, generator_dur_loss=1.998, generator_adv_loss=4.762, generator_feat_match_loss=0.2483, over 50.00 samples.], tot_loss[discriminator_loss=5.934, discriminator_real_loss=5.934, discriminator_fake_loss=0.001226, generator_loss=959.2, generator_mel_loss=77.12, generator_kl_loss=875.1, generator_dur_loss=1.998, generator_adv_loss=4.762, generator_feat_match_loss=0.2483, over 50.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 2.0
2023-11-12 20:53:41,222 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 20:53:52,381 INFO [train.py:517] (3/4) Epoch 1, validation: discriminator_loss=4.839, discriminator_real_loss=4.761, discriminator_fake_loss=0.0778, generator_loss=489.8, generator_mel_loss=72.73, generator_kl_loss=409.9, generator_dur_loss=2.185, generator_adv_loss=4.762, generator_feat_match_loss=0.2584, over 100.00 samples.
2023-11-12 20:53:52,382 INFO [train.py:518] (3/4) Maximum memory allocated so far is 25947MB
2023-11-12 20:57:06,441 INFO [train.py:811] (3/4) Start epoch 2
2023-11-12 20:58:36,608 INFO [train.py:467] (3/4) Epoch 2, batch 13, global_batch_idx: 50, batch size: 60, loss[discriminator_loss=3.35, discriminator_real_loss=2.432, discriminator_fake_loss=0.9185, generator_loss=62.22, generator_mel_loss=42.51, generator_kl_loss=13.25, generator_dur_loss=1.832, generator_adv_loss=2.266, generator_feat_match_loss=2.355, over 60.00 samples.], tot_loss[discriminator_loss=2.697, discriminator_real_loss=1.618, discriminator_fake_loss=1.08, generator_loss=66.01, generator_mel_loss=43.52, generator_kl_loss=15.22, generator_dur_loss=1.811, generator_adv_loss=2.489, generator_feat_match_loss=2.978, over 974.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 2.0
2023-11-12 21:00:35,262 INFO [train.py:811] (3/4) Start epoch 3
2023-11-12 21:02:59,039 INFO [train.py:467] (3/4) Epoch 3, batch 26, global_batch_idx: 100, batch size: 52, loss[discriminator_loss=3.309, discriminator_real_loss=1.707, discriminator_fake_loss=1.602, generator_loss=52.47, generator_mel_loss=39.39, generator_kl_loss=6.434, generator_dur_loss=1.871, generator_adv_loss=1.996, generator_feat_match_loss=2.781, over 52.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.495, discriminator_fake_loss=1.121, generator_loss=55.34, generator_mel_loss=40, generator_kl_loss=7.575, generator_dur_loss=1.86, generator_adv_loss=2.45, generator_feat_match_loss=3.459, over 1972.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 4.0
2023-11-12 21:04:04,516 INFO [train.py:811] (3/4) Start epoch 4
2023-11-12 21:07:36,947 INFO [train.py:811] (3/4) Start epoch 5
2023-11-12 21:08:06,178 INFO [train.py:467] (3/4) Epoch 5, batch 2, global_batch_idx: 150, batch size: 61, loss[discriminator_loss=2.707, discriminator_real_loss=1.482, discriminator_fake_loss=1.225, generator_loss=47.64, generator_mel_loss=37.63, generator_kl_loss=4.183, generator_dur_loss=1.89, generator_adv_loss=1.699, generator_feat_match_loss=2.236, over 61.00 samples.], tot_loss[discriminator_loss=2.815, discriminator_real_loss=1.738, discriminator_fake_loss=1.078, generator_loss=48.09, generator_mel_loss=37.88, generator_kl_loss=4.289, generator_dur_loss=1.893, generator_adv_loss=1.893, generator_feat_match_loss=2.131, over 225.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 4.0
2023-11-12 21:11:08,832 INFO [train.py:811] (3/4) Start epoch 6
2023-11-12 21:12:49,840 INFO [train.py:467] (3/4) Epoch 6, batch 15, global_batch_idx: 200, batch size: 49, loss[discriminator_loss=2.65, discriminator_real_loss=1.414, discriminator_fake_loss=1.236, generator_loss=45.52, generator_mel_loss=36.59, generator_kl_loss=3.233, generator_dur_loss=1.897, generator_adv_loss=1.734, generator_feat_match_loss=2.062, over 49.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.392, discriminator_fake_loss=1.26, generator_loss=46.04, generator_mel_loss=36.71, generator_kl_loss=3.488, generator_dur_loss=1.891, generator_adv_loss=1.913, generator_feat_match_loss=2.039, over 1247.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 8.0
2023-11-12 21:12:49,841 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 21:13:01,089 INFO [train.py:517] (3/4) Epoch 6, validation: discriminator_loss=2.551, discriminator_real_loss=1.182, discriminator_fake_loss=1.369, generator_loss=46.58, generator_mel_loss=36.93, generator_kl_loss=3.442, generator_dur_loss=1.942, generator_adv_loss=1.789, generator_feat_match_loss=2.469, over 100.00 samples.
2023-11-12 21:13:01,090 INFO [train.py:518] (3/4) Maximum memory allocated so far is 26784MB
2023-11-12 21:14:53,740 INFO [train.py:811] (3/4) Start epoch 7
2023-11-12 21:17:39,673 INFO [train.py:467] (3/4) Epoch 7, batch 28, global_batch_idx: 250, batch size: 90, loss[discriminator_loss=2.727, discriminator_real_loss=1.229, discriminator_fake_loss=1.499, generator_loss=45.08, generator_mel_loss=36.2, generator_kl_loss=2.815, generator_dur_loss=1.903, generator_adv_loss=2.057, generator_feat_match_loss=2.111, over 90.00 samples.], tot_loss[discriminator_loss=2.661, discriminator_real_loss=1.415, discriminator_fake_loss=1.246, generator_loss=45.48, generator_mel_loss=36.32, generator_kl_loss=2.935, generator_dur_loss=1.901, generator_adv_loss=2.07, generator_feat_match_loss=2.253, over 2066.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 8.0
2023-11-12 21:18:23,418 INFO [train.py:811] (3/4) Start epoch 8
2023-11-12 21:21:51,414 INFO [train.py:811] (3/4) Start epoch 9
2023-11-12 21:22:30,278 INFO [train.py:467] (3/4) Epoch 9, batch 4, global_batch_idx: 300, batch size: 69, loss[discriminator_loss=2.518, discriminator_real_loss=1.677, discriminator_fake_loss=0.8408, generator_loss=43.35, generator_mel_loss=34.61, generator_kl_loss=2.472, generator_dur_loss=1.9, generator_adv_loss=1.838, generator_feat_match_loss=2.531, over 69.00 samples.], tot_loss[discriminator_loss=2.527, discriminator_real_loss=1.317, discriminator_fake_loss=1.21, generator_loss=43.99, generator_mel_loss=35.05, generator_kl_loss=2.574, generator_dur_loss=1.892, generator_adv_loss=2.143, generator_feat_match_loss=2.33, over 490.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 8.0
2023-11-12 21:25:27,664 INFO [train.py:811] (3/4) Start epoch 10
2023-11-12 21:27:16,823 INFO [train.py:467] (3/4) Epoch 10, batch 17, global_batch_idx: 350, batch size: 126, loss[discriminator_loss=2.383, discriminator_real_loss=1.133, discriminator_fake_loss=1.251, generator_loss=44.58, generator_mel_loss=34.83, generator_kl_loss=2.388, generator_dur_loss=1.889, generator_adv_loss=2.379, generator_feat_match_loss=3.098, over 126.00 samples.], tot_loss[discriminator_loss=2.499, discriminator_real_loss=1.35, discriminator_fake_loss=1.15, generator_loss=43.95, generator_mel_loss=34.73, generator_kl_loss=2.361, generator_dur_loss=1.906, generator_adv_loss=2.212, generator_feat_match_loss=2.737, over 1322.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 8.0
2023-11-12 21:28:59,577 INFO [train.py:811] (3/4) Start epoch 11
2023-11-12 21:31:55,296 INFO [train.py:467] (3/4) Epoch 11, batch 30, global_batch_idx: 400, batch size: 51, loss[discriminator_loss=2.715, discriminator_real_loss=1.607, discriminator_fake_loss=1.107, generator_loss=41.66, generator_mel_loss=33.97, generator_kl_loss=2.122, generator_dur_loss=1.917, generator_adv_loss=1.953, generator_feat_match_loss=1.695, over 51.00 samples.], tot_loss[discriminator_loss=2.728, discriminator_real_loss=1.445, discriminator_fake_loss=1.284, generator_loss=42.68, generator_mel_loss=34.39, generator_kl_loss=2.215, generator_dur_loss=1.907, generator_adv_loss=1.993, generator_feat_match_loss=2.169, over 2278.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:31:55,297 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 21:32:06,358 INFO [train.py:517] (3/4) Epoch 11, validation: discriminator_loss=2.597, discriminator_real_loss=1.44, discriminator_fake_loss=1.157, generator_loss=43.15, generator_mel_loss=34.91, generator_kl_loss=2.243, generator_dur_loss=1.931, generator_adv_loss=1.948, generator_feat_match_loss=2.121, over 100.00 samples.
2023-11-12 21:32:06,359 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 21:32:43,611 INFO [train.py:811] (3/4) Start epoch 12
2023-11-12 21:36:16,232 INFO [train.py:811] (3/4) Start epoch 13
2023-11-12 21:37:05,186 INFO [train.py:467] (3/4) Epoch 13, batch 6, global_batch_idx: 450, batch size: 73, loss[discriminator_loss=2.707, discriminator_real_loss=1.568, discriminator_fake_loss=1.14, generator_loss=40.68, generator_mel_loss=33.06, generator_kl_loss=2.11, generator_dur_loss=1.897, generator_adv_loss=1.937, generator_feat_match_loss=1.67, over 73.00 samples.], tot_loss[discriminator_loss=2.683, discriminator_real_loss=1.387, discriminator_fake_loss=1.297, generator_loss=42, generator_mel_loss=33.91, generator_kl_loss=2.075, generator_dur_loss=1.921, generator_adv_loss=2.041, generator_feat_match_loss=2.054, over 495.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:39:51,376 INFO [train.py:811] (3/4) Start epoch 14
2023-11-12 21:41:39,934 INFO [train.py:467] (3/4) Epoch 14, batch 19, global_batch_idx: 500, batch size: 54, loss[discriminator_loss=2.719, discriminator_real_loss=1.411, discriminator_fake_loss=1.309, generator_loss=39.09, generator_mel_loss=31.78, generator_kl_loss=1.98, generator_dur_loss=1.944, generator_adv_loss=1.821, generator_feat_match_loss=1.571, over 54.00 samples.], tot_loss[discriminator_loss=2.788, discriminator_real_loss=1.467, discriminator_fake_loss=1.321, generator_loss=40.6, generator_mel_loss=33.15, generator_kl_loss=2.038, generator_dur_loss=1.916, generator_adv_loss=1.824, generator_feat_match_loss=1.676, over 1367.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:43:16,386 INFO [train.py:811] (3/4) Start epoch 15
2023-11-12 21:46:18,659 INFO [train.py:467] (3/4) Epoch 15, batch 32, global_batch_idx: 550, batch size: 69, loss[discriminator_loss=2.699, discriminator_real_loss=1.543, discriminator_fake_loss=1.157, generator_loss=40.94, generator_mel_loss=33.07, generator_kl_loss=2.133, generator_dur_loss=1.911, generator_adv_loss=1.977, generator_feat_match_loss=1.85, over 69.00 samples.], tot_loss[discriminator_loss=2.71, discriminator_real_loss=1.396, discriminator_fake_loss=1.314, generator_loss=40.54, generator_mel_loss=32.72, generator_kl_loss=2.061, generator_dur_loss=1.913, generator_adv_loss=1.922, generator_feat_match_loss=1.925, over 2268.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:46:42,723 INFO [train.py:811] (3/4) Start epoch 16
2023-11-12 21:50:15,619 INFO [train.py:811] (3/4) Start epoch 17
2023-11-12 21:51:20,393 INFO [train.py:467] (3/4) Epoch 17, batch 8, global_batch_idx: 600, batch size: 59, loss[discriminator_loss=2.697, discriminator_real_loss=1.589, discriminator_fake_loss=1.108, generator_loss=40.98, generator_mel_loss=33.32, generator_kl_loss=2.041, generator_dur_loss=1.923, generator_adv_loss=1.56, generator_feat_match_loss=2.139, over 59.00 samples.], tot_loss[discriminator_loss=2.67, discriminator_real_loss=1.383, discriminator_fake_loss=1.287, generator_loss=40.18, generator_mel_loss=32.16, generator_kl_loss=2.165, generator_dur_loss=1.904, generator_adv_loss=1.945, generator_feat_match_loss=2.007, over 735.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:51:20,394 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 21:51:31,351 INFO [train.py:517] (3/4) Epoch 17, validation: discriminator_loss=2.587, discriminator_real_loss=1.127, discriminator_fake_loss=1.46, generator_loss=40.38, generator_mel_loss=32.44, generator_kl_loss=2.307, generator_dur_loss=1.91, generator_adv_loss=1.636, generator_feat_match_loss=2.091, over 100.00 samples.
2023-11-12 21:51:31,352 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 21:54:04,302 INFO [train.py:811] (3/4) Start epoch 18
2023-11-12 21:56:14,755 INFO [train.py:467] (3/4) Epoch 18, batch 21, global_batch_idx: 650, batch size: 85, loss[discriminator_loss=2.613, discriminator_real_loss=1.188, discriminator_fake_loss=1.427, generator_loss=40.12, generator_mel_loss=31.77, generator_kl_loss=2.131, generator_dur_loss=1.885, generator_adv_loss=2.102, generator_feat_match_loss=2.229, over 85.00 samples.], tot_loss[discriminator_loss=2.679, discriminator_real_loss=1.416, discriminator_fake_loss=1.264, generator_loss=39.83, generator_mel_loss=31.66, generator_kl_loss=2.137, generator_dur_loss=1.915, generator_adv_loss=1.994, generator_feat_match_loss=2.126, over 1532.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 21:57:41,257 INFO [train.py:811] (3/4) Start epoch 19
2023-11-12 22:00:56,430 INFO [train.py:467] (3/4) Epoch 19, batch 34, global_batch_idx: 700, batch size: 126, loss[discriminator_loss=2.684, discriminator_real_loss=1.318, discriminator_fake_loss=1.364, generator_loss=40.48, generator_mel_loss=31.89, generator_kl_loss=2.073, generator_dur_loss=1.882, generator_adv_loss=2.18, generator_feat_match_loss=2.455, over 126.00 samples.], tot_loss[discriminator_loss=2.611, discriminator_real_loss=1.399, discriminator_fake_loss=1.212, generator_loss=39.8, generator_mel_loss=31.29, generator_kl_loss=2.078, generator_dur_loss=1.918, generator_adv_loss=2.047, generator_feat_match_loss=2.466, over 2400.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 22:01:12,326 INFO [train.py:811] (3/4) Start epoch 20
2023-11-12 22:04:47,442 INFO [train.py:811] (3/4) Start epoch 21
2023-11-12 22:06:00,295 INFO [train.py:467] (3/4) Epoch 21, batch 10, global_batch_idx: 750, batch size: 51, loss[discriminator_loss=2.641, discriminator_real_loss=1.434, discriminator_fake_loss=1.207, generator_loss=38.76, generator_mel_loss=30.45, generator_kl_loss=2.084, generator_dur_loss=1.956, generator_adv_loss=2.184, generator_feat_match_loss=2.086, over 51.00 samples.], tot_loss[discriminator_loss=2.638, discriminator_real_loss=1.41, discriminator_fake_loss=1.228, generator_loss=39.4, generator_mel_loss=31.11, generator_kl_loss=2.053, generator_dur_loss=1.902, generator_adv_loss=2.008, generator_feat_match_loss=2.335, over 887.00 samples.], cur_lr_g: 2.00e-04, cur_lr_d: 2.00e-04, grad_scale: 16.0
2023-11-12 22:08:26,146 INFO [train.py:811] (3/4) Start epoch 22
2023-11-12 22:10:46,133 INFO [train.py:467] (3/4) Epoch 22, batch 23, global_batch_idx: 800, batch size: 71, loss[discriminator_loss=2.609, discriminator_real_loss=1.255, discriminator_fake_loss=1.355, generator_loss=37.09, generator_mel_loss=29.45, generator_kl_loss=2.034, generator_dur_loss=1.906, generator_adv_loss=1.553, generator_feat_match_loss=2.152, over 71.00 samples.], tot_loss[discriminator_loss=2.619, discriminator_real_loss=1.416, discriminator_fake_loss=1.203, generator_loss=39.08, generator_mel_loss=30.83, generator_kl_loss=2.06, generator_dur_loss=1.913, generator_adv_loss=1.952, generator_feat_match_loss=2.325, over 1758.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:10:46,134 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 22:10:56,994 INFO [train.py:517] (3/4) Epoch 22, validation: discriminator_loss=2.534, discriminator_real_loss=0.9679, discriminator_fake_loss=1.566, generator_loss=40.08, generator_mel_loss=31.58, generator_kl_loss=2.227, generator_dur_loss=1.91, generator_adv_loss=1.604, generator_feat_match_loss=2.761, over 100.00 samples.
2023-11-12 22:10:56,995 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 22:12:09,517 INFO [train.py:811] (3/4) Start epoch 23
2023-11-12 22:15:46,199 INFO [train.py:467] (3/4) Epoch 23, batch 36, global_batch_idx: 850, batch size: 65, loss[discriminator_loss=2.777, discriminator_real_loss=1.379, discriminator_fake_loss=1.398, generator_loss=38.68, generator_mel_loss=31.01, generator_kl_loss=2.013, generator_dur_loss=1.938, generator_adv_loss=1.596, generator_feat_match_loss=2.125, over 65.00 samples.], tot_loss[discriminator_loss=2.595, discriminator_real_loss=1.363, discriminator_fake_loss=1.232, generator_loss=39.04, generator_mel_loss=30.6, generator_kl_loss=2.025, generator_dur_loss=1.915, generator_adv_loss=2.015, generator_feat_match_loss=2.484, over 2847.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:15:46,852 INFO [train.py:811] (3/4) Start epoch 24
2023-11-12 22:19:14,651 INFO [train.py:811] (3/4) Start epoch 25
2023-11-12 22:20:35,887 INFO [train.py:467] (3/4) Epoch 25, batch 12, global_batch_idx: 900, batch size: 69, loss[discriminator_loss=2.494, discriminator_real_loss=1.319, discriminator_fake_loss=1.175, generator_loss=38.44, generator_mel_loss=29.7, generator_kl_loss=2.067, generator_dur_loss=1.919, generator_adv_loss=2.254, generator_feat_match_loss=2.506, over 69.00 samples.], tot_loss[discriminator_loss=2.612, discriminator_real_loss=1.346, discriminator_fake_loss=1.266, generator_loss=38.3, generator_mel_loss=30.3, generator_kl_loss=1.979, generator_dur_loss=1.916, generator_adv_loss=1.958, generator_feat_match_loss=2.147, over 899.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:22:43,236 INFO [train.py:811] (3/4) Start epoch 26
2023-11-12 22:25:14,077 INFO [train.py:467] (3/4) Epoch 26, batch 25, global_batch_idx: 950, batch size: 101, loss[discriminator_loss=2.701, discriminator_real_loss=1.397, discriminator_fake_loss=1.304, generator_loss=38.59, generator_mel_loss=30.32, generator_kl_loss=2.064, generator_dur_loss=1.897, generator_adv_loss=2.094, generator_feat_match_loss=2.221, over 101.00 samples.], tot_loss[discriminator_loss=2.662, discriminator_real_loss=1.407, discriminator_fake_loss=1.255, generator_loss=38.19, generator_mel_loss=30.09, generator_kl_loss=1.965, generator_dur_loss=1.92, generator_adv_loss=1.994, generator_feat_match_loss=2.218, over 1714.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:26:18,429 INFO [train.py:811] (3/4) Start epoch 27
2023-11-12 22:29:49,131 INFO [train.py:811] (3/4) Start epoch 28
2023-11-12 22:30:10,010 INFO [train.py:467] (3/4) Epoch 28, batch 1, global_batch_idx: 1000, batch size: 110, loss[discriminator_loss=2.605, discriminator_real_loss=1.256, discriminator_fake_loss=1.35, generator_loss=38.24, generator_mel_loss=30.06, generator_kl_loss=2.093, generator_dur_loss=1.912, generator_adv_loss=1.961, generator_feat_match_loss=2.213, over 110.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.316, discriminator_fake_loss=1.301, generator_loss=38.34, generator_mel_loss=30.31, generator_kl_loss=2.05, generator_dur_loss=1.913, generator_adv_loss=1.881, generator_feat_match_loss=2.18, over 189.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:30:10,011 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 22:30:21,705 INFO [train.py:517] (3/4) Epoch 28, validation: discriminator_loss=2.516, discriminator_real_loss=1.224, discriminator_fake_loss=1.292, generator_loss=38.54, generator_mel_loss=30.3, generator_kl_loss=2.126, generator_dur_loss=1.907, generator_adv_loss=1.876, generator_feat_match_loss=2.332, over 100.00 samples.
2023-11-12 22:30:21,706 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 22:33:37,399 INFO [train.py:811] (3/4) Start epoch 29
2023-11-12 22:35:09,488 INFO [train.py:467] (3/4) Epoch 29, batch 14, global_batch_idx: 1050, batch size: 101, loss[discriminator_loss=2.648, discriminator_real_loss=1.343, discriminator_fake_loss=1.306, generator_loss=37.4, generator_mel_loss=29.52, generator_kl_loss=2.036, generator_dur_loss=1.901, generator_adv_loss=1.841, generator_feat_match_loss=2.098, over 101.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.383, discriminator_fake_loss=1.274, generator_loss=38.23, generator_mel_loss=30.03, generator_kl_loss=1.984, generator_dur_loss=1.912, generator_adv_loss=1.993, generator_feat_match_loss=2.306, over 1180.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:37:10,340 INFO [train.py:811] (3/4) Start epoch 30
2023-11-12 22:39:46,607 INFO [train.py:467] (3/4) Epoch 30, batch 27, global_batch_idx: 1100, batch size: 101, loss[discriminator_loss=2.523, discriminator_real_loss=1.18, discriminator_fake_loss=1.344, generator_loss=37.66, generator_mel_loss=29.28, generator_kl_loss=2.069, generator_dur_loss=1.887, generator_adv_loss=1.999, generator_feat_match_loss=2.428, over 101.00 samples.], tot_loss[discriminator_loss=2.649, discriminator_real_loss=1.393, discriminator_fake_loss=1.256, generator_loss=37.63, generator_mel_loss=29.46, generator_kl_loss=1.995, generator_dur_loss=1.916, generator_adv_loss=1.98, generator_feat_match_loss=2.284, over 2218.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:40:37,750 INFO [train.py:811] (3/4) Start epoch 31
2023-11-12 22:44:07,621 INFO [train.py:811] (3/4) Start epoch 32
2023-11-12 22:44:43,091 INFO [train.py:467] (3/4) Epoch 32, batch 3, global_batch_idx: 1150, batch size: 95, loss[discriminator_loss=2.582, discriminator_real_loss=1.309, discriminator_fake_loss=1.272, generator_loss=38.4, generator_mel_loss=29.75, generator_kl_loss=1.979, generator_dur_loss=1.888, generator_adv_loss=2.229, generator_feat_match_loss=2.559, over 95.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.325, discriminator_fake_loss=1.298, generator_loss=37.86, generator_mel_loss=29.64, generator_kl_loss=2.01, generator_dur_loss=1.901, generator_adv_loss=1.929, generator_feat_match_loss=2.379, over 355.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:47:35,289 INFO [train.py:811] (3/4) Start epoch 33
2023-11-12 22:49:25,312 INFO [train.py:467] (3/4) Epoch 33, batch 16, global_batch_idx: 1200, batch size: 52, loss[discriminator_loss=2.6, discriminator_real_loss=1.333, discriminator_fake_loss=1.267, generator_loss=37.16, generator_mel_loss=28.87, generator_kl_loss=2.066, generator_dur_loss=1.943, generator_adv_loss=2.203, generator_feat_match_loss=2.078, over 52.00 samples.], tot_loss[discriminator_loss=2.806, discriminator_real_loss=1.517, discriminator_fake_loss=1.288, generator_loss=37.26, generator_mel_loss=29.25, generator_kl_loss=1.949, generator_dur_loss=1.925, generator_adv_loss=2.058, generator_feat_match_loss=2.072, over 1259.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:49:25,313 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 22:49:36,161 INFO [train.py:517] (3/4) Epoch 33, validation: discriminator_loss=2.588, discriminator_real_loss=1.459, discriminator_fake_loss=1.129, generator_loss=39.25, generator_mel_loss=30.77, generator_kl_loss=2.195, generator_dur_loss=1.925, generator_adv_loss=2.007, generator_feat_match_loss=2.35, over 100.00 samples.
2023-11-12 22:49:36,163 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 22:51:17,679 INFO [train.py:811] (3/4) Start epoch 34
2023-11-12 22:54:07,310 INFO [train.py:467] (3/4) Epoch 34, batch 29, global_batch_idx: 1250, batch size: 53, loss[discriminator_loss=2.73, discriminator_real_loss=1.479, discriminator_fake_loss=1.251, generator_loss=37.36, generator_mel_loss=29.43, generator_kl_loss=1.998, generator_dur_loss=1.947, generator_adv_loss=1.866, generator_feat_match_loss=2.115, over 53.00 samples.], tot_loss[discriminator_loss=2.687, discriminator_real_loss=1.394, discriminator_fake_loss=1.293, generator_loss=37.11, generator_mel_loss=29.13, generator_kl_loss=1.978, generator_dur_loss=1.919, generator_adv_loss=1.935, generator_feat_match_loss=2.148, over 2103.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 22:54:46,073 INFO [train.py:811] (3/4) Start epoch 35
2023-11-12 22:58:21,160 INFO [train.py:811] (3/4) Start epoch 36
2023-11-12 22:59:01,726 INFO [train.py:467] (3/4) Epoch 36, batch 5, global_batch_idx: 1300, batch size: 55, loss[discriminator_loss=2.449, discriminator_real_loss=1.281, discriminator_fake_loss=1.169, generator_loss=38.33, generator_mel_loss=29.45, generator_kl_loss=1.812, generator_dur_loss=1.968, generator_adv_loss=2.057, generator_feat_match_loss=3.047, over 55.00 samples.], tot_loss[discriminator_loss=2.629, discriminator_real_loss=1.362, discriminator_fake_loss=1.267, generator_loss=37.13, generator_mel_loss=28.85, generator_kl_loss=1.953, generator_dur_loss=1.922, generator_adv_loss=1.984, generator_feat_match_loss=2.423, over 437.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:01:54,278 INFO [train.py:811] (3/4) Start epoch 37
2023-11-12 23:03:44,576 INFO [train.py:467] (3/4) Epoch 37, batch 18, global_batch_idx: 1350, batch size: 81, loss[discriminator_loss=2.559, discriminator_real_loss=1.106, discriminator_fake_loss=1.451, generator_loss=37.63, generator_mel_loss=28.96, generator_kl_loss=1.986, generator_dur_loss=1.924, generator_adv_loss=2.148, generator_feat_match_loss=2.613, over 81.00 samples.], tot_loss[discriminator_loss=2.661, discriminator_real_loss=1.388, discriminator_fake_loss=1.274, generator_loss=37.1, generator_mel_loss=28.84, generator_kl_loss=1.987, generator_dur_loss=1.934, generator_adv_loss=2.028, generator_feat_match_loss=2.315, over 1307.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:05:25,668 INFO [train.py:811] (3/4) Start epoch 38
2023-11-12 23:08:29,498 INFO [train.py:467] (3/4) Epoch 38, batch 31, global_batch_idx: 1400, batch size: 61, loss[discriminator_loss=2.641, discriminator_real_loss=1.559, discriminator_fake_loss=1.082, generator_loss=37.23, generator_mel_loss=28.87, generator_kl_loss=1.852, generator_dur_loss=1.934, generator_adv_loss=2.105, generator_feat_match_loss=2.461, over 61.00 samples.], tot_loss[discriminator_loss=2.682, discriminator_real_loss=1.391, discriminator_fake_loss=1.291, generator_loss=36.73, generator_mel_loss=28.62, generator_kl_loss=1.967, generator_dur_loss=1.93, generator_adv_loss=1.981, generator_feat_match_loss=2.228, over 2409.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:08:29,500 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 23:08:40,525 INFO [train.py:517] (3/4) Epoch 38, validation: discriminator_loss=2.645, discriminator_real_loss=1.425, discriminator_fake_loss=1.22, generator_loss=39.17, generator_mel_loss=30.61, generator_kl_loss=2.023, generator_dur_loss=1.978, generator_adv_loss=1.968, generator_feat_match_loss=2.591, over 100.00 samples.
2023-11-12 23:08:40,526 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 23:09:05,434 INFO [train.py:811] (3/4) Start epoch 39
2023-11-12 23:12:38,060 INFO [train.py:811] (3/4) Start epoch 40
2023-11-12 23:13:39,067 INFO [train.py:467] (3/4) Epoch 40, batch 7, global_batch_idx: 1450, batch size: 53, loss[discriminator_loss=2.555, discriminator_real_loss=1.272, discriminator_fake_loss=1.283, generator_loss=37.59, generator_mel_loss=29.28, generator_kl_loss=1.914, generator_dur_loss=1.951, generator_adv_loss=1.955, generator_feat_match_loss=2.498, over 53.00 samples.], tot_loss[discriminator_loss=2.658, discriminator_real_loss=1.406, discriminator_fake_loss=1.252, generator_loss=36.64, generator_mel_loss=28.6, generator_kl_loss=1.911, generator_dur_loss=1.927, generator_adv_loss=1.933, generator_feat_match_loss=2.279, over 621.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:16:10,952 INFO [train.py:811] (3/4) Start epoch 41
2023-11-12 23:18:22,781 INFO [train.py:467] (3/4) Epoch 41, batch 20, global_batch_idx: 1500, batch size: 65, loss[discriminator_loss=2.816, discriminator_real_loss=1.367, discriminator_fake_loss=1.449, generator_loss=36.85, generator_mel_loss=28.57, generator_kl_loss=1.875, generator_dur_loss=1.94, generator_adv_loss=2.27, generator_feat_match_loss=2.197, over 65.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.372, discriminator_fake_loss=1.28, generator_loss=36.79, generator_mel_loss=28.61, generator_kl_loss=1.897, generator_dur_loss=1.931, generator_adv_loss=2.013, generator_feat_match_loss=2.34, over 1560.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:19:52,323 INFO [train.py:811] (3/4) Start epoch 42
2023-11-12 23:23:08,822 INFO [train.py:467] (3/4) Epoch 42, batch 33, global_batch_idx: 1550, batch size: 58, loss[discriminator_loss=2.695, discriminator_real_loss=1.312, discriminator_fake_loss=1.384, generator_loss=35.33, generator_mel_loss=27.22, generator_kl_loss=1.921, generator_dur_loss=1.934, generator_adv_loss=2.121, generator_feat_match_loss=2.133, over 58.00 samples.], tot_loss[discriminator_loss=2.649, discriminator_real_loss=1.394, discriminator_fake_loss=1.255, generator_loss=36.22, generator_mel_loss=28.09, generator_kl_loss=1.921, generator_dur_loss=1.931, generator_adv_loss=1.951, generator_feat_match_loss=2.325, over 2492.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:23:26,158 INFO [train.py:811] (3/4) Start epoch 43
2023-11-12 23:26:52,114 INFO [train.py:811] (3/4) Start epoch 44
2023-11-12 23:27:57,931 INFO [train.py:467] (3/4) Epoch 44, batch 9, global_batch_idx: 1600, batch size: 64, loss[discriminator_loss=2.617, discriminator_real_loss=1.478, discriminator_fake_loss=1.141, generator_loss=36.13, generator_mel_loss=27.67, generator_kl_loss=1.865, generator_dur_loss=1.93, generator_adv_loss=2.287, generator_feat_match_loss=2.379, over 64.00 samples.], tot_loss[discriminator_loss=2.65, discriminator_real_loss=1.407, discriminator_fake_loss=1.242, generator_loss=35.83, generator_mel_loss=27.68, generator_kl_loss=1.958, generator_dur_loss=1.939, generator_adv_loss=1.978, generator_feat_match_loss=2.278, over 755.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:27:57,933 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 23:28:08,905 INFO [train.py:517] (3/4) Epoch 44, validation: discriminator_loss=2.572, discriminator_real_loss=1.465, discriminator_fake_loss=1.107, generator_loss=37.81, generator_mel_loss=29.18, generator_kl_loss=1.971, generator_dur_loss=1.932, generator_adv_loss=2.111, generator_feat_match_loss=2.616, over 100.00 samples.
2023-11-12 23:28:08,908 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 23:30:29,633 INFO [train.py:811] (3/4) Start epoch 45
2023-11-12 23:32:41,413 INFO [train.py:467] (3/4) Epoch 45, batch 22, global_batch_idx: 1650, batch size: 79, loss[discriminator_loss=2.697, discriminator_real_loss=1.381, discriminator_fake_loss=1.316, generator_loss=34.57, generator_mel_loss=26.88, generator_kl_loss=1.869, generator_dur_loss=1.939, generator_adv_loss=1.967, generator_feat_match_loss=1.918, over 79.00 samples.], tot_loss[discriminator_loss=2.707, discriminator_real_loss=1.418, discriminator_fake_loss=1.289, generator_loss=35.91, generator_mel_loss=27.88, generator_kl_loss=1.958, generator_dur_loss=1.942, generator_adv_loss=1.936, generator_feat_match_loss=2.199, over 1665.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:33:56,775 INFO [train.py:811] (3/4) Start epoch 46
2023-11-12 23:37:24,911 INFO [train.py:467] (3/4) Epoch 46, batch 35, global_batch_idx: 1700, batch size: 52, loss[discriminator_loss=2.697, discriminator_real_loss=1.179, discriminator_fake_loss=1.519, generator_loss=35.04, generator_mel_loss=26.74, generator_kl_loss=1.961, generator_dur_loss=1.946, generator_adv_loss=2.07, generator_feat_match_loss=2.322, over 52.00 samples.], tot_loss[discriminator_loss=2.705, discriminator_real_loss=1.411, discriminator_fake_loss=1.294, generator_loss=35.43, generator_mel_loss=27.44, generator_kl_loss=1.967, generator_dur_loss=1.934, generator_adv_loss=1.922, generator_feat_match_loss=2.166, over 2702.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:37:29,967 INFO [train.py:811] (3/4) Start epoch 47
2023-11-12 23:41:00,957 INFO [train.py:811] (3/4) Start epoch 48
2023-11-12 23:42:16,591 INFO [train.py:467] (3/4) Epoch 48, batch 11, global_batch_idx: 1750, batch size: 60, loss[discriminator_loss=2.604, discriminator_real_loss=1.209, discriminator_fake_loss=1.395, generator_loss=35.82, generator_mel_loss=27.39, generator_kl_loss=2.026, generator_dur_loss=1.965, generator_adv_loss=1.991, generator_feat_match_loss=2.451, over 60.00 samples.], tot_loss[discriminator_loss=2.68, discriminator_real_loss=1.389, discriminator_fake_loss=1.291, generator_loss=34.89, generator_mel_loss=26.95, generator_kl_loss=1.986, generator_dur_loss=1.949, generator_adv_loss=1.892, generator_feat_match_loss=2.112, over 790.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:44:30,583 INFO [train.py:811] (3/4) Start epoch 49
2023-11-12 23:46:57,481 INFO [train.py:467] (3/4) Epoch 49, batch 24, global_batch_idx: 1800, batch size: 50, loss[discriminator_loss=2.732, discriminator_real_loss=1.527, discriminator_fake_loss=1.205, generator_loss=34.13, generator_mel_loss=26.21, generator_kl_loss=2.024, generator_dur_loss=1.96, generator_adv_loss=1.829, generator_feat_match_loss=2.105, over 50.00 samples.], tot_loss[discriminator_loss=2.707, discriminator_real_loss=1.396, discriminator_fake_loss=1.311, generator_loss=34.93, generator_mel_loss=26.96, generator_kl_loss=1.986, generator_dur_loss=1.941, generator_adv_loss=1.943, generator_feat_match_loss=2.108, over 1807.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:46:57,482 INFO [train.py:508] (3/4) Computing validation loss
2023-11-12 23:47:08,424 INFO [train.py:517] (3/4) Epoch 49, validation: discriminator_loss=2.643, discriminator_real_loss=1.239, discriminator_fake_loss=1.404, generator_loss=36.36, generator_mel_loss=28.35, generator_kl_loss=2.026, generator_dur_loss=1.943, generator_adv_loss=1.79, generator_feat_match_loss=2.245, over 100.00 samples.
2023-11-12 23:47:08,425 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-12 23:48:15,320 INFO [train.py:811] (3/4) Start epoch 50
2023-11-12 23:51:47,389 INFO [train.py:811] (3/4) Start epoch 51
2023-11-12 23:52:02,386 INFO [train.py:467] (3/4) Epoch 51, batch 0, global_batch_idx: 1850, batch size: 71, loss[discriminator_loss=2.723, discriminator_real_loss=1.334, discriminator_fake_loss=1.39, generator_loss=34.16, generator_mel_loss=26.23, generator_kl_loss=1.922, generator_dur_loss=1.918, generator_adv_loss=2.166, generator_feat_match_loss=1.928, over 71.00 samples.], tot_loss[discriminator_loss=2.723, discriminator_real_loss=1.334, discriminator_fake_loss=1.39, generator_loss=34.16, generator_mel_loss=26.23, generator_kl_loss=1.922, generator_dur_loss=1.918, generator_adv_loss=2.166, generator_feat_match_loss=1.928, over 71.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:55:18,280 INFO [train.py:811] (3/4) Start epoch 52
2023-11-12 23:56:40,076 INFO [train.py:467] (3/4) Epoch 52, batch 13, global_batch_idx: 1900, batch size: 50, loss[discriminator_loss=2.676, discriminator_real_loss=1.462, discriminator_fake_loss=1.214, generator_loss=34.81, generator_mel_loss=26.98, generator_kl_loss=2.037, generator_dur_loss=1.987, generator_adv_loss=1.721, generator_feat_match_loss=2.078, over 50.00 samples.], tot_loss[discriminator_loss=2.722, discriminator_real_loss=1.394, discriminator_fake_loss=1.328, generator_loss=34.7, generator_mel_loss=26.78, generator_kl_loss=2.046, generator_dur_loss=1.941, generator_adv_loss=1.872, generator_feat_match_loss=2.066, over 983.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-12 23:58:45,786 INFO [train.py:811] (3/4) Start epoch 53
2023-11-13 00:01:25,377 INFO [train.py:467] (3/4) Epoch 53, batch 26, global_batch_idx: 1950, batch size: 56, loss[discriminator_loss=2.781, discriminator_real_loss=1.526, discriminator_fake_loss=1.256, generator_loss=33.48, generator_mel_loss=26, generator_kl_loss=1.98, generator_dur_loss=1.951, generator_adv_loss=1.738, generator_feat_match_loss=1.817, over 56.00 samples.], tot_loss[discriminator_loss=2.768, discriminator_real_loss=1.439, discriminator_fake_loss=1.328, generator_loss=33.98, generator_mel_loss=26.11, generator_kl_loss=2.028, generator_dur_loss=1.94, generator_adv_loss=1.902, generator_feat_match_loss=1.999, over 1859.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 32.0
2023-11-13 00:02:18,335 INFO [train.py:811] (3/4) Start epoch 54
2023-11-13 00:05:44,626 INFO [train.py:811] (3/4) Start epoch 55
2023-11-13 00:06:08,607 INFO [train.py:467] (3/4) Epoch 55, batch 2, global_batch_idx: 2000, batch size: 73, loss[discriminator_loss=2.693, discriminator_real_loss=1.434, discriminator_fake_loss=1.26, generator_loss=33.76, generator_mel_loss=25.92, generator_kl_loss=1.907, generator_dur_loss=1.924, generator_adv_loss=1.975, generator_feat_match_loss=2.037, over 73.00 samples.], tot_loss[discriminator_loss=2.711, discriminator_real_loss=1.346, discriminator_fake_loss=1.365, generator_loss=33.5, generator_mel_loss=25.67, generator_kl_loss=1.968, generator_dur_loss=1.925, generator_adv_loss=1.943, generator_feat_match_loss=1.995, over 198.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:06:08,609 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 00:06:20,173 INFO [train.py:517] (3/4) Epoch 55, validation: discriminator_loss=2.637, discriminator_real_loss=1.39, discriminator_fake_loss=1.247, generator_loss=34.51, generator_mel_loss=26.79, generator_kl_loss=1.921, generator_dur_loss=1.915, generator_adv_loss=1.903, generator_feat_match_loss=1.974, over 100.00 samples.
2023-11-13 00:06:20,174 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 00:09:25,022 INFO [train.py:811] (3/4) Start epoch 56
2023-11-13 00:11:05,082 INFO [train.py:467] (3/4) Epoch 56, batch 15, global_batch_idx: 2050, batch size: 73, loss[discriminator_loss=2.852, discriminator_real_loss=1.447, discriminator_fake_loss=1.403, generator_loss=33.66, generator_mel_loss=26.06, generator_kl_loss=1.995, generator_dur_loss=1.898, generator_adv_loss=1.76, generator_feat_match_loss=1.941, over 73.00 samples.], tot_loss[discriminator_loss=2.762, discriminator_real_loss=1.442, discriminator_fake_loss=1.32, generator_loss=33.67, generator_mel_loss=25.89, generator_kl_loss=1.989, generator_dur_loss=1.92, generator_adv_loss=1.907, generator_feat_match_loss=1.965, over 1048.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:12:58,307 INFO [train.py:811] (3/4) Start epoch 57
2023-11-13 00:15:54,470 INFO [train.py:467] (3/4) Epoch 57, batch 28, global_batch_idx: 2100, batch size: 63, loss[discriminator_loss=2.734, discriminator_real_loss=1.329, discriminator_fake_loss=1.404, generator_loss=33.01, generator_mel_loss=25.15, generator_kl_loss=2.071, generator_dur_loss=1.926, generator_adv_loss=1.995, generator_feat_match_loss=1.86, over 63.00 samples.], tot_loss[discriminator_loss=2.723, discriminator_real_loss=1.399, discriminator_fake_loss=1.325, generator_loss=33.59, generator_mel_loss=25.73, generator_kl_loss=1.988, generator_dur_loss=1.909, generator_adv_loss=1.875, generator_feat_match_loss=2.083, over 2007.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:16:33,740 INFO [train.py:811] (3/4) Start epoch 58
2023-11-13 00:20:09,538 INFO [train.py:811] (3/4) Start epoch 59
2023-11-13 00:20:44,073 INFO [train.py:467] (3/4) Epoch 59, batch 4, global_batch_idx: 2150, batch size: 76, loss[discriminator_loss=2.764, discriminator_real_loss=1.078, discriminator_fake_loss=1.686, generator_loss=33.65, generator_mel_loss=25.74, generator_kl_loss=1.996, generator_dur_loss=1.901, generator_adv_loss=1.963, generator_feat_match_loss=2.051, over 76.00 samples.], tot_loss[discriminator_loss=2.754, discriminator_real_loss=1.385, discriminator_fake_loss=1.369, generator_loss=33.09, generator_mel_loss=25.37, generator_kl_loss=1.945, generator_dur_loss=1.899, generator_adv_loss=1.895, generator_feat_match_loss=1.981, over 353.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:23:42,604 INFO [train.py:811] (3/4) Start epoch 60
2023-11-13 00:25:37,852 INFO [train.py:467] (3/4) Epoch 60, batch 17, global_batch_idx: 2200, batch size: 60, loss[discriminator_loss=2.717, discriminator_real_loss=1.43, discriminator_fake_loss=1.287, generator_loss=32.74, generator_mel_loss=24.99, generator_kl_loss=1.988, generator_dur_loss=1.9, generator_adv_loss=1.868, generator_feat_match_loss=1.988, over 60.00 samples.], tot_loss[discriminator_loss=2.713, discriminator_real_loss=1.39, discriminator_fake_loss=1.323, generator_loss=33.19, generator_mel_loss=25.32, generator_kl_loss=1.979, generator_dur_loss=1.885, generator_adv_loss=1.914, generator_feat_match_loss=2.087, over 1203.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:25:37,854 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 00:25:49,056 INFO [train.py:517] (3/4) Epoch 60, validation: discriminator_loss=2.641, discriminator_real_loss=1.307, discriminator_fake_loss=1.334, generator_loss=34.97, generator_mel_loss=27.14, generator_kl_loss=1.974, generator_dur_loss=1.873, generator_adv_loss=1.789, generator_feat_match_loss=2.193, over 100.00 samples.
2023-11-13 00:25:49,058 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 00:27:28,428 INFO [train.py:811] (3/4) Start epoch 61
2023-11-13 00:30:39,990 INFO [train.py:467] (3/4) Epoch 61, batch 30, global_batch_idx: 2250, batch size: 52, loss[discriminator_loss=2.734, discriminator_real_loss=1.414, discriminator_fake_loss=1.32, generator_loss=34.39, generator_mel_loss=26.47, generator_kl_loss=2.028, generator_dur_loss=1.897, generator_adv_loss=1.832, generator_feat_match_loss=2.164, over 52.00 samples.], tot_loss[discriminator_loss=2.718, discriminator_real_loss=1.398, discriminator_fake_loss=1.321, generator_loss=33.14, generator_mel_loss=25.3, generator_kl_loss=1.962, generator_dur_loss=1.87, generator_adv_loss=1.918, generator_feat_match_loss=2.087, over 2223.00 samples.], cur_lr_g: 1.99e-04, cur_lr_d: 1.99e-04, grad_scale: 64.0
2023-11-13 00:31:07,286 INFO [train.py:811] (3/4) Start epoch 62
2023-11-13 00:34:46,082 INFO [train.py:811] (3/4) Start epoch 63
2023-11-13 00:35:39,834 INFO [train.py:467] (3/4) Epoch 63, batch 6, global_batch_idx: 2300, batch size: 53, loss[discriminator_loss=2.717, discriminator_real_loss=1.33, discriminator_fake_loss=1.387, generator_loss=32.38, generator_mel_loss=24.77, generator_kl_loss=1.963, generator_dur_loss=1.911, generator_adv_loss=1.695, generator_feat_match_loss=2.045, over 53.00 samples.], tot_loss[discriminator_loss=2.767, discriminator_real_loss=1.401, discriminator_fake_loss=1.366, generator_loss=33.13, generator_mel_loss=25.33, generator_kl_loss=1.951, generator_dur_loss=1.879, generator_adv_loss=1.902, generator_feat_match_loss=2.069, over 538.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:38:19,864 INFO [train.py:811] (3/4) Start epoch 64
2023-11-13 00:40:24,050 INFO [train.py:467] (3/4) Epoch 64, batch 19, global_batch_idx: 2350, batch size: 73, loss[discriminator_loss=2.783, discriminator_real_loss=1.448, discriminator_fake_loss=1.335, generator_loss=32.64, generator_mel_loss=25.26, generator_kl_loss=1.894, generator_dur_loss=1.852, generator_adv_loss=1.78, generator_feat_match_loss=1.857, over 73.00 samples.], tot_loss[discriminator_loss=2.743, discriminator_real_loss=1.408, discriminator_fake_loss=1.335, generator_loss=32.79, generator_mel_loss=25.22, generator_kl_loss=1.97, generator_dur_loss=1.862, generator_adv_loss=1.816, generator_feat_match_loss=1.926, over 1592.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:41:52,809 INFO [train.py:811] (3/4) Start epoch 65
2023-11-13 00:45:07,195 INFO [train.py:467] (3/4) Epoch 65, batch 32, global_batch_idx: 2400, batch size: 54, loss[discriminator_loss=2.742, discriminator_real_loss=1.385, discriminator_fake_loss=1.357, generator_loss=31.86, generator_mel_loss=24.09, generator_kl_loss=2.014, generator_dur_loss=1.894, generator_adv_loss=1.889, generator_feat_match_loss=1.974, over 54.00 samples.], tot_loss[discriminator_loss=2.75, discriminator_real_loss=1.403, discriminator_fake_loss=1.347, generator_loss=32.57, generator_mel_loss=24.92, generator_kl_loss=1.963, generator_dur_loss=1.865, generator_adv_loss=1.881, generator_feat_match_loss=1.946, over 2232.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:45:07,197 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 00:45:18,048 INFO [train.py:517] (3/4) Epoch 65, validation: discriminator_loss=2.64, discriminator_real_loss=1.369, discriminator_fake_loss=1.272, generator_loss=33.82, generator_mel_loss=25.91, generator_kl_loss=2.019, generator_dur_loss=1.845, generator_adv_loss=1.855, generator_feat_match_loss=2.192, over 100.00 samples.
2023-11-13 00:45:18,049 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 00:45:36,346 INFO [train.py:811] (3/4) Start epoch 66
2023-11-13 00:49:07,817 INFO [train.py:811] (3/4) Start epoch 67
2023-11-13 00:50:03,771 INFO [train.py:467] (3/4) Epoch 67, batch 8, global_batch_idx: 2450, batch size: 63, loss[discriminator_loss=2.67, discriminator_real_loss=1.371, discriminator_fake_loss=1.299, generator_loss=32.17, generator_mel_loss=24.59, generator_kl_loss=1.986, generator_dur_loss=1.863, generator_adv_loss=1.789, generator_feat_match_loss=1.948, over 63.00 samples.], tot_loss[discriminator_loss=2.721, discriminator_real_loss=1.405, discriminator_fake_loss=1.317, generator_loss=32.42, generator_mel_loss=24.74, generator_kl_loss=1.964, generator_dur_loss=1.86, generator_adv_loss=1.849, generator_feat_match_loss=2.011, over 540.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:52:45,270 INFO [train.py:811] (3/4) Start epoch 68
2023-11-13 00:55:03,175 INFO [train.py:467] (3/4) Epoch 68, batch 21, global_batch_idx: 2500, batch size: 85, loss[discriminator_loss=2.777, discriminator_real_loss=1.506, discriminator_fake_loss=1.271, generator_loss=31.43, generator_mel_loss=24.23, generator_kl_loss=1.981, generator_dur_loss=1.826, generator_adv_loss=1.561, generator_feat_match_loss=1.836, over 85.00 samples.], tot_loss[discriminator_loss=2.747, discriminator_real_loss=1.444, discriminator_fake_loss=1.303, generator_loss=32.48, generator_mel_loss=24.78, generator_kl_loss=2.005, generator_dur_loss=1.853, generator_adv_loss=1.843, generator_feat_match_loss=2.004, over 1613.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:56:22,254 INFO [train.py:811] (3/4) Start epoch 69
2023-11-13 00:59:44,467 INFO [train.py:467] (3/4) Epoch 69, batch 34, global_batch_idx: 2550, batch size: 50, loss[discriminator_loss=2.742, discriminator_real_loss=1.288, discriminator_fake_loss=1.454, generator_loss=32.71, generator_mel_loss=24.94, generator_kl_loss=1.925, generator_dur_loss=1.851, generator_adv_loss=1.91, generator_feat_match_loss=2.078, over 50.00 samples.], tot_loss[discriminator_loss=2.743, discriminator_real_loss=1.389, discriminator_fake_loss=1.354, generator_loss=32.61, generator_mel_loss=24.85, generator_kl_loss=1.991, generator_dur_loss=1.842, generator_adv_loss=1.886, generator_feat_match_loss=2.048, over 2874.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 00:59:55,263 INFO [train.py:811] (3/4) Start epoch 70
2023-11-13 01:03:19,349 INFO [train.py:811] (3/4) Start epoch 71
2023-11-13 01:04:28,846 INFO [train.py:467] (3/4) Epoch 71, batch 10, global_batch_idx: 2600, batch size: 53, loss[discriminator_loss=2.691, discriminator_real_loss=1.359, discriminator_fake_loss=1.333, generator_loss=31.32, generator_mel_loss=23.54, generator_kl_loss=1.985, generator_dur_loss=1.844, generator_adv_loss=1.881, generator_feat_match_loss=2.074, over 53.00 samples.], tot_loss[discriminator_loss=2.718, discriminator_real_loss=1.373, discriminator_fake_loss=1.345, generator_loss=32.45, generator_mel_loss=24.66, generator_kl_loss=1.995, generator_dur_loss=1.844, generator_adv_loss=1.871, generator_feat_match_loss=2.077, over 837.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:04:28,847 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 01:04:39,259 INFO [train.py:517] (3/4) Epoch 71, validation: discriminator_loss=2.675, discriminator_real_loss=1.302, discriminator_fake_loss=1.373, generator_loss=33.04, generator_mel_loss=25.21, generator_kl_loss=2.058, generator_dur_loss=1.819, generator_adv_loss=1.796, generator_feat_match_loss=2.157, over 100.00 samples.
2023-11-13 01:04:39,261 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 01:06:57,306 INFO [train.py:811] (3/4) Start epoch 72
2023-11-13 01:09:24,988 INFO [train.py:467] (3/4) Epoch 72, batch 23, global_batch_idx: 2650, batch size: 110, loss[discriminator_loss=2.762, discriminator_real_loss=1.659, discriminator_fake_loss=1.104, generator_loss=31.52, generator_mel_loss=23.9, generator_kl_loss=2.013, generator_dur_loss=1.836, generator_adv_loss=1.738, generator_feat_match_loss=2.029, over 110.00 samples.], tot_loss[discriminator_loss=2.739, discriminator_real_loss=1.424, discriminator_fake_loss=1.315, generator_loss=32.08, generator_mel_loss=24.38, generator_kl_loss=1.97, generator_dur_loss=1.836, generator_adv_loss=1.873, generator_feat_match_loss=2.021, over 1762.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:10:28,851 INFO [train.py:811] (3/4) Start epoch 73
2023-11-13 01:14:02,898 INFO [train.py:467] (3/4) Epoch 73, batch 36, global_batch_idx: 2700, batch size: 61, loss[discriminator_loss=2.854, discriminator_real_loss=1.24, discriminator_fake_loss=1.613, generator_loss=31.86, generator_mel_loss=24.26, generator_kl_loss=1.925, generator_dur_loss=1.845, generator_adv_loss=1.906, generator_feat_match_loss=1.92, over 61.00 samples.], tot_loss[discriminator_loss=2.73, discriminator_real_loss=1.398, discriminator_fake_loss=1.332, generator_loss=32.2, generator_mel_loss=24.42, generator_kl_loss=1.971, generator_dur_loss=1.833, generator_adv_loss=1.891, generator_feat_match_loss=2.082, over 2686.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:14:03,592 INFO [train.py:811] (3/4) Start epoch 74
2023-11-13 01:17:38,642 INFO [train.py:811] (3/4) Start epoch 75
2023-11-13 01:18:57,346 INFO [train.py:467] (3/4) Epoch 75, batch 12, global_batch_idx: 2750, batch size: 60, loss[discriminator_loss=2.82, discriminator_real_loss=1.233, discriminator_fake_loss=1.588, generator_loss=32.38, generator_mel_loss=24.13, generator_kl_loss=1.925, generator_dur_loss=1.851, generator_adv_loss=2.283, generator_feat_match_loss=2.182, over 60.00 samples.], tot_loss[discriminator_loss=2.73, discriminator_real_loss=1.39, discriminator_fake_loss=1.34, generator_loss=32.22, generator_mel_loss=24.37, generator_kl_loss=2.006, generator_dur_loss=1.833, generator_adv_loss=1.932, generator_feat_match_loss=2.079, over 936.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:21:05,249 INFO [train.py:811] (3/4) Start epoch 76
2023-11-13 01:23:36,736 INFO [train.py:467] (3/4) Epoch 76, batch 25, global_batch_idx: 2800, batch size: 63, loss[discriminator_loss=2.777, discriminator_real_loss=1.473, discriminator_fake_loss=1.305, generator_loss=31.99, generator_mel_loss=24.23, generator_kl_loss=1.895, generator_dur_loss=1.862, generator_adv_loss=1.997, generator_feat_match_loss=2.012, over 63.00 samples.], tot_loss[discriminator_loss=2.738, discriminator_real_loss=1.396, discriminator_fake_loss=1.341, generator_loss=31.94, generator_mel_loss=24.16, generator_kl_loss=1.971, generator_dur_loss=1.825, generator_adv_loss=1.902, generator_feat_match_loss=2.076, over 1754.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:23:36,738 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 01:23:47,554 INFO [train.py:517] (3/4) Epoch 76, validation: discriminator_loss=2.666, discriminator_real_loss=1.444, discriminator_fake_loss=1.222, generator_loss=33.52, generator_mel_loss=25.64, generator_kl_loss=2.048, generator_dur_loss=1.806, generator_adv_loss=1.952, generator_feat_match_loss=2.076, over 100.00 samples.
2023-11-13 01:23:47,555 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 01:24:48,420 INFO [train.py:811] (3/4) Start epoch 77
2023-11-13 01:28:18,986 INFO [train.py:811] (3/4) Start epoch 78
2023-11-13 01:28:39,661 INFO [train.py:467] (3/4) Epoch 78, batch 1, global_batch_idx: 2850, batch size: 101, loss[discriminator_loss=2.824, discriminator_real_loss=1.669, discriminator_fake_loss=1.154, generator_loss=31.95, generator_mel_loss=24.46, generator_kl_loss=1.952, generator_dur_loss=1.808, generator_adv_loss=1.745, generator_feat_match_loss=1.99, over 101.00 samples.], tot_loss[discriminator_loss=2.794, discriminator_real_loss=1.412, discriminator_fake_loss=1.381, generator_loss=32.24, generator_mel_loss=24.43, generator_kl_loss=1.995, generator_dur_loss=1.808, generator_adv_loss=1.95, generator_feat_match_loss=2.05, over 186.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:31:48,768 INFO [train.py:811] (3/4) Start epoch 79
2023-11-13 01:33:16,516 INFO [train.py:467] (3/4) Epoch 79, batch 14, global_batch_idx: 2900, batch size: 79, loss[discriminator_loss=2.707, discriminator_real_loss=1.286, discriminator_fake_loss=1.421, generator_loss=31.52, generator_mel_loss=23.71, generator_kl_loss=1.932, generator_dur_loss=1.817, generator_adv_loss=1.993, generator_feat_match_loss=2.066, over 79.00 samples.], tot_loss[discriminator_loss=2.759, discriminator_real_loss=1.416, discriminator_fake_loss=1.344, generator_loss=31.91, generator_mel_loss=24.12, generator_kl_loss=1.995, generator_dur_loss=1.817, generator_adv_loss=1.893, generator_feat_match_loss=2.082, over 1006.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:35:20,898 INFO [train.py:811] (3/4) Start epoch 80
2023-11-13 01:38:04,839 INFO [train.py:467] (3/4) Epoch 80, batch 27, global_batch_idx: 2950, batch size: 52, loss[discriminator_loss=2.773, discriminator_real_loss=1.37, discriminator_fake_loss=1.403, generator_loss=31.73, generator_mel_loss=24.27, generator_kl_loss=1.985, generator_dur_loss=1.802, generator_adv_loss=1.693, generator_feat_match_loss=1.977, over 52.00 samples.], tot_loss[discriminator_loss=2.73, discriminator_real_loss=1.397, discriminator_fake_loss=1.332, generator_loss=31.82, generator_mel_loss=24.04, generator_kl_loss=1.984, generator_dur_loss=1.818, generator_adv_loss=1.879, generator_feat_match_loss=2.102, over 2126.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:38:48,578 INFO [train.py:811] (3/4) Start epoch 81
2023-11-13 01:42:32,760 INFO [train.py:811] (3/4) Start epoch 82
2023-11-13 01:43:03,060 INFO [train.py:467] (3/4) Epoch 82, batch 3, global_batch_idx: 3000, batch size: 52, loss[discriminator_loss=2.752, discriminator_real_loss=1.176, discriminator_fake_loss=1.576, generator_loss=31.98, generator_mel_loss=24.05, generator_kl_loss=1.924, generator_dur_loss=1.804, generator_adv_loss=2.066, generator_feat_match_loss=2.133, over 52.00 samples.], tot_loss[discriminator_loss=2.733, discriminator_real_loss=1.343, discriminator_fake_loss=1.39, generator_loss=32.2, generator_mel_loss=24.32, generator_kl_loss=1.962, generator_dur_loss=1.802, generator_adv_loss=1.93, generator_feat_match_loss=2.187, over 284.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:43:03,062 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 01:43:14,687 INFO [train.py:517] (3/4) Epoch 82, validation: discriminator_loss=2.629, discriminator_real_loss=1.475, discriminator_fake_loss=1.154, generator_loss=32.79, generator_mel_loss=24.73, generator_kl_loss=2.072, generator_dur_loss=1.789, generator_adv_loss=2.068, generator_feat_match_loss=2.126, over 100.00 samples.
2023-11-13 01:43:14,688 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 01:46:15,592 INFO [train.py:811] (3/4) Start epoch 83
2023-11-13 01:48:02,215 INFO [train.py:467] (3/4) Epoch 83, batch 16, global_batch_idx: 3050, batch size: 85, loss[discriminator_loss=2.746, discriminator_real_loss=1.474, discriminator_fake_loss=1.273, generator_loss=30.88, generator_mel_loss=23.42, generator_kl_loss=2.011, generator_dur_loss=1.818, generator_adv_loss=1.707, generator_feat_match_loss=1.918, over 85.00 samples.], tot_loss[discriminator_loss=2.766, discriminator_real_loss=1.422, discriminator_fake_loss=1.344, generator_loss=31.95, generator_mel_loss=24.1, generator_kl_loss=2.003, generator_dur_loss=1.814, generator_adv_loss=1.904, generator_feat_match_loss=2.13, over 1398.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:49:53,465 INFO [train.py:811] (3/4) Start epoch 84
2023-11-13 01:52:41,967 INFO [train.py:467] (3/4) Epoch 84, batch 29, global_batch_idx: 3100, batch size: 63, loss[discriminator_loss=2.691, discriminator_real_loss=1.293, discriminator_fake_loss=1.398, generator_loss=31.23, generator_mel_loss=23.14, generator_kl_loss=1.968, generator_dur_loss=1.815, generator_adv_loss=2.191, generator_feat_match_loss=2.117, over 63.00 samples.], tot_loss[discriminator_loss=2.734, discriminator_real_loss=1.396, discriminator_fake_loss=1.338, generator_loss=31.65, generator_mel_loss=23.84, generator_kl_loss=1.989, generator_dur_loss=1.812, generator_adv_loss=1.903, generator_feat_match_loss=2.105, over 1950.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 01:53:25,290 INFO [train.py:811] (3/4) Start epoch 85
2023-11-13 01:56:53,107 INFO [train.py:811] (3/4) Start epoch 86
2023-11-13 01:57:32,063 INFO [train.py:467] (3/4) Epoch 86, batch 5, global_batch_idx: 3150, batch size: 81, loss[discriminator_loss=2.779, discriminator_real_loss=1.354, discriminator_fake_loss=1.426, generator_loss=31.81, generator_mel_loss=24.08, generator_kl_loss=1.979, generator_dur_loss=1.822, generator_adv_loss=1.939, generator_feat_match_loss=1.988, over 81.00 samples.], tot_loss[discriminator_loss=2.766, discriminator_real_loss=1.4, discriminator_fake_loss=1.365, generator_loss=31.29, generator_mel_loss=23.65, generator_kl_loss=1.966, generator_dur_loss=1.82, generator_adv_loss=1.862, generator_feat_match_loss=1.995, over 392.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:00:25,721 INFO [train.py:811] (3/4) Start epoch 87
2023-11-13 02:02:21,717 INFO [train.py:467] (3/4) Epoch 87, batch 18, global_batch_idx: 3200, batch size: 65, loss[discriminator_loss=2.768, discriminator_real_loss=1.408, discriminator_fake_loss=1.359, generator_loss=31.74, generator_mel_loss=23.78, generator_kl_loss=2.045, generator_dur_loss=1.802, generator_adv_loss=1.842, generator_feat_match_loss=2.27, over 65.00 samples.], tot_loss[discriminator_loss=2.73, discriminator_real_loss=1.393, discriminator_fake_loss=1.337, generator_loss=31.67, generator_mel_loss=23.79, generator_kl_loss=1.995, generator_dur_loss=1.803, generator_adv_loss=1.898, generator_feat_match_loss=2.184, over 1450.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:02:21,718 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 02:02:32,500 INFO [train.py:517] (3/4) Epoch 87, validation: discriminator_loss=2.654, discriminator_real_loss=1.294, discriminator_fake_loss=1.36, generator_loss=32.68, generator_mel_loss=24.71, generator_kl_loss=2.052, generator_dur_loss=1.781, generator_adv_loss=1.804, generator_feat_match_loss=2.333, over 100.00 samples.
2023-11-13 02:02:32,501 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 02:04:11,508 INFO [train.py:811] (3/4) Start epoch 88
2023-11-13 02:07:12,341 INFO [train.py:467] (3/4) Epoch 88, batch 31, global_batch_idx: 3250, batch size: 52, loss[discriminator_loss=2.721, discriminator_real_loss=1.611, discriminator_fake_loss=1.109, generator_loss=32.02, generator_mel_loss=24.38, generator_kl_loss=1.966, generator_dur_loss=1.81, generator_adv_loss=1.597, generator_feat_match_loss=2.273, over 52.00 samples.], tot_loss[discriminator_loss=2.727, discriminator_real_loss=1.4, discriminator_fake_loss=1.328, generator_loss=31.51, generator_mel_loss=23.67, generator_kl_loss=1.967, generator_dur_loss=1.806, generator_adv_loss=1.9, generator_feat_match_loss=2.172, over 2147.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:07:43,323 INFO [train.py:811] (3/4) Start epoch 89
2023-11-13 02:11:19,120 INFO [train.py:811] (3/4) Start epoch 90
2023-11-13 02:12:13,396 INFO [train.py:467] (3/4) Epoch 90, batch 7, global_batch_idx: 3300, batch size: 52, loss[discriminator_loss=2.793, discriminator_real_loss=1.364, discriminator_fake_loss=1.43, generator_loss=31.23, generator_mel_loss=23.46, generator_kl_loss=1.986, generator_dur_loss=1.818, generator_adv_loss=2.041, generator_feat_match_loss=1.921, over 52.00 samples.], tot_loss[discriminator_loss=2.775, discriminator_real_loss=1.428, discriminator_fake_loss=1.347, generator_loss=31.39, generator_mel_loss=23.68, generator_kl_loss=1.976, generator_dur_loss=1.797, generator_adv_loss=1.891, generator_feat_match_loss=2.046, over 571.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:14:56,210 INFO [train.py:811] (3/4) Start epoch 91
2023-11-13 02:17:00,000 INFO [train.py:467] (3/4) Epoch 91, batch 20, global_batch_idx: 3350, batch size: 59, loss[discriminator_loss=2.719, discriminator_real_loss=1.393, discriminator_fake_loss=1.326, generator_loss=31.29, generator_mel_loss=23.61, generator_kl_loss=1.945, generator_dur_loss=1.82, generator_adv_loss=1.889, generator_feat_match_loss=2.027, over 59.00 samples.], tot_loss[discriminator_loss=2.754, discriminator_real_loss=1.412, discriminator_fake_loss=1.342, generator_loss=31.37, generator_mel_loss=23.55, generator_kl_loss=1.99, generator_dur_loss=1.799, generator_adv_loss=1.9, generator_feat_match_loss=2.137, over 1519.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:18:22,990 INFO [train.py:811] (3/4) Start epoch 92
2023-11-13 02:21:30,595 INFO [train.py:467] (3/4) Epoch 92, batch 33, global_batch_idx: 3400, batch size: 51, loss[discriminator_loss=2.816, discriminator_real_loss=1.304, discriminator_fake_loss=1.513, generator_loss=32.02, generator_mel_loss=24.09, generator_kl_loss=1.871, generator_dur_loss=1.798, generator_adv_loss=2.133, generator_feat_match_loss=2.131, over 51.00 samples.], tot_loss[discriminator_loss=2.731, discriminator_real_loss=1.396, discriminator_fake_loss=1.334, generator_loss=31.62, generator_mel_loss=23.74, generator_kl_loss=1.99, generator_dur_loss=1.795, generator_adv_loss=1.895, generator_feat_match_loss=2.208, over 2384.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:21:30,596 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 02:21:41,597 INFO [train.py:517] (3/4) Epoch 92, validation: discriminator_loss=2.768, discriminator_real_loss=1.521, discriminator_fake_loss=1.246, generator_loss=33.17, generator_mel_loss=24.83, generator_kl_loss=2.05, generator_dur_loss=1.77, generator_adv_loss=2.067, generator_feat_match_loss=2.454, over 100.00 samples.
2023-11-13 02:21:41,598 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 02:21:59,767 INFO [train.py:811] (3/4) Start epoch 93
2023-11-13 02:25:30,851 INFO [train.py:811] (3/4) Start epoch 94
2023-11-13 02:26:32,235 INFO [train.py:467] (3/4) Epoch 94, batch 9, global_batch_idx: 3450, batch size: 64, loss[discriminator_loss=2.775, discriminator_real_loss=1.383, discriminator_fake_loss=1.393, generator_loss=31.57, generator_mel_loss=23.51, generator_kl_loss=2.146, generator_dur_loss=1.786, generator_adv_loss=1.962, generator_feat_match_loss=2.166, over 64.00 samples.], tot_loss[discriminator_loss=2.776, discriminator_real_loss=1.427, discriminator_fake_loss=1.35, generator_loss=31.38, generator_mel_loss=23.61, generator_kl_loss=1.966, generator_dur_loss=1.796, generator_adv_loss=1.88, generator_feat_match_loss=2.125, over 689.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:29:02,009 INFO [train.py:811] (3/4) Start epoch 95
2023-11-13 02:31:26,332 INFO [train.py:467] (3/4) Epoch 95, batch 22, global_batch_idx: 3500, batch size: 90, loss[discriminator_loss=2.746, discriminator_real_loss=1.583, discriminator_fake_loss=1.162, generator_loss=31.07, generator_mel_loss=23.27, generator_kl_loss=1.962, generator_dur_loss=1.814, generator_adv_loss=1.891, generator_feat_match_loss=2.133, over 90.00 samples.], tot_loss[discriminator_loss=2.768, discriminator_real_loss=1.422, discriminator_fake_loss=1.347, generator_loss=31.07, generator_mel_loss=23.4, generator_kl_loss=1.96, generator_dur_loss=1.797, generator_adv_loss=1.838, generator_feat_match_loss=2.069, over 1591.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:32:38,825 INFO [train.py:811] (3/4) Start epoch 96
2023-11-13 02:35:56,134 INFO [train.py:467] (3/4) Epoch 96, batch 35, global_batch_idx: 3550, batch size: 52, loss[discriminator_loss=2.811, discriminator_real_loss=1.339, discriminator_fake_loss=1.472, generator_loss=30.75, generator_mel_loss=22.94, generator_kl_loss=1.965, generator_dur_loss=1.781, generator_adv_loss=1.916, generator_feat_match_loss=2.143, over 52.00 samples.], tot_loss[discriminator_loss=2.743, discriminator_real_loss=1.388, discriminator_fake_loss=1.355, generator_loss=31.09, generator_mel_loss=23.32, generator_kl_loss=1.961, generator_dur_loss=1.789, generator_adv_loss=1.89, generator_feat_match_loss=2.134, over 2607.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:36:02,695 INFO [train.py:811] (3/4) Start epoch 97
2023-11-13 02:39:34,222 INFO [train.py:811] (3/4) Start epoch 98
2023-11-13 02:40:52,654 INFO [train.py:467] (3/4) Epoch 98, batch 11, global_batch_idx: 3600, batch size: 54, loss[discriminator_loss=2.777, discriminator_real_loss=1.393, discriminator_fake_loss=1.385, generator_loss=31.42, generator_mel_loss=23.72, generator_kl_loss=1.868, generator_dur_loss=1.792, generator_adv_loss=1.9, generator_feat_match_loss=2.137, over 54.00 samples.], tot_loss[discriminator_loss=2.751, discriminator_real_loss=1.402, discriminator_fake_loss=1.349, generator_loss=30.88, generator_mel_loss=23.3, generator_kl_loss=1.931, generator_dur_loss=1.796, generator_adv_loss=1.82, generator_feat_match_loss=2.033, over 731.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:40:52,656 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 02:41:03,531 INFO [train.py:517] (3/4) Epoch 98, validation: discriminator_loss=2.675, discriminator_real_loss=1.354, discriminator_fake_loss=1.321, generator_loss=32.06, generator_mel_loss=24.26, generator_kl_loss=2.058, generator_dur_loss=1.766, generator_adv_loss=1.862, generator_feat_match_loss=2.108, over 100.00 samples.
2023-11-13 02:41:03,533 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 02:43:14,235 INFO [train.py:811] (3/4) Start epoch 99
2023-11-13 02:45:42,503 INFO [train.py:467] (3/4) Epoch 99, batch 24, global_batch_idx: 3650, batch size: 69, loss[discriminator_loss=2.828, discriminator_real_loss=1.689, discriminator_fake_loss=1.14, generator_loss=31.24, generator_mel_loss=23.5, generator_kl_loss=2.029, generator_dur_loss=1.776, generator_adv_loss=1.734, generator_feat_match_loss=2.197, over 69.00 samples.], tot_loss[discriminator_loss=2.752, discriminator_real_loss=1.396, discriminator_fake_loss=1.355, generator_loss=31.28, generator_mel_loss=23.46, generator_kl_loss=1.977, generator_dur_loss=1.793, generator_adv_loss=1.893, generator_feat_match_loss=2.154, over 1636.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:46:48,967 INFO [train.py:811] (3/4) Start epoch 100
2023-11-13 02:50:27,438 INFO [train.py:811] (3/4) Start epoch 101
2023-11-13 02:50:46,135 INFO [train.py:467] (3/4) Epoch 101, batch 0, global_batch_idx: 3700, batch size: 53, loss[discriminator_loss=2.824, discriminator_real_loss=1.339, discriminator_fake_loss=1.484, generator_loss=29.54, generator_mel_loss=22.21, generator_kl_loss=1.912, generator_dur_loss=1.79, generator_adv_loss=1.775, generator_feat_match_loss=1.854, over 53.00 samples.], tot_loss[discriminator_loss=2.824, discriminator_real_loss=1.339, discriminator_fake_loss=1.484, generator_loss=29.54, generator_mel_loss=22.21, generator_kl_loss=1.912, generator_dur_loss=1.79, generator_adv_loss=1.775, generator_feat_match_loss=1.854, over 53.00 samples.], cur_lr_g: 1.98e-04, cur_lr_d: 1.98e-04, grad_scale: 64.0
2023-11-13 02:54:03,308 INFO [train.py:811] (3/4) Start epoch 102
2023-11-13 02:55:22,967 INFO [train.py:467] (3/4) Epoch 102, batch 13, global_batch_idx: 3750, batch size: 67, loss[discriminator_loss=2.699, discriminator_real_loss=1.521, discriminator_fake_loss=1.178, generator_loss=31.02, generator_mel_loss=23.29, generator_kl_loss=1.982, generator_dur_loss=1.797, generator_adv_loss=1.668, generator_feat_match_loss=2.287, over 67.00 samples.], tot_loss[discriminator_loss=2.718, discriminator_real_loss=1.383, discriminator_fake_loss=1.334, generator_loss=30.79, generator_mel_loss=23.09, generator_kl_loss=1.929, generator_dur_loss=1.784, generator_adv_loss=1.852, generator_feat_match_loss=2.134, over 989.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 64.0
2023-11-13 02:57:40,842 INFO [train.py:811] (3/4) Start epoch 103
2023-11-13 03:00:04,355 INFO [train.py:467] (3/4) Epoch 103, batch 26, global_batch_idx: 3800, batch size: 55, loss[discriminator_loss=2.797, discriminator_real_loss=1.396, discriminator_fake_loss=1.399, generator_loss=30.52, generator_mel_loss=22.8, generator_kl_loss=1.942, generator_dur_loss=1.79, generator_adv_loss=1.96, generator_feat_match_loss=2.021, over 55.00 samples.], tot_loss[discriminator_loss=2.725, discriminator_real_loss=1.387, discriminator_fake_loss=1.338, generator_loss=31.06, generator_mel_loss=23.29, generator_kl_loss=1.975, generator_dur_loss=1.786, generator_adv_loss=1.85, generator_feat_match_loss=2.149, over 1791.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 64.0
2023-11-13 03:00:04,356 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 03:00:15,274 INFO [train.py:517] (3/4) Epoch 103, validation: discriminator_loss=2.699, discriminator_real_loss=1.437, discriminator_fake_loss=1.263, generator_loss=31.77, generator_mel_loss=23.74, generator_kl_loss=2.119, generator_dur_loss=1.75, generator_adv_loss=1.923, generator_feat_match_loss=2.244, over 100.00 samples.
2023-11-13 03:00:15,275 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 03:01:13,844 INFO [train.py:811] (3/4) Start epoch 104
2023-11-13 03:04:45,309 INFO [train.py:811] (3/4) Start epoch 105
2023-11-13 03:05:16,136 INFO [train.py:467] (3/4) Epoch 105, batch 2, global_batch_idx: 3850, batch size: 52, loss[discriminator_loss=2.754, discriminator_real_loss=1.416, discriminator_fake_loss=1.338, generator_loss=31.38, generator_mel_loss=23.33, generator_kl_loss=1.956, generator_dur_loss=1.8, generator_adv_loss=2.051, generator_feat_match_loss=2.242, over 52.00 samples.], tot_loss[discriminator_loss=2.73, discriminator_real_loss=1.406, discriminator_fake_loss=1.325, generator_loss=30.87, generator_mel_loss=22.96, generator_kl_loss=1.955, generator_dur_loss=1.799, generator_adv_loss=1.993, generator_feat_match_loss=2.164, over 179.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 64.0
2023-11-13 03:08:13,871 INFO [train.py:811] (3/4) Start epoch 106
2023-11-13 03:09:43,373 INFO [train.py:467] (3/4) Epoch 106, batch 15, global_batch_idx: 3900, batch size: 67, loss[discriminator_loss=2.789, discriminator_real_loss=1.592, discriminator_fake_loss=1.197, generator_loss=31.46, generator_mel_loss=23.47, generator_kl_loss=2.023, generator_dur_loss=1.799, generator_adv_loss=2.035, generator_feat_match_loss=2.131, over 67.00 samples.], tot_loss[discriminator_loss=2.766, discriminator_real_loss=1.403, discriminator_fake_loss=1.363, generator_loss=31.14, generator_mel_loss=23.39, generator_kl_loss=1.965, generator_dur_loss=1.789, generator_adv_loss=1.873, generator_feat_match_loss=2.118, over 1165.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 64.0
2023-11-13 03:11:40,327 INFO [train.py:811] (3/4) Start epoch 107
2023-11-13 03:14:19,930 INFO [train.py:467] (3/4) Epoch 107, batch 28, global_batch_idx: 3950, batch size: 153, loss[discriminator_loss=2.744, discriminator_real_loss=1.451, discriminator_fake_loss=1.293, generator_loss=30.63, generator_mel_loss=23.03, generator_kl_loss=1.866, generator_dur_loss=1.778, generator_adv_loss=1.808, generator_feat_match_loss=2.145, over 153.00 samples.], tot_loss[discriminator_loss=2.752, discriminator_real_loss=1.4, discriminator_fake_loss=1.352, generator_loss=31.06, generator_mel_loss=23.26, generator_kl_loss=1.959, generator_dur_loss=1.78, generator_adv_loss=1.882, generator_feat_match_loss=2.185, over 2116.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 64.0
2023-11-13 03:15:13,839 INFO [train.py:811] (3/4) Start epoch 108
2023-11-13 03:18:42,982 INFO [train.py:811] (3/4) Start epoch 109
2023-11-13 03:19:19,575 INFO [train.py:467] (3/4) Epoch 109, batch 4, global_batch_idx: 4000, batch size: 85, loss[discriminator_loss=2.738, discriminator_real_loss=1.317, discriminator_fake_loss=1.42, generator_loss=30.55, generator_mel_loss=22.78, generator_kl_loss=1.911, generator_dur_loss=1.776, generator_adv_loss=2, generator_feat_match_loss=2.08, over 85.00 samples.], tot_loss[discriminator_loss=2.72, discriminator_real_loss=1.395, discriminator_fake_loss=1.325, generator_loss=30.87, generator_mel_loss=23.2, generator_kl_loss=1.946, generator_dur_loss=1.769, generator_adv_loss=1.854, generator_feat_match_loss=2.102, over 444.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:19:19,576 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 03:19:31,525 INFO [train.py:517] (3/4) Epoch 109, validation: discriminator_loss=2.658, discriminator_real_loss=1.349, discriminator_fake_loss=1.309, generator_loss=31.43, generator_mel_loss=23.69, generator_kl_loss=1.953, generator_dur_loss=1.745, generator_adv_loss=1.882, generator_feat_match_loss=2.155, over 100.00 samples.
2023-11-13 03:19:31,526 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 03:22:31,739 INFO [train.py:811] (3/4) Start epoch 110
2023-11-13 03:24:17,771 INFO [train.py:467] (3/4) Epoch 110, batch 17, global_batch_idx: 4050, batch size: 65, loss[discriminator_loss=2.707, discriminator_real_loss=1.226, discriminator_fake_loss=1.481, generator_loss=30.86, generator_mel_loss=22.8, generator_kl_loss=1.894, generator_dur_loss=1.751, generator_adv_loss=2.135, generator_feat_match_loss=2.283, over 65.00 samples.], tot_loss[discriminator_loss=2.738, discriminator_real_loss=1.384, discriminator_fake_loss=1.354, generator_loss=30.89, generator_mel_loss=23.05, generator_kl_loss=1.955, generator_dur_loss=1.772, generator_adv_loss=1.879, generator_feat_match_loss=2.231, over 1223.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:26:04,295 INFO [train.py:811] (3/4) Start epoch 111
2023-11-13 03:29:07,986 INFO [train.py:467] (3/4) Epoch 111, batch 30, global_batch_idx: 4100, batch size: 56, loss[discriminator_loss=2.887, discriminator_real_loss=1.764, discriminator_fake_loss=1.124, generator_loss=29.88, generator_mel_loss=22.53, generator_kl_loss=1.852, generator_dur_loss=1.791, generator_adv_loss=1.625, generator_feat_match_loss=2.084, over 56.00 samples.], tot_loss[discriminator_loss=2.771, discriminator_real_loss=1.423, discriminator_fake_loss=1.349, generator_loss=30.88, generator_mel_loss=23.06, generator_kl_loss=1.976, generator_dur_loss=1.771, generator_adv_loss=1.892, generator_feat_match_loss=2.182, over 2468.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:29:39,137 INFO [train.py:811] (3/4) Start epoch 112
2023-11-13 03:33:13,488 INFO [train.py:811] (3/4) Start epoch 113
2023-11-13 03:33:57,745 INFO [train.py:467] (3/4) Epoch 113, batch 6, global_batch_idx: 4150, batch size: 81, loss[discriminator_loss=2.736, discriminator_real_loss=1.22, discriminator_fake_loss=1.517, generator_loss=31.51, generator_mel_loss=23.24, generator_kl_loss=2.093, generator_dur_loss=1.794, generator_adv_loss=2.127, generator_feat_match_loss=2.248, over 81.00 samples.], tot_loss[discriminator_loss=2.755, discriminator_real_loss=1.385, discriminator_fake_loss=1.37, generator_loss=30.84, generator_mel_loss=23.01, generator_kl_loss=2.005, generator_dur_loss=1.788, generator_adv_loss=1.89, generator_feat_match_loss=2.148, over 450.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:36:40,073 INFO [train.py:811] (3/4) Start epoch 114
2023-11-13 03:38:42,379 INFO [train.py:467] (3/4) Epoch 114, batch 19, global_batch_idx: 4200, batch size: 61, loss[discriminator_loss=2.727, discriminator_real_loss=1.385, discriminator_fake_loss=1.342, generator_loss=30.86, generator_mel_loss=22.95, generator_kl_loss=1.916, generator_dur_loss=1.782, generator_adv_loss=2.01, generator_feat_match_loss=2.199, over 61.00 samples.], tot_loss[discriminator_loss=2.749, discriminator_real_loss=1.393, discriminator_fake_loss=1.356, generator_loss=30.79, generator_mel_loss=22.99, generator_kl_loss=1.966, generator_dur_loss=1.775, generator_adv_loss=1.867, generator_feat_match_loss=2.196, over 1372.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:38:42,381 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 03:38:54,191 INFO [train.py:517] (3/4) Epoch 114, validation: discriminator_loss=2.653, discriminator_real_loss=1.35, discriminator_fake_loss=1.303, generator_loss=31.92, generator_mel_loss=23.96, generator_kl_loss=2.078, generator_dur_loss=1.744, generator_adv_loss=1.877, generator_feat_match_loss=2.255, over 100.00 samples.
2023-11-13 03:38:54,191 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 03:40:19,591 INFO [train.py:811] (3/4) Start epoch 115
2023-11-13 03:43:25,641 INFO [train.py:467] (3/4) Epoch 115, batch 32, global_batch_idx: 4250, batch size: 53, loss[discriminator_loss=2.707, discriminator_real_loss=1.405, discriminator_fake_loss=1.303, generator_loss=29.9, generator_mel_loss=22.13, generator_kl_loss=1.904, generator_dur_loss=1.741, generator_adv_loss=1.893, generator_feat_match_loss=2.229, over 53.00 samples.], tot_loss[discriminator_loss=2.758, discriminator_real_loss=1.392, discriminator_fake_loss=1.366, generator_loss=30.65, generator_mel_loss=22.84, generator_kl_loss=1.969, generator_dur_loss=1.771, generator_adv_loss=1.88, generator_feat_match_loss=2.182, over 2317.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:43:48,523 INFO [train.py:811] (3/4) Start epoch 116
2023-11-13 03:47:13,957 INFO [train.py:811] (3/4) Start epoch 117
2023-11-13 03:48:13,617 INFO [train.py:467] (3/4) Epoch 117, batch 8, global_batch_idx: 4300, batch size: 101, loss[discriminator_loss=2.781, discriminator_real_loss=1.547, discriminator_fake_loss=1.233, generator_loss=30.93, generator_mel_loss=23.04, generator_kl_loss=1.995, generator_dur_loss=1.751, generator_adv_loss=1.819, generator_feat_match_loss=2.316, over 101.00 samples.], tot_loss[discriminator_loss=2.765, discriminator_real_loss=1.392, discriminator_fake_loss=1.373, generator_loss=30.76, generator_mel_loss=22.92, generator_kl_loss=1.988, generator_dur_loss=1.772, generator_adv_loss=1.888, generator_feat_match_loss=2.189, over 745.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:50:52,073 INFO [train.py:811] (3/4) Start epoch 118
2023-11-13 03:53:08,233 INFO [train.py:467] (3/4) Epoch 118, batch 21, global_batch_idx: 4350, batch size: 59, loss[discriminator_loss=2.721, discriminator_real_loss=1.369, discriminator_fake_loss=1.352, generator_loss=31.64, generator_mel_loss=23.57, generator_kl_loss=2.019, generator_dur_loss=1.799, generator_adv_loss=2.033, generator_feat_match_loss=2.217, over 59.00 samples.], tot_loss[discriminator_loss=2.768, discriminator_real_loss=1.406, discriminator_fake_loss=1.363, generator_loss=30.75, generator_mel_loss=23.01, generator_kl_loss=1.928, generator_dur_loss=1.776, generator_adv_loss=1.872, generator_feat_match_loss=2.17, over 1603.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:54:26,469 INFO [train.py:811] (3/4) Start epoch 119
2023-11-13 03:57:50,227 INFO [train.py:467] (3/4) Epoch 119, batch 34, global_batch_idx: 4400, batch size: 69, loss[discriminator_loss=2.703, discriminator_real_loss=1.307, discriminator_fake_loss=1.396, generator_loss=30.17, generator_mel_loss=22.61, generator_kl_loss=1.843, generator_dur_loss=1.745, generator_adv_loss=1.83, generator_feat_match_loss=2.145, over 69.00 samples.], tot_loss[discriminator_loss=2.754, discriminator_real_loss=1.395, discriminator_fake_loss=1.359, generator_loss=30.68, generator_mel_loss=22.87, generator_kl_loss=1.96, generator_dur_loss=1.767, generator_adv_loss=1.879, generator_feat_match_loss=2.202, over 2526.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 03:57:50,228 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 03:58:00,794 INFO [train.py:517] (3/4) Epoch 119, validation: discriminator_loss=2.664, discriminator_real_loss=1.22, discriminator_fake_loss=1.443, generator_loss=31.92, generator_mel_loss=23.95, generator_kl_loss=1.953, generator_dur_loss=1.737, generator_adv_loss=1.761, generator_feat_match_loss=2.518, over 100.00 samples.
2023-11-13 03:58:00,795 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 03:58:10,946 INFO [train.py:811] (3/4) Start epoch 120
2023-11-13 04:01:42,981 INFO [train.py:811] (3/4) Start epoch 121
2023-11-13 04:02:50,446 INFO [train.py:467] (3/4) Epoch 121, batch 10, global_batch_idx: 4450, batch size: 101, loss[discriminator_loss=2.719, discriminator_real_loss=1.348, discriminator_fake_loss=1.372, generator_loss=31.58, generator_mel_loss=23.62, generator_kl_loss=2.077, generator_dur_loss=1.763, generator_adv_loss=1.812, generator_feat_match_loss=2.309, over 101.00 samples.], tot_loss[discriminator_loss=2.755, discriminator_real_loss=1.403, discriminator_fake_loss=1.351, generator_loss=30.56, generator_mel_loss=22.77, generator_kl_loss=1.976, generator_dur_loss=1.774, generator_adv_loss=1.849, generator_feat_match_loss=2.185, over 712.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:05:14,856 INFO [train.py:811] (3/4) Start epoch 122
2023-11-13 04:07:37,379 INFO [train.py:467] (3/4) Epoch 122, batch 23, global_batch_idx: 4500, batch size: 53, loss[discriminator_loss=2.676, discriminator_real_loss=1.361, discriminator_fake_loss=1.313, generator_loss=30.47, generator_mel_loss=22.58, generator_kl_loss=1.91, generator_dur_loss=1.772, generator_adv_loss=1.863, generator_feat_match_loss=2.346, over 53.00 samples.], tot_loss[discriminator_loss=2.742, discriminator_real_loss=1.396, discriminator_fake_loss=1.346, generator_loss=30.55, generator_mel_loss=22.74, generator_kl_loss=1.965, generator_dur_loss=1.757, generator_adv_loss=1.862, generator_feat_match_loss=2.227, over 1875.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:08:46,429 INFO [train.py:811] (3/4) Start epoch 123
2023-11-13 04:12:13,053 INFO [train.py:467] (3/4) Epoch 123, batch 36, global_batch_idx: 4550, batch size: 60, loss[discriminator_loss=2.721, discriminator_real_loss=1.189, discriminator_fake_loss=1.531, generator_loss=30.94, generator_mel_loss=22.7, generator_kl_loss=1.997, generator_dur_loss=1.789, generator_adv_loss=2.168, generator_feat_match_loss=2.287, over 60.00 samples.], tot_loss[discriminator_loss=2.773, discriminator_real_loss=1.404, discriminator_fake_loss=1.369, generator_loss=30.61, generator_mel_loss=22.89, generator_kl_loss=1.973, generator_dur_loss=1.766, generator_adv_loss=1.852, generator_feat_match_loss=2.134, over 2761.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:12:13,832 INFO [train.py:811] (3/4) Start epoch 124
2023-11-13 04:15:43,210 INFO [train.py:811] (3/4) Start epoch 125
2023-11-13 04:17:08,794 INFO [train.py:467] (3/4) Epoch 125, batch 12, global_batch_idx: 4600, batch size: 59, loss[discriminator_loss=2.695, discriminator_real_loss=1.371, discriminator_fake_loss=1.323, generator_loss=30.74, generator_mel_loss=22.81, generator_kl_loss=1.971, generator_dur_loss=1.767, generator_adv_loss=1.848, generator_feat_match_loss=2.352, over 59.00 samples.], tot_loss[discriminator_loss=2.736, discriminator_real_loss=1.4, discriminator_fake_loss=1.336, generator_loss=30.44, generator_mel_loss=22.62, generator_kl_loss=1.981, generator_dur_loss=1.766, generator_adv_loss=1.876, generator_feat_match_loss=2.2, over 895.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:17:08,795 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 04:17:19,765 INFO [train.py:517] (3/4) Epoch 125, validation: discriminator_loss=2.659, discriminator_real_loss=1.309, discriminator_fake_loss=1.35, generator_loss=31.88, generator_mel_loss=23.86, generator_kl_loss=1.967, generator_dur_loss=1.736, generator_adv_loss=1.793, generator_feat_match_loss=2.523, over 100.00 samples.
2023-11-13 04:17:19,767 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 04:19:29,534 INFO [train.py:811] (3/4) Start epoch 126
2023-11-13 04:21:56,594 INFO [train.py:467] (3/4) Epoch 126, batch 25, global_batch_idx: 4650, batch size: 56, loss[discriminator_loss=2.75, discriminator_real_loss=1.424, discriminator_fake_loss=1.325, generator_loss=30.13, generator_mel_loss=22.32, generator_kl_loss=1.978, generator_dur_loss=1.747, generator_adv_loss=1.816, generator_feat_match_loss=2.268, over 56.00 samples.], tot_loss[discriminator_loss=2.741, discriminator_real_loss=1.396, discriminator_fake_loss=1.346, generator_loss=30.45, generator_mel_loss=22.62, generator_kl_loss=1.956, generator_dur_loss=1.76, generator_adv_loss=1.874, generator_feat_match_loss=2.248, over 1667.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:22:55,997 INFO [train.py:811] (3/4) Start epoch 127
2023-11-13 04:26:33,983 INFO [train.py:811] (3/4) Start epoch 128
2023-11-13 04:26:58,845 INFO [train.py:467] (3/4) Epoch 128, batch 1, global_batch_idx: 4700, batch size: 153, loss[discriminator_loss=2.715, discriminator_real_loss=1.35, discriminator_fake_loss=1.366, generator_loss=31.51, generator_mel_loss=23.15, generator_kl_loss=2.093, generator_dur_loss=1.724, generator_adv_loss=2.086, generator_feat_match_loss=2.457, over 153.00 samples.], tot_loss[discriminator_loss=2.739, discriminator_real_loss=1.327, discriminator_fake_loss=1.412, generator_loss=31.13, generator_mel_loss=22.96, generator_kl_loss=2.053, generator_dur_loss=1.732, generator_adv_loss=2.01, generator_feat_match_loss=2.369, over 213.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:30:11,201 INFO [train.py:811] (3/4) Start epoch 129
2023-11-13 04:31:40,401 INFO [train.py:467] (3/4) Epoch 129, batch 14, global_batch_idx: 4750, batch size: 110, loss[discriminator_loss=2.801, discriminator_real_loss=1.187, discriminator_fake_loss=1.615, generator_loss=30.8, generator_mel_loss=22.67, generator_kl_loss=1.951, generator_dur_loss=1.754, generator_adv_loss=2.184, generator_feat_match_loss=2.242, over 110.00 samples.], tot_loss[discriminator_loss=2.769, discriminator_real_loss=1.394, discriminator_fake_loss=1.374, generator_loss=30.58, generator_mel_loss=22.72, generator_kl_loss=1.953, generator_dur_loss=1.753, generator_adv_loss=1.871, generator_feat_match_loss=2.273, over 1323.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:33:45,599 INFO [train.py:811] (3/4) Start epoch 130
2023-11-13 04:36:24,713 INFO [train.py:467] (3/4) Epoch 130, batch 27, global_batch_idx: 4800, batch size: 64, loss[discriminator_loss=2.783, discriminator_real_loss=1.438, discriminator_fake_loss=1.346, generator_loss=29.65, generator_mel_loss=21.93, generator_kl_loss=1.982, generator_dur_loss=1.748, generator_adv_loss=1.772, generator_feat_match_loss=2.211, over 64.00 samples.], tot_loss[discriminator_loss=2.799, discriminator_real_loss=1.413, discriminator_fake_loss=1.387, generator_loss=30.31, generator_mel_loss=22.61, generator_kl_loss=1.967, generator_dur_loss=1.756, generator_adv_loss=1.838, generator_feat_match_loss=2.135, over 2088.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:36:24,714 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 04:36:35,810 INFO [train.py:517] (3/4) Epoch 130, validation: discriminator_loss=2.648, discriminator_real_loss=1.241, discriminator_fake_loss=1.407, generator_loss=30.79, generator_mel_loss=23.05, generator_kl_loss=2.004, generator_dur_loss=1.733, generator_adv_loss=1.731, generator_feat_match_loss=2.275, over 100.00 samples.
2023-11-13 04:36:35,811 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 04:37:26,031 INFO [train.py:811] (3/4) Start epoch 131
2023-11-13 04:40:55,960 INFO [train.py:811] (3/4) Start epoch 132
2023-11-13 04:41:29,813 INFO [train.py:467] (3/4) Epoch 132, batch 3, global_batch_idx: 4850, batch size: 101, loss[discriminator_loss=2.734, discriminator_real_loss=1.319, discriminator_fake_loss=1.414, generator_loss=30.93, generator_mel_loss=22.98, generator_kl_loss=2.039, generator_dur_loss=1.737, generator_adv_loss=1.782, generator_feat_match_loss=2.395, over 101.00 samples.], tot_loss[discriminator_loss=2.739, discriminator_real_loss=1.377, discriminator_fake_loss=1.362, generator_loss=31.1, generator_mel_loss=23.07, generator_kl_loss=2.031, generator_dur_loss=1.75, generator_adv_loss=1.851, generator_feat_match_loss=2.404, over 418.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:44:25,951 INFO [train.py:811] (3/4) Start epoch 133
2023-11-13 04:46:03,010 INFO [train.py:467] (3/4) Epoch 133, batch 16, global_batch_idx: 4900, batch size: 110, loss[discriminator_loss=2.783, discriminator_real_loss=1.457, discriminator_fake_loss=1.326, generator_loss=30.75, generator_mel_loss=22.9, generator_kl_loss=2.012, generator_dur_loss=1.725, generator_adv_loss=1.886, generator_feat_match_loss=2.227, over 110.00 samples.], tot_loss[discriminator_loss=2.767, discriminator_real_loss=1.417, discriminator_fake_loss=1.349, generator_loss=30.64, generator_mel_loss=22.77, generator_kl_loss=1.975, generator_dur_loss=1.753, generator_adv_loss=1.844, generator_feat_match_loss=2.289, over 1411.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:47:57,064 INFO [train.py:811] (3/4) Start epoch 134
2023-11-13 04:50:46,790 INFO [train.py:467] (3/4) Epoch 134, batch 29, global_batch_idx: 4950, batch size: 52, loss[discriminator_loss=2.77, discriminator_real_loss=1.368, discriminator_fake_loss=1.4, generator_loss=30.83, generator_mel_loss=22.76, generator_kl_loss=2.067, generator_dur_loss=1.762, generator_adv_loss=1.924, generator_feat_match_loss=2.314, over 52.00 samples.], tot_loss[discriminator_loss=2.789, discriminator_real_loss=1.417, discriminator_fake_loss=1.372, generator_loss=30.53, generator_mel_loss=22.71, generator_kl_loss=1.974, generator_dur_loss=1.754, generator_adv_loss=1.859, generator_feat_match_loss=2.228, over 2249.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:51:26,132 INFO [train.py:811] (3/4) Start epoch 135
2023-11-13 04:54:55,836 INFO [train.py:811] (3/4) Start epoch 136
2023-11-13 04:55:38,417 INFO [train.py:467] (3/4) Epoch 136, batch 5, global_batch_idx: 5000, batch size: 76, loss[discriminator_loss=2.676, discriminator_real_loss=1.308, discriminator_fake_loss=1.369, generator_loss=30.83, generator_mel_loss=22.89, generator_kl_loss=1.929, generator_dur_loss=1.744, generator_adv_loss=1.863, generator_feat_match_loss=2.402, over 76.00 samples.], tot_loss[discriminator_loss=2.723, discriminator_real_loss=1.378, discriminator_fake_loss=1.345, generator_loss=30.31, generator_mel_loss=22.48, generator_kl_loss=1.896, generator_dur_loss=1.751, generator_adv_loss=1.861, generator_feat_match_loss=2.33, over 406.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 04:55:38,419 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 04:55:50,229 INFO [train.py:517] (3/4) Epoch 136, validation: discriminator_loss=2.601, discriminator_real_loss=1.183, discriminator_fake_loss=1.418, generator_loss=31.35, generator_mel_loss=23.4, generator_kl_loss=1.903, generator_dur_loss=1.72, generator_adv_loss=1.731, generator_feat_match_loss=2.595, over 100.00 samples.
2023-11-13 04:55:50,230 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 04:58:35,221 INFO [train.py:811] (3/4) Start epoch 137
2023-11-13 05:00:30,155 INFO [train.py:467] (3/4) Epoch 137, batch 18, global_batch_idx: 5050, batch size: 79, loss[discriminator_loss=2.77, discriminator_real_loss=1.191, discriminator_fake_loss=1.578, generator_loss=30.43, generator_mel_loss=22.44, generator_kl_loss=1.966, generator_dur_loss=1.736, generator_adv_loss=2.18, generator_feat_match_loss=2.113, over 79.00 samples.], tot_loss[discriminator_loss=2.766, discriminator_real_loss=1.398, discriminator_fake_loss=1.368, generator_loss=30.37, generator_mel_loss=22.54, generator_kl_loss=1.981, generator_dur_loss=1.751, generator_adv_loss=1.873, generator_feat_match_loss=2.229, over 1405.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 05:02:10,249 INFO [train.py:811] (3/4) Start epoch 138
2023-11-13 05:05:17,214 INFO [train.py:467] (3/4) Epoch 138, batch 31, global_batch_idx: 5100, batch size: 56, loss[discriminator_loss=2.766, discriminator_real_loss=1.389, discriminator_fake_loss=1.378, generator_loss=29.53, generator_mel_loss=21.99, generator_kl_loss=1.864, generator_dur_loss=1.755, generator_adv_loss=1.898, generator_feat_match_loss=2.023, over 56.00 samples.], tot_loss[discriminator_loss=2.778, discriminator_real_loss=1.414, discriminator_fake_loss=1.365, generator_loss=30.42, generator_mel_loss=22.61, generator_kl_loss=1.976, generator_dur_loss=1.75, generator_adv_loss=1.857, generator_feat_match_loss=2.222, over 2365.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 05:05:45,635 INFO [train.py:811] (3/4) Start epoch 139
2023-11-13 05:09:14,754 INFO [train.py:811] (3/4) Start epoch 140
2023-11-13 05:10:06,743 INFO [train.py:467] (3/4) Epoch 140, batch 7, global_batch_idx: 5150, batch size: 52, loss[discriminator_loss=2.805, discriminator_real_loss=1.536, discriminator_fake_loss=1.27, generator_loss=30.15, generator_mel_loss=22.48, generator_kl_loss=1.886, generator_dur_loss=1.774, generator_adv_loss=1.786, generator_feat_match_loss=2.223, over 52.00 samples.], tot_loss[discriminator_loss=2.807, discriminator_real_loss=1.422, discriminator_fake_loss=1.385, generator_loss=30.41, generator_mel_loss=22.68, generator_kl_loss=1.965, generator_dur_loss=1.751, generator_adv_loss=1.836, generator_feat_match_loss=2.181, over 448.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 05:12:47,561 INFO [train.py:811] (3/4) Start epoch 141
2023-11-13 05:14:54,162 INFO [train.py:467] (3/4) Epoch 141, batch 20, global_batch_idx: 5200, batch size: 60, loss[discriminator_loss=2.758, discriminator_real_loss=1.413, discriminator_fake_loss=1.344, generator_loss=31.1, generator_mel_loss=23.2, generator_kl_loss=2.026, generator_dur_loss=1.761, generator_adv_loss=1.751, generator_feat_match_loss=2.365, over 60.00 samples.], tot_loss[discriminator_loss=2.756, discriminator_real_loss=1.387, discriminator_fake_loss=1.369, generator_loss=30.61, generator_mel_loss=22.65, generator_kl_loss=1.991, generator_dur_loss=1.746, generator_adv_loss=1.9, generator_feat_match_loss=2.314, over 1660.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 05:14:54,163 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 05:15:05,080 INFO [train.py:517] (3/4) Epoch 141, validation: discriminator_loss=2.734, discriminator_real_loss=1.313, discriminator_fake_loss=1.421, generator_loss=31.4, generator_mel_loss=23.61, generator_kl_loss=2.005, generator_dur_loss=1.718, generator_adv_loss=1.729, generator_feat_match_loss=2.339, over 100.00 samples.
2023-11-13 05:15:05,081 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 05:16:27,047 INFO [train.py:811] (3/4) Start epoch 142
2023-11-13 05:19:39,453 INFO [train.py:467] (3/4) Epoch 142, batch 33, global_batch_idx: 5250, batch size: 63, loss[discriminator_loss=2.734, discriminator_real_loss=1.331, discriminator_fake_loss=1.402, generator_loss=29.91, generator_mel_loss=22.1, generator_kl_loss=1.914, generator_dur_loss=1.718, generator_adv_loss=1.798, generator_feat_match_loss=2.375, over 63.00 samples.], tot_loss[discriminator_loss=2.749, discriminator_real_loss=1.397, discriminator_fake_loss=1.352, generator_loss=30.37, generator_mel_loss=22.44, generator_kl_loss=1.98, generator_dur_loss=1.744, generator_adv_loss=1.88, generator_feat_match_loss=2.321, over 2641.00 samples.], cur_lr_g: 1.97e-04, cur_lr_d: 1.97e-04, grad_scale: 128.0
2023-11-13 05:19:57,142 INFO [train.py:811] (3/4) Start epoch 143
2023-11-13 05:23:26,044 INFO [train.py:811] (3/4) Start epoch 144
2023-11-13 05:24:33,437 INFO [train.py:467] (3/4) Epoch 144, batch 9, global_batch_idx: 5300, batch size: 110, loss[discriminator_loss=2.672, discriminator_real_loss=1.431, discriminator_fake_loss=1.241, generator_loss=30.63, generator_mel_loss=22.73, generator_kl_loss=1.991, generator_dur_loss=1.729, generator_adv_loss=1.745, generator_feat_match_loss=2.426, over 110.00 samples.], tot_loss[discriminator_loss=2.702, discriminator_real_loss=1.373, discriminator_fake_loss=1.33, generator_loss=30.38, generator_mel_loss=22.45, generator_kl_loss=1.954, generator_dur_loss=1.757, generator_adv_loss=1.85, generator_feat_match_loss=2.369, over 746.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:27:00,389 INFO [train.py:811] (3/4) Start epoch 145
2023-11-13 05:29:15,247 INFO [train.py:467] (3/4) Epoch 145, batch 22, global_batch_idx: 5350, batch size: 102, loss[discriminator_loss=2.68, discriminator_real_loss=1.227, discriminator_fake_loss=1.452, generator_loss=30.76, generator_mel_loss=22.69, generator_kl_loss=1.986, generator_dur_loss=1.728, generator_adv_loss=1.873, generator_feat_match_loss=2.484, over 102.00 samples.], tot_loss[discriminator_loss=2.749, discriminator_real_loss=1.387, discriminator_fake_loss=1.362, generator_loss=30.58, generator_mel_loss=22.65, generator_kl_loss=1.993, generator_dur_loss=1.745, generator_adv_loss=1.864, generator_feat_match_loss=2.319, over 1798.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:30:35,932 INFO [train.py:811] (3/4) Start epoch 146
2023-11-13 05:34:02,582 INFO [train.py:467] (3/4) Epoch 146, batch 35, global_batch_idx: 5400, batch size: 61, loss[discriminator_loss=2.789, discriminator_real_loss=1.227, discriminator_fake_loss=1.562, generator_loss=30.8, generator_mel_loss=22.55, generator_kl_loss=1.982, generator_dur_loss=1.741, generator_adv_loss=2.088, generator_feat_match_loss=2.438, over 61.00 samples.], tot_loss[discriminator_loss=2.747, discriminator_real_loss=1.379, discriminator_fake_loss=1.368, generator_loss=30.49, generator_mel_loss=22.55, generator_kl_loss=1.968, generator_dur_loss=1.739, generator_adv_loss=1.892, generator_feat_match_loss=2.343, over 2786.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:34:02,583 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 05:34:13,539 INFO [train.py:517] (3/4) Epoch 146, validation: discriminator_loss=2.781, discriminator_real_loss=1.589, discriminator_fake_loss=1.192, generator_loss=31.94, generator_mel_loss=23.76, generator_kl_loss=2.033, generator_dur_loss=1.711, generator_adv_loss=2.037, generator_feat_match_loss=2.401, over 100.00 samples.
2023-11-13 05:34:13,541 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 05:34:18,945 INFO [train.py:811] (3/4) Start epoch 147
2023-11-13 05:37:48,527 INFO [train.py:811] (3/4) Start epoch 148
2023-11-13 05:38:57,718 INFO [train.py:467] (3/4) Epoch 148, batch 11, global_batch_idx: 5450, batch size: 85, loss[discriminator_loss=2.762, discriminator_real_loss=1.404, discriminator_fake_loss=1.356, generator_loss=30.02, generator_mel_loss=22.26, generator_kl_loss=2.031, generator_dur_loss=1.755, generator_adv_loss=1.886, generator_feat_match_loss=2.09, over 85.00 samples.], tot_loss[discriminator_loss=2.77, discriminator_real_loss=1.414, discriminator_fake_loss=1.356, generator_loss=30.05, generator_mel_loss=22.32, generator_kl_loss=1.98, generator_dur_loss=1.746, generator_adv_loss=1.841, generator_feat_match_loss=2.163, over 871.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:41:17,293 INFO [train.py:811] (3/4) Start epoch 149
2023-11-13 05:43:45,114 INFO [train.py:467] (3/4) Epoch 149, batch 24, global_batch_idx: 5500, batch size: 60, loss[discriminator_loss=2.746, discriminator_real_loss=1.498, discriminator_fake_loss=1.249, generator_loss=30.46, generator_mel_loss=22.44, generator_kl_loss=1.978, generator_dur_loss=1.765, generator_adv_loss=1.992, generator_feat_match_loss=2.285, over 60.00 samples.], tot_loss[discriminator_loss=2.767, discriminator_real_loss=1.4, discriminator_fake_loss=1.367, generator_loss=30.14, generator_mel_loss=22.26, generator_kl_loss=1.964, generator_dur_loss=1.745, generator_adv_loss=1.877, generator_feat_match_loss=2.289, over 1660.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:44:52,147 INFO [train.py:811] (3/4) Start epoch 150
2023-11-13 05:48:19,094 INFO [train.py:811] (3/4) Start epoch 151
2023-11-13 05:48:33,665 INFO [train.py:467] (3/4) Epoch 151, batch 0, global_batch_idx: 5550, batch size: 69, loss[discriminator_loss=2.66, discriminator_real_loss=1.317, discriminator_fake_loss=1.343, generator_loss=30.02, generator_mel_loss=21.89, generator_kl_loss=1.993, generator_dur_loss=1.733, generator_adv_loss=1.986, generator_feat_match_loss=2.422, over 69.00 samples.], tot_loss[discriminator_loss=2.66, discriminator_real_loss=1.317, discriminator_fake_loss=1.343, generator_loss=30.02, generator_mel_loss=21.89, generator_kl_loss=1.993, generator_dur_loss=1.733, generator_adv_loss=1.986, generator_feat_match_loss=2.422, over 69.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:51:49,084 INFO [train.py:811] (3/4) Start epoch 152
2023-11-13 05:53:09,420 INFO [train.py:467] (3/4) Epoch 152, batch 13, global_batch_idx: 5600, batch size: 53, loss[discriminator_loss=2.727, discriminator_real_loss=1.372, discriminator_fake_loss=1.355, generator_loss=30.16, generator_mel_loss=22.22, generator_kl_loss=2.002, generator_dur_loss=1.744, generator_adv_loss=1.899, generator_feat_match_loss=2.297, over 53.00 samples.], tot_loss[discriminator_loss=2.736, discriminator_real_loss=1.397, discriminator_fake_loss=1.339, generator_loss=30.1, generator_mel_loss=22.26, generator_kl_loss=1.953, generator_dur_loss=1.74, generator_adv_loss=1.847, generator_feat_match_loss=2.295, over 969.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:53:09,422 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 05:53:20,499 INFO [train.py:517] (3/4) Epoch 152, validation: discriminator_loss=2.72, discriminator_real_loss=1.382, discriminator_fake_loss=1.337, generator_loss=31.62, generator_mel_loss=23.69, generator_kl_loss=2.044, generator_dur_loss=1.705, generator_adv_loss=1.797, generator_feat_match_loss=2.381, over 100.00 samples.
2023-11-13 05:53:20,500 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 05:55:27,664 INFO [train.py:811] (3/4) Start epoch 153
2023-11-13 05:58:05,086 INFO [train.py:467] (3/4) Epoch 153, batch 26, global_batch_idx: 5650, batch size: 55, loss[discriminator_loss=2.738, discriminator_real_loss=1.347, discriminator_fake_loss=1.393, generator_loss=29.97, generator_mel_loss=22.03, generator_kl_loss=1.927, generator_dur_loss=1.771, generator_adv_loss=1.842, generator_feat_match_loss=2.404, over 55.00 samples.], tot_loss[discriminator_loss=2.757, discriminator_real_loss=1.392, discriminator_fake_loss=1.365, generator_loss=30.13, generator_mel_loss=22.29, generator_kl_loss=1.986, generator_dur_loss=1.742, generator_adv_loss=1.848, generator_feat_match_loss=2.269, over 2022.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 05:59:03,311 INFO [train.py:811] (3/4) Start epoch 154
2023-11-13 06:02:31,619 INFO [train.py:811] (3/4) Start epoch 155
2023-11-13 06:02:58,863 INFO [train.py:467] (3/4) Epoch 155, batch 2, global_batch_idx: 5700, batch size: 50, loss[discriminator_loss=2.801, discriminator_real_loss=1.354, discriminator_fake_loss=1.446, generator_loss=29.03, generator_mel_loss=21.33, generator_kl_loss=1.932, generator_dur_loss=1.746, generator_adv_loss=1.962, generator_feat_match_loss=2.066, over 50.00 samples.], tot_loss[discriminator_loss=2.775, discriminator_real_loss=1.422, discriminator_fake_loss=1.353, generator_loss=29.92, generator_mel_loss=22.12, generator_kl_loss=1.915, generator_dur_loss=1.74, generator_adv_loss=1.877, generator_feat_match_loss=2.27, over 187.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 06:06:06,791 INFO [train.py:811] (3/4) Start epoch 156
2023-11-13 06:07:45,324 INFO [train.py:467] (3/4) Epoch 156, batch 15, global_batch_idx: 5750, batch size: 126, loss[discriminator_loss=2.715, discriminator_real_loss=1.27, discriminator_fake_loss=1.446, generator_loss=30.55, generator_mel_loss=22.4, generator_kl_loss=2.014, generator_dur_loss=1.72, generator_adv_loss=2.021, generator_feat_match_loss=2.402, over 126.00 samples.], tot_loss[discriminator_loss=2.742, discriminator_real_loss=1.381, discriminator_fake_loss=1.36, generator_loss=30.22, generator_mel_loss=22.26, generator_kl_loss=1.958, generator_dur_loss=1.733, generator_adv_loss=1.902, generator_feat_match_loss=2.372, over 1258.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 128.0
2023-11-13 06:09:43,617 INFO [train.py:811] (3/4) Start epoch 157
2023-11-13 06:12:34,808 INFO [train.py:467] (3/4) Epoch 157, batch 28, global_batch_idx: 5800, batch size: 52, loss[discriminator_loss=2.816, discriminator_real_loss=1.274, discriminator_fake_loss=1.543, generator_loss=29.7, generator_mel_loss=21.89, generator_kl_loss=2.058, generator_dur_loss=1.747, generator_adv_loss=2.021, generator_feat_match_loss=1.986, over 52.00 samples.], tot_loss[discriminator_loss=2.754, discriminator_real_loss=1.374, discriminator_fake_loss=1.38, generator_loss=30.45, generator_mel_loss=22.38, generator_kl_loss=1.971, generator_dur_loss=1.737, generator_adv_loss=1.932, generator_feat_match_loss=2.432, over 2075.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 64.0
2023-11-13 06:12:34,810 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 06:12:46,006 INFO [train.py:517] (3/4) Epoch 157, validation: discriminator_loss=2.75, discriminator_real_loss=1.503, discriminator_fake_loss=1.247, generator_loss=31.28, generator_mel_loss=23.14, generator_kl_loss=2.07, generator_dur_loss=1.715, generator_adv_loss=1.982, generator_feat_match_loss=2.37, over 100.00 samples.
2023-11-13 06:12:46,007 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 06:13:25,127 INFO [train.py:811] (3/4) Start epoch 158
2023-11-13 06:16:56,466 INFO [train.py:811] (3/4) Start epoch 159
2023-11-13 06:17:37,074 INFO [train.py:467] (3/4) Epoch 159, batch 4, global_batch_idx: 5850, batch size: 61, loss[discriminator_loss=2.674, discriminator_real_loss=1.26, discriminator_fake_loss=1.414, generator_loss=30.61, generator_mel_loss=22.56, generator_kl_loss=2.046, generator_dur_loss=1.738, generator_adv_loss=1.809, generator_feat_match_loss=2.459, over 61.00 samples.], tot_loss[discriminator_loss=2.72, discriminator_real_loss=1.396, discriminator_fake_loss=1.324, generator_loss=30.59, generator_mel_loss=22.51, generator_kl_loss=1.996, generator_dur_loss=1.748, generator_adv_loss=1.908, generator_feat_match_loss=2.437, over 341.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 64.0
2023-11-13 06:20:24,543 INFO [train.py:811] (3/4) Start epoch 160
2023-11-13 06:22:15,076 INFO [train.py:467] (3/4) Epoch 160, batch 17, global_batch_idx: 5900, batch size: 54, loss[discriminator_loss=2.555, discriminator_real_loss=1.376, discriminator_fake_loss=1.178, generator_loss=31.41, generator_mel_loss=22.65, generator_kl_loss=1.922, generator_dur_loss=1.762, generator_adv_loss=2.047, generator_feat_match_loss=3.033, over 54.00 samples.], tot_loss[discriminator_loss=2.718, discriminator_real_loss=1.357, discriminator_fake_loss=1.361, generator_loss=31.06, generator_mel_loss=22.46, generator_kl_loss=1.975, generator_dur_loss=1.738, generator_adv_loss=2.108, generator_feat_match_loss=2.778, over 1330.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 32.0
2023-11-13 06:23:58,052 INFO [train.py:811] (3/4) Start epoch 161
2023-11-13 06:27:03,377 INFO [train.py:467] (3/4) Epoch 161, batch 30, global_batch_idx: 5950, batch size: 79, loss[discriminator_loss=2.801, discriminator_real_loss=1.212, discriminator_fake_loss=1.59, generator_loss=30.4, generator_mel_loss=22.05, generator_kl_loss=1.963, generator_dur_loss=1.751, generator_adv_loss=2.016, generator_feat_match_loss=2.617, over 79.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.318, discriminator_fake_loss=1.329, generator_loss=31.23, generator_mel_loss=22.51, generator_kl_loss=1.984, generator_dur_loss=1.731, generator_adv_loss=2.111, generator_feat_match_loss=2.892, over 2382.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 32.0
2023-11-13 06:27:31,326 INFO [train.py:811] (3/4) Start epoch 162
2023-11-13 06:31:06,566 INFO [train.py:811] (3/4) Start epoch 163
2023-11-13 06:31:59,803 INFO [train.py:467] (3/4) Epoch 163, batch 6, global_batch_idx: 6000, batch size: 58, loss[discriminator_loss=2.668, discriminator_real_loss=1.316, discriminator_fake_loss=1.353, generator_loss=30.4, generator_mel_loss=22.38, generator_kl_loss=1.953, generator_dur_loss=1.713, generator_adv_loss=1.828, generator_feat_match_loss=2.529, over 58.00 samples.], tot_loss[discriminator_loss=2.724, discriminator_real_loss=1.395, discriminator_fake_loss=1.328, generator_loss=30.04, generator_mel_loss=22.19, generator_kl_loss=1.92, generator_dur_loss=1.737, generator_adv_loss=1.86, generator_feat_match_loss=2.326, over 532.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 32.0
2023-11-13 06:31:59,805 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 06:32:13,077 INFO [train.py:517] (3/4) Epoch 163, validation: discriminator_loss=2.666, discriminator_real_loss=1.218, discriminator_fake_loss=1.447, generator_loss=31.02, generator_mel_loss=23.15, generator_kl_loss=2.002, generator_dur_loss=1.701, generator_adv_loss=1.687, generator_feat_match_loss=2.479, over 100.00 samples.
2023-11-13 06:32:13,079 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 06:34:52,735 INFO [train.py:811] (3/4) Start epoch 164
2023-11-13 06:36:51,839 INFO [train.py:467] (3/4) Epoch 164, batch 19, global_batch_idx: 6050, batch size: 153, loss[discriminator_loss=2.529, discriminator_real_loss=1.268, discriminator_fake_loss=1.262, generator_loss=30.74, generator_mel_loss=22.23, generator_kl_loss=2.032, generator_dur_loss=1.718, generator_adv_loss=1.883, generator_feat_match_loss=2.877, over 153.00 samples.], tot_loss[discriminator_loss=2.746, discriminator_real_loss=1.409, discriminator_fake_loss=1.337, generator_loss=30.89, generator_mel_loss=22.26, generator_kl_loss=1.953, generator_dur_loss=1.728, generator_adv_loss=2.124, generator_feat_match_loss=2.822, over 1617.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 32.0
2023-11-13 06:38:26,596 INFO [train.py:811] (3/4) Start epoch 165
2023-11-13 06:41:30,882 INFO [train.py:467] (3/4) Epoch 165, batch 32, global_batch_idx: 6100, batch size: 85, loss[discriminator_loss=2.543, discriminator_real_loss=1.297, discriminator_fake_loss=1.247, generator_loss=30.65, generator_mel_loss=21.9, generator_kl_loss=1.905, generator_dur_loss=1.736, generator_adv_loss=2.121, generator_feat_match_loss=2.986, over 85.00 samples.], tot_loss[discriminator_loss=2.641, discriminator_real_loss=1.324, discriminator_fake_loss=1.317, generator_loss=30.63, generator_mel_loss=22.01, generator_kl_loss=1.953, generator_dur_loss=1.737, generator_adv_loss=2.117, generator_feat_match_loss=2.809, over 2302.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 06:41:51,961 INFO [train.py:811] (3/4) Start epoch 166
2023-11-13 06:45:29,211 INFO [train.py:811] (3/4) Start epoch 167
2023-11-13 06:46:33,464 INFO [train.py:467] (3/4) Epoch 167, batch 8, global_batch_idx: 6150, batch size: 64, loss[discriminator_loss=2.529, discriminator_real_loss=1.313, discriminator_fake_loss=1.216, generator_loss=30.95, generator_mel_loss=22.34, generator_kl_loss=1.948, generator_dur_loss=1.72, generator_adv_loss=2.174, generator_feat_match_loss=2.768, over 64.00 samples.], tot_loss[discriminator_loss=2.653, discriminator_real_loss=1.33, discriminator_fake_loss=1.322, generator_loss=30.68, generator_mel_loss=22.17, generator_kl_loss=1.98, generator_dur_loss=1.736, generator_adv_loss=2.034, generator_feat_match_loss=2.76, over 597.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 06:49:03,082 INFO [train.py:811] (3/4) Start epoch 168
2023-11-13 06:51:24,124 INFO [train.py:467] (3/4) Epoch 168, batch 21, global_batch_idx: 6200, batch size: 73, loss[discriminator_loss=3.021, discriminator_real_loss=1.491, discriminator_fake_loss=1.53, generator_loss=30.21, generator_mel_loss=22.46, generator_kl_loss=1.92, generator_dur_loss=1.716, generator_adv_loss=1.723, generator_feat_match_loss=2.387, over 73.00 samples.], tot_loss[discriminator_loss=2.703, discriminator_real_loss=1.384, discriminator_fake_loss=1.319, generator_loss=31.21, generator_mel_loss=22.28, generator_kl_loss=1.973, generator_dur_loss=1.73, generator_adv_loss=2.194, generator_feat_match_loss=3.035, over 1725.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 06:51:24,125 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 06:51:35,085 INFO [train.py:517] (3/4) Epoch 168, validation: discriminator_loss=2.937, discriminator_real_loss=1.581, discriminator_fake_loss=1.356, generator_loss=30.96, generator_mel_loss=22.99, generator_kl_loss=1.986, generator_dur_loss=1.7, generator_adv_loss=1.922, generator_feat_match_loss=2.362, over 100.00 samples.
2023-11-13 06:51:35,086 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 06:52:50,437 INFO [train.py:811] (3/4) Start epoch 169
2023-11-13 06:56:14,832 INFO [train.py:467] (3/4) Epoch 169, batch 34, global_batch_idx: 6250, batch size: 67, loss[discriminator_loss=2.758, discriminator_real_loss=1.34, discriminator_fake_loss=1.418, generator_loss=30.86, generator_mel_loss=22.15, generator_kl_loss=1.816, generator_dur_loss=1.753, generator_adv_loss=2.221, generator_feat_match_loss=2.914, over 67.00 samples.], tot_loss[discriminator_loss=2.673, discriminator_real_loss=1.348, discriminator_fake_loss=1.324, generator_loss=30.43, generator_mel_loss=22.1, generator_kl_loss=1.965, generator_dur_loss=1.733, generator_adv_loss=1.998, generator_feat_match_loss=2.625, over 2504.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 06:56:24,883 INFO [train.py:811] (3/4) Start epoch 170
2023-11-13 06:59:57,630 INFO [train.py:811] (3/4) Start epoch 171
2023-11-13 07:01:11,273 INFO [train.py:467] (3/4) Epoch 171, batch 10, global_batch_idx: 6300, batch size: 64, loss[discriminator_loss=2.641, discriminator_real_loss=1.374, discriminator_fake_loss=1.268, generator_loss=30.77, generator_mel_loss=21.43, generator_kl_loss=1.906, generator_dur_loss=1.732, generator_adv_loss=2.289, generator_feat_match_loss=3.414, over 64.00 samples.], tot_loss[discriminator_loss=2.688, discriminator_real_loss=1.37, discriminator_fake_loss=1.319, generator_loss=31.29, generator_mel_loss=22.15, generator_kl_loss=1.926, generator_dur_loss=1.734, generator_adv_loss=2.278, generator_feat_match_loss=3.206, over 789.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:03:36,349 INFO [train.py:811] (3/4) Start epoch 172
2023-11-13 07:05:49,547 INFO [train.py:467] (3/4) Epoch 172, batch 23, global_batch_idx: 6350, batch size: 69, loss[discriminator_loss=2.602, discriminator_real_loss=1.365, discriminator_fake_loss=1.236, generator_loss=30.35, generator_mel_loss=21.93, generator_kl_loss=1.914, generator_dur_loss=1.733, generator_adv_loss=2.072, generator_feat_match_loss=2.699, over 69.00 samples.], tot_loss[discriminator_loss=2.665, discriminator_real_loss=1.349, discriminator_fake_loss=1.316, generator_loss=30.62, generator_mel_loss=22.18, generator_kl_loss=1.948, generator_dur_loss=1.735, generator_adv_loss=2.032, generator_feat_match_loss=2.722, over 1710.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:07:06,220 INFO [train.py:811] (3/4) Start epoch 173
2023-11-13 07:10:37,997 INFO [train.py:467] (3/4) Epoch 173, batch 36, global_batch_idx: 6400, batch size: 61, loss[discriminator_loss=2.717, discriminator_real_loss=1.533, discriminator_fake_loss=1.184, generator_loss=30.23, generator_mel_loss=22.08, generator_kl_loss=2.033, generator_dur_loss=1.741, generator_adv_loss=1.891, generator_feat_match_loss=2.486, over 61.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.337, discriminator_fake_loss=1.311, generator_loss=31.14, generator_mel_loss=22.14, generator_kl_loss=1.962, generator_dur_loss=1.728, generator_adv_loss=2.219, generator_feat_match_loss=3.091, over 2808.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 32.0
2023-11-13 07:10:37,999 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 07:10:50,207 INFO [train.py:517] (3/4) Epoch 173, validation: discriminator_loss=2.858, discriminator_real_loss=1.201, discriminator_fake_loss=1.657, generator_loss=30.66, generator_mel_loss=22.74, generator_kl_loss=1.988, generator_dur_loss=1.701, generator_adv_loss=1.712, generator_feat_match_loss=2.527, over 100.00 samples.
2023-11-13 07:10:50,208 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 07:10:50,860 INFO [train.py:811] (3/4) Start epoch 174
2023-11-13 07:14:25,845 INFO [train.py:811] (3/4) Start epoch 175
2023-11-13 07:15:51,287 INFO [train.py:467] (3/4) Epoch 175, batch 12, global_batch_idx: 6450, batch size: 51, loss[discriminator_loss=2.449, discriminator_real_loss=1.165, discriminator_fake_loss=1.285, generator_loss=31.6, generator_mel_loss=22.09, generator_kl_loss=1.937, generator_dur_loss=1.746, generator_adv_loss=2.113, generator_feat_match_loss=3.715, over 51.00 samples.], tot_loss[discriminator_loss=2.744, discriminator_real_loss=1.431, discriminator_fake_loss=1.314, generator_loss=30.97, generator_mel_loss=22.1, generator_kl_loss=1.987, generator_dur_loss=1.728, generator_adv_loss=2.17, generator_feat_match_loss=2.989, over 1035.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:17:58,997 INFO [train.py:811] (3/4) Start epoch 176
2023-11-13 07:20:24,782 INFO [train.py:467] (3/4) Epoch 176, batch 25, global_batch_idx: 6500, batch size: 110, loss[discriminator_loss=2.768, discriminator_real_loss=1.467, discriminator_fake_loss=1.301, generator_loss=30.24, generator_mel_loss=22.18, generator_kl_loss=1.891, generator_dur_loss=1.713, generator_adv_loss=2.025, generator_feat_match_loss=2.426, over 110.00 samples.], tot_loss[discriminator_loss=2.686, discriminator_real_loss=1.358, discriminator_fake_loss=1.329, generator_loss=30.23, generator_mel_loss=22.08, generator_kl_loss=1.952, generator_dur_loss=1.734, generator_adv_loss=1.946, generator_feat_match_loss=2.519, over 1803.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:21:29,752 INFO [train.py:811] (3/4) Start epoch 177
2023-11-13 07:25:01,258 INFO [train.py:811] (3/4) Start epoch 178
2023-11-13 07:25:21,827 INFO [train.py:467] (3/4) Epoch 178, batch 1, global_batch_idx: 6550, batch size: 69, loss[discriminator_loss=2.574, discriminator_real_loss=1.194, discriminator_fake_loss=1.381, generator_loss=31.81, generator_mel_loss=22.66, generator_kl_loss=1.903, generator_dur_loss=1.738, generator_adv_loss=2.469, generator_feat_match_loss=3.039, over 69.00 samples.], tot_loss[discriminator_loss=2.547, discriminator_real_loss=1.181, discriminator_fake_loss=1.367, generator_loss=32.07, generator_mel_loss=22.69, generator_kl_loss=1.931, generator_dur_loss=1.741, generator_adv_loss=2.407, generator_feat_match_loss=3.298, over 129.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:28:34,310 INFO [train.py:811] (3/4) Start epoch 179
2023-11-13 07:30:03,340 INFO [train.py:467] (3/4) Epoch 179, batch 14, global_batch_idx: 6600, batch size: 85, loss[discriminator_loss=3.25, discriminator_real_loss=2.02, discriminator_fake_loss=1.23, generator_loss=30.57, generator_mel_loss=22.09, generator_kl_loss=2.088, generator_dur_loss=1.711, generator_adv_loss=2.367, generator_feat_match_loss=2.316, over 85.00 samples.], tot_loss[discriminator_loss=2.713, discriminator_real_loss=1.395, discriminator_fake_loss=1.318, generator_loss=31.31, generator_mel_loss=22.35, generator_kl_loss=1.974, generator_dur_loss=1.728, generator_adv_loss=2.247, generator_feat_match_loss=3.009, over 1093.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:30:03,341 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 07:30:14,356 INFO [train.py:517] (3/4) Epoch 179, validation: discriminator_loss=2.851, discriminator_real_loss=1.733, discriminator_fake_loss=1.118, generator_loss=31.68, generator_mel_loss=23.22, generator_kl_loss=1.976, generator_dur_loss=1.695, generator_adv_loss=2.28, generator_feat_match_loss=2.509, over 100.00 samples.
2023-11-13 07:30:14,358 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 07:32:14,061 INFO [train.py:811] (3/4) Start epoch 180
2023-11-13 07:35:02,503 INFO [train.py:467] (3/4) Epoch 180, batch 27, global_batch_idx: 6650, batch size: 59, loss[discriminator_loss=2.629, discriminator_real_loss=1.354, discriminator_fake_loss=1.275, generator_loss=30.95, generator_mel_loss=22.1, generator_kl_loss=1.946, generator_dur_loss=1.757, generator_adv_loss=2.152, generator_feat_match_loss=3, over 59.00 samples.], tot_loss[discriminator_loss=2.689, discriminator_real_loss=1.361, discriminator_fake_loss=1.328, generator_loss=30.09, generator_mel_loss=21.96, generator_kl_loss=1.932, generator_dur_loss=1.723, generator_adv_loss=2.006, generator_feat_match_loss=2.475, over 2145.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:35:48,341 INFO [train.py:811] (3/4) Start epoch 181
2023-11-13 07:39:21,121 INFO [train.py:811] (3/4) Start epoch 182
2023-11-13 07:39:52,280 INFO [train.py:467] (3/4) Epoch 182, batch 3, global_batch_idx: 6700, batch size: 65, loss[discriminator_loss=2.918, discriminator_real_loss=1.377, discriminator_fake_loss=1.54, generator_loss=30.48, generator_mel_loss=22.53, generator_kl_loss=1.968, generator_dur_loss=1.728, generator_adv_loss=1.863, generator_feat_match_loss=2.389, over 65.00 samples.], tot_loss[discriminator_loss=2.708, discriminator_real_loss=1.407, discriminator_fake_loss=1.301, generator_loss=31.1, generator_mel_loss=22.06, generator_kl_loss=1.938, generator_dur_loss=1.736, generator_adv_loss=2.293, generator_feat_match_loss=3.08, over 296.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:42:47,424 INFO [train.py:811] (3/4) Start epoch 183
2023-11-13 07:44:30,921 INFO [train.py:467] (3/4) Epoch 183, batch 16, global_batch_idx: 6750, batch size: 51, loss[discriminator_loss=2.594, discriminator_real_loss=1.254, discriminator_fake_loss=1.341, generator_loss=29.65, generator_mel_loss=21.16, generator_kl_loss=2.038, generator_dur_loss=1.693, generator_adv_loss=2.078, generator_feat_match_loss=2.682, over 51.00 samples.], tot_loss[discriminator_loss=2.638, discriminator_real_loss=1.336, discriminator_fake_loss=1.303, generator_loss=30.09, generator_mel_loss=21.85, generator_kl_loss=1.942, generator_dur_loss=1.726, generator_adv_loss=1.969, generator_feat_match_loss=2.597, over 1064.00 samples.], cur_lr_g: 1.96e-04, cur_lr_d: 1.96e-04, grad_scale: 16.0
2023-11-13 07:46:19,870 INFO [train.py:811] (3/4) Start epoch 184
2023-11-13 07:49:18,138 INFO [train.py:467] (3/4) Epoch 184, batch 29, global_batch_idx: 6800, batch size: 85, loss[discriminator_loss=2.654, discriminator_real_loss=1.26, discriminator_fake_loss=1.395, generator_loss=30.11, generator_mel_loss=21.79, generator_kl_loss=1.98, generator_dur_loss=1.73, generator_adv_loss=2.059, generator_feat_match_loss=2.557, over 85.00 samples.], tot_loss[discriminator_loss=2.639, discriminator_real_loss=1.335, discriminator_fake_loss=1.304, generator_loss=31.04, generator_mel_loss=22.07, generator_kl_loss=1.945, generator_dur_loss=1.722, generator_adv_loss=2.19, generator_feat_match_loss=3.112, over 2298.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 32.0
2023-11-13 07:49:18,139 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 07:49:29,713 INFO [train.py:517] (3/4) Epoch 184, validation: discriminator_loss=2.926, discriminator_real_loss=1.293, discriminator_fake_loss=1.633, generator_loss=30.87, generator_mel_loss=23.04, generator_kl_loss=2.013, generator_dur_loss=1.695, generator_adv_loss=1.598, generator_feat_match_loss=2.517, over 100.00 samples.
2023-11-13 07:49:29,714 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 07:50:04,040 INFO [train.py:811] (3/4) Start epoch 185
2023-11-13 07:53:35,970 INFO [train.py:811] (3/4) Start epoch 186
2023-11-13 07:54:17,591 INFO [train.py:467] (3/4) Epoch 186, batch 5, global_batch_idx: 6850, batch size: 50, loss[discriminator_loss=3.027, discriminator_real_loss=1.331, discriminator_fake_loss=1.697, generator_loss=29.57, generator_mel_loss=21.77, generator_kl_loss=1.965, generator_dur_loss=1.715, generator_adv_loss=1.92, generator_feat_match_loss=2.197, over 50.00 samples.], tot_loss[discriminator_loss=2.612, discriminator_real_loss=1.337, discriminator_fake_loss=1.276, generator_loss=30.9, generator_mel_loss=22.02, generator_kl_loss=1.898, generator_dur_loss=1.725, generator_adv_loss=2.13, generator_feat_match_loss=3.126, over 367.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 07:57:03,978 INFO [train.py:811] (3/4) Start epoch 187
2023-11-13 07:58:58,585 INFO [train.py:467] (3/4) Epoch 187, batch 18, global_batch_idx: 6900, batch size: 60, loss[discriminator_loss=2.752, discriminator_real_loss=1.575, discriminator_fake_loss=1.177, generator_loss=31.19, generator_mel_loss=22.12, generator_kl_loss=1.872, generator_dur_loss=1.721, generator_adv_loss=2.422, generator_feat_match_loss=3.051, over 60.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.32, discriminator_fake_loss=1.292, generator_loss=30.75, generator_mel_loss=21.9, generator_kl_loss=1.946, generator_dur_loss=1.722, generator_adv_loss=2.151, generator_feat_match_loss=3.027, over 1439.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:00:37,750 INFO [train.py:811] (3/4) Start epoch 188
2023-11-13 08:03:33,557 INFO [train.py:467] (3/4) Epoch 188, batch 31, global_batch_idx: 6950, batch size: 82, loss[discriminator_loss=2.438, discriminator_real_loss=1.287, discriminator_fake_loss=1.149, generator_loss=32.15, generator_mel_loss=21.53, generator_kl_loss=2.079, generator_dur_loss=1.743, generator_adv_loss=2.57, generator_feat_match_loss=4.223, over 82.00 samples.], tot_loss[discriminator_loss=2.587, discriminator_real_loss=1.316, discriminator_fake_loss=1.272, generator_loss=31, generator_mel_loss=21.91, generator_kl_loss=1.969, generator_dur_loss=1.722, generator_adv_loss=2.239, generator_feat_match_loss=3.159, over 2277.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:04:05,127 INFO [train.py:811] (3/4) Start epoch 189
2023-11-13 08:07:35,029 INFO [train.py:811] (3/4) Start epoch 190
2023-11-13 08:08:35,970 INFO [train.py:467] (3/4) Epoch 190, batch 7, global_batch_idx: 7000, batch size: 69, loss[discriminator_loss=2.629, discriminator_real_loss=1.234, discriminator_fake_loss=1.396, generator_loss=29.84, generator_mel_loss=21.88, generator_kl_loss=1.924, generator_dur_loss=1.725, generator_adv_loss=1.925, generator_feat_match_loss=2.383, over 69.00 samples.], tot_loss[discriminator_loss=2.607, discriminator_real_loss=1.306, discriminator_fake_loss=1.302, generator_loss=30.2, generator_mel_loss=21.8, generator_kl_loss=1.948, generator_dur_loss=1.727, generator_adv_loss=2.015, generator_feat_match_loss=2.716, over 614.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:08:35,971 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 08:08:47,370 INFO [train.py:517] (3/4) Epoch 190, validation: discriminator_loss=2.559, discriminator_real_loss=1.268, discriminator_fake_loss=1.291, generator_loss=31.28, generator_mel_loss=22.84, generator_kl_loss=1.986, generator_dur_loss=1.694, generator_adv_loss=1.96, generator_feat_match_loss=2.801, over 100.00 samples.
2023-11-13 08:08:47,371 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 08:11:22,333 INFO [train.py:811] (3/4) Start epoch 191
2023-11-13 08:13:23,846 INFO [train.py:467] (3/4) Epoch 191, batch 20, global_batch_idx: 7050, batch size: 126, loss[discriminator_loss=2.473, discriminator_real_loss=1.189, discriminator_fake_loss=1.283, generator_loss=31.43, generator_mel_loss=22.17, generator_kl_loss=1.984, generator_dur_loss=1.721, generator_adv_loss=2.223, generator_feat_match_loss=3.336, over 126.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.375, discriminator_fake_loss=1.301, generator_loss=30.63, generator_mel_loss=21.78, generator_kl_loss=1.954, generator_dur_loss=1.722, generator_adv_loss=2.133, generator_feat_match_loss=3.043, over 1641.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:14:52,730 INFO [train.py:811] (3/4) Start epoch 192
2023-11-13 08:18:07,485 INFO [train.py:467] (3/4) Epoch 192, batch 33, global_batch_idx: 7100, batch size: 55, loss[discriminator_loss=2.551, discriminator_real_loss=1.206, discriminator_fake_loss=1.344, generator_loss=31.53, generator_mel_loss=22.54, generator_kl_loss=1.857, generator_dur_loss=1.727, generator_adv_loss=2.211, generator_feat_match_loss=3.197, over 55.00 samples.], tot_loss[discriminator_loss=2.642, discriminator_real_loss=1.336, discriminator_fake_loss=1.306, generator_loss=30.68, generator_mel_loss=22.1, generator_kl_loss=1.958, generator_dur_loss=1.722, generator_adv_loss=2.048, generator_feat_match_loss=2.846, over 2478.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:18:24,274 INFO [train.py:811] (3/4) Start epoch 193
2023-11-13 08:21:46,840 INFO [train.py:811] (3/4) Start epoch 194
2023-11-13 08:22:44,465 INFO [train.py:467] (3/4) Epoch 194, batch 9, global_batch_idx: 7150, batch size: 53, loss[discriminator_loss=2.562, discriminator_real_loss=1.224, discriminator_fake_loss=1.34, generator_loss=31.45, generator_mel_loss=21.56, generator_kl_loss=1.876, generator_dur_loss=1.728, generator_adv_loss=2.719, generator_feat_match_loss=3.566, over 53.00 samples.], tot_loss[discriminator_loss=2.619, discriminator_real_loss=1.322, discriminator_fake_loss=1.296, generator_loss=31.28, generator_mel_loss=21.87, generator_kl_loss=1.968, generator_dur_loss=1.725, generator_adv_loss=2.366, generator_feat_match_loss=3.356, over 567.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:25:12,719 INFO [train.py:811] (3/4) Start epoch 195
2023-11-13 08:27:14,704 INFO [train.py:467] (3/4) Epoch 195, batch 22, global_batch_idx: 7200, batch size: 64, loss[discriminator_loss=2.719, discriminator_real_loss=1.362, discriminator_fake_loss=1.356, generator_loss=30.94, generator_mel_loss=21.92, generator_kl_loss=1.967, generator_dur_loss=1.706, generator_adv_loss=2.242, generator_feat_match_loss=3.104, over 64.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.317, discriminator_fake_loss=1.295, generator_loss=30.6, generator_mel_loss=21.89, generator_kl_loss=1.965, generator_dur_loss=1.72, generator_adv_loss=2.09, generator_feat_match_loss=2.935, over 1589.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 32.0
2023-11-13 08:27:14,706 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 08:27:25,456 INFO [train.py:517] (3/4) Epoch 195, validation: discriminator_loss=2.731, discriminator_real_loss=1.228, discriminator_fake_loss=1.504, generator_loss=31.59, generator_mel_loss=23.37, generator_kl_loss=2.062, generator_dur_loss=1.684, generator_adv_loss=1.696, generator_feat_match_loss=2.776, over 100.00 samples.
2023-11-13 08:27:25,457 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 08:28:48,378 INFO [train.py:811] (3/4) Start epoch 196
2023-11-13 08:32:12,708 INFO [train.py:467] (3/4) Epoch 196, batch 35, global_batch_idx: 7250, batch size: 60, loss[discriminator_loss=2.443, discriminator_real_loss=1.214, discriminator_fake_loss=1.229, generator_loss=31.52, generator_mel_loss=22.08, generator_kl_loss=1.913, generator_dur_loss=1.744, generator_adv_loss=2.322, generator_feat_match_loss=3.467, over 60.00 samples.], tot_loss[discriminator_loss=2.643, discriminator_real_loss=1.351, discriminator_fake_loss=1.292, generator_loss=30.76, generator_mel_loss=21.74, generator_kl_loss=1.959, generator_dur_loss=1.718, generator_adv_loss=2.218, generator_feat_match_loss=3.128, over 2665.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:32:18,323 INFO [train.py:811] (3/4) Start epoch 197
2023-11-13 08:35:48,429 INFO [train.py:811] (3/4) Start epoch 198
2023-11-13 08:37:03,825 INFO [train.py:467] (3/4) Epoch 198, batch 11, global_batch_idx: 7300, batch size: 76, loss[discriminator_loss=2.666, discriminator_real_loss=1.347, discriminator_fake_loss=1.319, generator_loss=30.82, generator_mel_loss=21.83, generator_kl_loss=1.927, generator_dur_loss=1.701, generator_adv_loss=2.201, generator_feat_match_loss=3.164, over 76.00 samples.], tot_loss[discriminator_loss=2.62, discriminator_real_loss=1.314, discriminator_fake_loss=1.306, generator_loss=30.5, generator_mel_loss=21.59, generator_kl_loss=1.94, generator_dur_loss=1.717, generator_adv_loss=2.186, generator_feat_match_loss=3.067, over 848.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:39:16,777 INFO [train.py:811] (3/4) Start epoch 199
2023-11-13 08:41:45,039 INFO [train.py:467] (3/4) Epoch 199, batch 24, global_batch_idx: 7350, batch size: 79, loss[discriminator_loss=2.508, discriminator_real_loss=1.378, discriminator_fake_loss=1.13, generator_loss=30.97, generator_mel_loss=21.67, generator_kl_loss=1.898, generator_dur_loss=1.728, generator_adv_loss=2.369, generator_feat_match_loss=3.309, over 79.00 samples.], tot_loss[discriminator_loss=2.63, discriminator_real_loss=1.337, discriminator_fake_loss=1.294, generator_loss=30.98, generator_mel_loss=22.05, generator_kl_loss=1.944, generator_dur_loss=1.719, generator_adv_loss=2.189, generator_feat_match_loss=3.073, over 1824.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:42:50,815 INFO [train.py:811] (3/4) Start epoch 200
2023-11-13 08:46:22,695 INFO [train.py:811] (3/4) Start epoch 201
2023-11-13 08:46:43,028 INFO [train.py:467] (3/4) Epoch 201, batch 0, global_batch_idx: 7400, batch size: 54, loss[discriminator_loss=2.5, discriminator_real_loss=1.276, discriminator_fake_loss=1.225, generator_loss=30.83, generator_mel_loss=21.7, generator_kl_loss=1.924, generator_dur_loss=1.735, generator_adv_loss=2.182, generator_feat_match_loss=3.297, over 54.00 samples.], tot_loss[discriminator_loss=2.5, discriminator_real_loss=1.276, discriminator_fake_loss=1.225, generator_loss=30.83, generator_mel_loss=21.7, generator_kl_loss=1.924, generator_dur_loss=1.735, generator_adv_loss=2.182, generator_feat_match_loss=3.297, over 54.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:46:43,030 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 08:46:55,221 INFO [train.py:517] (3/4) Epoch 201, validation: discriminator_loss=2.565, discriminator_real_loss=1.226, discriminator_fake_loss=1.339, generator_loss=31.02, generator_mel_loss=22.61, generator_kl_loss=2.034, generator_dur_loss=1.688, generator_adv_loss=1.874, generator_feat_match_loss=2.82, over 100.00 samples.
2023-11-13 08:46:55,222 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 08:50:15,822 INFO [train.py:811] (3/4) Start epoch 202
2023-11-13 08:51:44,148 INFO [train.py:467] (3/4) Epoch 202, batch 13, global_batch_idx: 7450, batch size: 49, loss[discriminator_loss=2.34, discriminator_real_loss=1.244, discriminator_fake_loss=1.096, generator_loss=31.46, generator_mel_loss=20.9, generator_kl_loss=1.991, generator_dur_loss=1.701, generator_adv_loss=2.555, generator_feat_match_loss=4.316, over 49.00 samples.], tot_loss[discriminator_loss=2.498, discriminator_real_loss=1.273, discriminator_fake_loss=1.225, generator_loss=30.93, generator_mel_loss=21.52, generator_kl_loss=1.923, generator_dur_loss=1.719, generator_adv_loss=2.325, generator_feat_match_loss=3.45, over 880.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:53:50,287 INFO [train.py:811] (3/4) Start epoch 203
2023-11-13 08:56:20,634 INFO [train.py:467] (3/4) Epoch 203, batch 26, global_batch_idx: 7500, batch size: 56, loss[discriminator_loss=2.41, discriminator_real_loss=1.195, discriminator_fake_loss=1.215, generator_loss=30.85, generator_mel_loss=21.42, generator_kl_loss=1.93, generator_dur_loss=1.717, generator_adv_loss=2.477, generator_feat_match_loss=3.309, over 56.00 samples.], tot_loss[discriminator_loss=2.449, discriminator_real_loss=1.234, discriminator_fake_loss=1.215, generator_loss=31.13, generator_mel_loss=21.41, generator_kl_loss=1.944, generator_dur_loss=1.715, generator_adv_loss=2.387, generator_feat_match_loss=3.683, over 2010.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 08:57:21,272 INFO [train.py:811] (3/4) Start epoch 204
2023-11-13 09:00:58,044 INFO [train.py:811] (3/4) Start epoch 205
2023-11-13 09:01:25,579 INFO [train.py:467] (3/4) Epoch 205, batch 2, global_batch_idx: 7550, batch size: 101, loss[discriminator_loss=2.25, discriminator_real_loss=1.121, discriminator_fake_loss=1.13, generator_loss=32.31, generator_mel_loss=21.45, generator_kl_loss=1.937, generator_dur_loss=1.705, generator_adv_loss=2.537, generator_feat_match_loss=4.676, over 101.00 samples.], tot_loss[discriminator_loss=2.296, discriminator_real_loss=1.172, discriminator_fake_loss=1.124, generator_loss=32.23, generator_mel_loss=21.64, generator_kl_loss=1.954, generator_dur_loss=1.713, generator_adv_loss=2.481, generator_feat_match_loss=4.449, over 217.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:04:25,180 INFO [train.py:811] (3/4) Start epoch 206
2023-11-13 09:06:00,689 INFO [train.py:467] (3/4) Epoch 206, batch 15, global_batch_idx: 7600, batch size: 110, loss[discriminator_loss=2.348, discriminator_real_loss=1.078, discriminator_fake_loss=1.269, generator_loss=31.86, generator_mel_loss=21.83, generator_kl_loss=1.951, generator_dur_loss=1.706, generator_adv_loss=2.523, generator_feat_match_loss=3.859, over 110.00 samples.], tot_loss[discriminator_loss=2.529, discriminator_real_loss=1.253, discriminator_fake_loss=1.276, generator_loss=30.76, generator_mel_loss=21.39, generator_kl_loss=1.932, generator_dur_loss=1.713, generator_adv_loss=2.327, generator_feat_match_loss=3.406, over 1226.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 32.0
2023-11-13 09:06:00,691 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 09:06:11,530 INFO [train.py:517] (3/4) Epoch 206, validation: discriminator_loss=2.259, discriminator_real_loss=1.189, discriminator_fake_loss=1.07, generator_loss=32.32, generator_mel_loss=22.16, generator_kl_loss=2.046, generator_dur_loss=1.682, generator_adv_loss=2.501, generator_feat_match_loss=3.928, over 100.00 samples.
2023-11-13 09:06:11,531 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 09:08:09,401 INFO [train.py:811] (3/4) Start epoch 207
2023-11-13 09:10:59,138 INFO [train.py:467] (3/4) Epoch 207, batch 28, global_batch_idx: 7650, batch size: 126, loss[discriminator_loss=2.674, discriminator_real_loss=1.395, discriminator_fake_loss=1.279, generator_loss=30.41, generator_mel_loss=22.13, generator_kl_loss=1.933, generator_dur_loss=1.692, generator_adv_loss=2.018, generator_feat_match_loss=2.637, over 126.00 samples.], tot_loss[discriminator_loss=2.621, discriminator_real_loss=1.326, discriminator_fake_loss=1.295, generator_loss=30.73, generator_mel_loss=21.8, generator_kl_loss=1.976, generator_dur_loss=1.712, generator_adv_loss=2.193, generator_feat_match_loss=3.051, over 2235.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:11:41,243 INFO [train.py:811] (3/4) Start epoch 208
2023-11-13 09:15:09,976 INFO [train.py:811] (3/4) Start epoch 209
2023-11-13 09:15:44,332 INFO [train.py:467] (3/4) Epoch 209, batch 4, global_batch_idx: 7700, batch size: 65, loss[discriminator_loss=2.48, discriminator_real_loss=1.246, discriminator_fake_loss=1.235, generator_loss=31.59, generator_mel_loss=22.11, generator_kl_loss=1.948, generator_dur_loss=1.734, generator_adv_loss=2.043, generator_feat_match_loss=3.754, over 65.00 samples.], tot_loss[discriminator_loss=2.644, discriminator_real_loss=1.36, discriminator_fake_loss=1.285, generator_loss=30.83, generator_mel_loss=21.75, generator_kl_loss=1.946, generator_dur_loss=1.732, generator_adv_loss=2.213, generator_feat_match_loss=3.187, over 321.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:18:42,585 INFO [train.py:811] (3/4) Start epoch 210
2023-11-13 09:20:35,022 INFO [train.py:467] (3/4) Epoch 210, batch 17, global_batch_idx: 7750, batch size: 65, loss[discriminator_loss=2.904, discriminator_real_loss=1.743, discriminator_fake_loss=1.161, generator_loss=29.89, generator_mel_loss=21.24, generator_kl_loss=1.945, generator_dur_loss=1.696, generator_adv_loss=2.342, generator_feat_match_loss=2.666, over 65.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.289, discriminator_fake_loss=1.328, generator_loss=30.93, generator_mel_loss=21.64, generator_kl_loss=1.957, generator_dur_loss=1.711, generator_adv_loss=2.257, generator_feat_match_loss=3.366, over 1304.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:22:21,464 INFO [train.py:811] (3/4) Start epoch 211
2023-11-13 09:25:29,362 INFO [train.py:467] (3/4) Epoch 211, batch 30, global_batch_idx: 7800, batch size: 50, loss[discriminator_loss=2.695, discriminator_real_loss=1.491, discriminator_fake_loss=1.203, generator_loss=30.47, generator_mel_loss=21.44, generator_kl_loss=1.937, generator_dur_loss=1.714, generator_adv_loss=2.197, generator_feat_match_loss=3.178, over 50.00 samples.], tot_loss[discriminator_loss=2.654, discriminator_real_loss=1.333, discriminator_fake_loss=1.321, generator_loss=30.55, generator_mel_loss=21.75, generator_kl_loss=1.952, generator_dur_loss=1.711, generator_adv_loss=2.153, generator_feat_match_loss=2.979, over 2199.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:25:29,363 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 09:25:40,565 INFO [train.py:517] (3/4) Epoch 211, validation: discriminator_loss=2.683, discriminator_real_loss=1.262, discriminator_fake_loss=1.421, generator_loss=30.94, generator_mel_loss=22.57, generator_kl_loss=1.999, generator_dur_loss=1.684, generator_adv_loss=1.803, generator_feat_match_loss=2.877, over 100.00 samples.
2023-11-13 09:25:40,566 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 09:26:08,248 INFO [train.py:811] (3/4) Start epoch 212
2023-11-13 09:29:28,963 INFO [train.py:811] (3/4) Start epoch 213
2023-11-13 09:30:20,022 INFO [train.py:467] (3/4) Epoch 213, batch 6, global_batch_idx: 7850, batch size: 76, loss[discriminator_loss=2.414, discriminator_real_loss=1.299, discriminator_fake_loss=1.115, generator_loss=30.84, generator_mel_loss=21.39, generator_kl_loss=1.819, generator_dur_loss=1.723, generator_adv_loss=2.492, generator_feat_match_loss=3.418, over 76.00 samples.], tot_loss[discriminator_loss=2.817, discriminator_real_loss=1.422, discriminator_fake_loss=1.394, generator_loss=30.21, generator_mel_loss=21.39, generator_kl_loss=1.905, generator_dur_loss=1.724, generator_adv_loss=2.202, generator_feat_match_loss=2.986, over 488.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:32:56,718 INFO [train.py:811] (3/4) Start epoch 214
2023-11-13 09:34:53,771 INFO [train.py:467] (3/4) Epoch 214, batch 19, global_batch_idx: 7900, batch size: 154, loss[discriminator_loss=2.309, discriminator_real_loss=1.187, discriminator_fake_loss=1.121, generator_loss=31.72, generator_mel_loss=21.84, generator_kl_loss=1.931, generator_dur_loss=1.694, generator_adv_loss=2.49, generator_feat_match_loss=3.771, over 154.00 samples.], tot_loss[discriminator_loss=2.586, discriminator_real_loss=1.303, discriminator_fake_loss=1.282, generator_loss=30.87, generator_mel_loss=21.58, generator_kl_loss=1.936, generator_dur_loss=1.711, generator_adv_loss=2.286, generator_feat_match_loss=3.354, over 1423.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:36:31,128 INFO [train.py:811] (3/4) Start epoch 215
2023-11-13 09:39:38,460 INFO [train.py:467] (3/4) Epoch 215, batch 32, global_batch_idx: 7950, batch size: 101, loss[discriminator_loss=2.881, discriminator_real_loss=1.302, discriminator_fake_loss=1.579, generator_loss=30.78, generator_mel_loss=21.77, generator_kl_loss=1.914, generator_dur_loss=1.692, generator_adv_loss=2.295, generator_feat_match_loss=3.113, over 101.00 samples.], tot_loss[discriminator_loss=2.614, discriminator_real_loss=1.315, discriminator_fake_loss=1.299, generator_loss=30.77, generator_mel_loss=21.84, generator_kl_loss=1.98, generator_dur_loss=1.712, generator_adv_loss=2.176, generator_feat_match_loss=3.06, over 2406.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 8.0
2023-11-13 09:40:02,135 INFO [train.py:811] (3/4) Start epoch 216
2023-11-13 09:43:37,331 INFO [train.py:811] (3/4) Start epoch 217
2023-11-13 09:44:42,976 INFO [train.py:467] (3/4) Epoch 217, batch 8, global_batch_idx: 8000, batch size: 69, loss[discriminator_loss=2.758, discriminator_real_loss=1.432, discriminator_fake_loss=1.326, generator_loss=31.31, generator_mel_loss=22.3, generator_kl_loss=1.915, generator_dur_loss=1.71, generator_adv_loss=2.355, generator_feat_match_loss=3.035, over 69.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.337, discriminator_fake_loss=1.339, generator_loss=30.69, generator_mel_loss=21.91, generator_kl_loss=1.972, generator_dur_loss=1.701, generator_adv_loss=2.137, generator_feat_match_loss=2.971, over 854.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:44:42,977 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 09:44:53,747 INFO [train.py:517] (3/4) Epoch 217, validation: discriminator_loss=2.619, discriminator_real_loss=1.454, discriminator_fake_loss=1.165, generator_loss=31.28, generator_mel_loss=22.22, generator_kl_loss=1.973, generator_dur_loss=1.683, generator_adv_loss=2.305, generator_feat_match_loss=3.1, over 100.00 samples.
2023-11-13 09:44:53,748 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 09:47:21,233 INFO [train.py:811] (3/4) Start epoch 218
2023-11-13 09:49:24,594 INFO [train.py:467] (3/4) Epoch 218, batch 21, global_batch_idx: 8050, batch size: 85, loss[discriminator_loss=2.793, discriminator_real_loss=1.651, discriminator_fake_loss=1.141, generator_loss=30.73, generator_mel_loss=21.59, generator_kl_loss=1.971, generator_dur_loss=1.709, generator_adv_loss=2.246, generator_feat_match_loss=3.217, over 85.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.287, discriminator_fake_loss=1.277, generator_loss=30.9, generator_mel_loss=21.47, generator_kl_loss=1.944, generator_dur_loss=1.71, generator_adv_loss=2.307, generator_feat_match_loss=3.463, over 1544.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:50:44,427 INFO [train.py:811] (3/4) Start epoch 219
2023-11-13 09:54:06,213 INFO [train.py:467] (3/4) Epoch 219, batch 34, global_batch_idx: 8100, batch size: 69, loss[discriminator_loss=2.324, discriminator_real_loss=1.117, discriminator_fake_loss=1.206, generator_loss=31.54, generator_mel_loss=21.31, generator_kl_loss=1.917, generator_dur_loss=1.7, generator_adv_loss=2.609, generator_feat_match_loss=4.004, over 69.00 samples.], tot_loss[discriminator_loss=2.566, discriminator_real_loss=1.296, discriminator_fake_loss=1.269, generator_loss=30.94, generator_mel_loss=21.72, generator_kl_loss=1.956, generator_dur_loss=1.713, generator_adv_loss=2.238, generator_feat_match_loss=3.309, over 2412.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 09:54:15,095 INFO [train.py:811] (3/4) Start epoch 220
2023-11-13 09:57:43,858 INFO [train.py:811] (3/4) Start epoch 221
2023-11-13 09:58:59,670 INFO [train.py:467] (3/4) Epoch 221, batch 10, global_batch_idx: 8150, batch size: 55, loss[discriminator_loss=2.668, discriminator_real_loss=1.388, discriminator_fake_loss=1.28, generator_loss=29.78, generator_mel_loss=21.04, generator_kl_loss=1.999, generator_dur_loss=1.748, generator_adv_loss=2.152, generator_feat_match_loss=2.836, over 55.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.268, discriminator_fake_loss=1.354, generator_loss=30.6, generator_mel_loss=21.44, generator_kl_loss=1.956, generator_dur_loss=1.705, generator_adv_loss=2.261, generator_feat_match_loss=3.243, over 906.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 10:01:15,667 INFO [train.py:811] (3/4) Start epoch 222
2023-11-13 10:03:37,499 INFO [train.py:467] (3/4) Epoch 222, batch 23, global_batch_idx: 8200, batch size: 52, loss[discriminator_loss=2.9, discriminator_real_loss=1.441, discriminator_fake_loss=1.459, generator_loss=30.43, generator_mel_loss=22.16, generator_kl_loss=1.873, generator_dur_loss=1.718, generator_adv_loss=2.004, generator_feat_match_loss=2.684, over 52.00 samples.], tot_loss[discriminator_loss=2.633, discriminator_real_loss=1.311, discriminator_fake_loss=1.323, generator_loss=31.16, generator_mel_loss=21.97, generator_kl_loss=1.973, generator_dur_loss=1.709, generator_adv_loss=2.265, generator_feat_match_loss=3.245, over 1696.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 10:03:37,500 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 10:03:48,042 INFO [train.py:517] (3/4) Epoch 222, validation: discriminator_loss=2.788, discriminator_real_loss=1.489, discriminator_fake_loss=1.298, generator_loss=31.44, generator_mel_loss=22.81, generator_kl_loss=1.963, generator_dur_loss=1.681, generator_adv_loss=2.115, generator_feat_match_loss=2.867, over 100.00 samples.
2023-11-13 10:03:48,043 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 10:05:01,389 INFO [train.py:811] (3/4) Start epoch 223
2023-11-13 10:08:33,702 INFO [train.py:467] (3/4) Epoch 223, batch 36, global_batch_idx: 8250, batch size: 126, loss[discriminator_loss=2.328, discriminator_real_loss=1.131, discriminator_fake_loss=1.197, generator_loss=31.97, generator_mel_loss=21.71, generator_kl_loss=1.907, generator_dur_loss=1.695, generator_adv_loss=2.484, generator_feat_match_loss=4.168, over 126.00 samples.], tot_loss[discriminator_loss=2.62, discriminator_real_loss=1.327, discriminator_fake_loss=1.293, generator_loss=30.6, generator_mel_loss=21.64, generator_kl_loss=1.923, generator_dur_loss=1.707, generator_adv_loss=2.182, generator_feat_match_loss=3.15, over 2519.00 samples.], cur_lr_g: 1.95e-04, cur_lr_d: 1.95e-04, grad_scale: 16.0
2023-11-13 10:08:34,387 INFO [train.py:811] (3/4) Start epoch 224
2023-11-13 10:12:11,100 INFO [train.py:811] (3/4) Start epoch 225
2023-11-13 10:13:30,030 INFO [train.py:467] (3/4) Epoch 225, batch 12, global_batch_idx: 8300, batch size: 126, loss[discriminator_loss=2.668, discriminator_real_loss=1.23, discriminator_fake_loss=1.438, generator_loss=29.97, generator_mel_loss=21.69, generator_kl_loss=1.978, generator_dur_loss=1.676, generator_adv_loss=1.909, generator_feat_match_loss=2.715, over 126.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.294, discriminator_fake_loss=1.322, generator_loss=30.12, generator_mel_loss=21.52, generator_kl_loss=1.953, generator_dur_loss=1.7, generator_adv_loss=2.067, generator_feat_match_loss=2.874, over 943.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:15:42,939 INFO [train.py:811] (3/4) Start epoch 226
2023-11-13 10:18:17,269 INFO [train.py:467] (3/4) Epoch 226, batch 25, global_batch_idx: 8350, batch size: 60, loss[discriminator_loss=2.65, discriminator_real_loss=1.242, discriminator_fake_loss=1.408, generator_loss=30.06, generator_mel_loss=21.22, generator_kl_loss=2.048, generator_dur_loss=1.701, generator_adv_loss=2.236, generator_feat_match_loss=2.85, over 60.00 samples.], tot_loss[discriminator_loss=2.601, discriminator_real_loss=1.316, discriminator_fake_loss=1.284, generator_loss=30.83, generator_mel_loss=21.83, generator_kl_loss=1.961, generator_dur_loss=1.707, generator_adv_loss=2.185, generator_feat_match_loss=3.15, over 1888.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:19:17,376 INFO [train.py:811] (3/4) Start epoch 227
2023-11-13 10:22:52,814 INFO [train.py:811] (3/4) Start epoch 228
2023-11-13 10:23:13,934 INFO [train.py:467] (3/4) Epoch 228, batch 1, global_batch_idx: 8400, batch size: 52, loss[discriminator_loss=2.617, discriminator_real_loss=1.274, discriminator_fake_loss=1.343, generator_loss=29.56, generator_mel_loss=21.02, generator_kl_loss=1.935, generator_dur_loss=1.731, generator_adv_loss=2.047, generator_feat_match_loss=2.82, over 52.00 samples.], tot_loss[discriminator_loss=2.621, discriminator_real_loss=1.359, discriminator_fake_loss=1.262, generator_loss=29.77, generator_mel_loss=21.11, generator_kl_loss=1.897, generator_dur_loss=1.712, generator_adv_loss=2.134, generator_feat_match_loss=2.918, over 108.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 32.0
2023-11-13 10:23:13,936 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 10:23:25,529 INFO [train.py:517] (3/4) Epoch 228, validation: discriminator_loss=2.709, discriminator_real_loss=1.193, discriminator_fake_loss=1.516, generator_loss=29.8, generator_mel_loss=22.04, generator_kl_loss=1.973, generator_dur_loss=1.681, generator_adv_loss=1.645, generator_feat_match_loss=2.456, over 100.00 samples.
2023-11-13 10:23:25,530 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27148MB
2023-11-13 10:26:38,247 INFO [train.py:811] (3/4) Start epoch 229
2023-11-13 10:27:57,100 INFO [train.py:467] (3/4) Epoch 229, batch 14, global_batch_idx: 8450, batch size: 67, loss[discriminator_loss=2.318, discriminator_real_loss=1.178, discriminator_fake_loss=1.141, generator_loss=31.94, generator_mel_loss=21.32, generator_kl_loss=1.983, generator_dur_loss=1.74, generator_adv_loss=2.553, generator_feat_match_loss=4.344, over 67.00 samples.], tot_loss[discriminator_loss=2.501, discriminator_real_loss=1.266, discriminator_fake_loss=1.235, generator_loss=31.37, generator_mel_loss=21.86, generator_kl_loss=1.955, generator_dur_loss=1.712, generator_adv_loss=2.293, generator_feat_match_loss=3.553, over 1008.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:30:04,560 INFO [train.py:811] (3/4) Start epoch 230
2023-11-13 10:32:46,517 INFO [train.py:467] (3/4) Epoch 230, batch 27, global_batch_idx: 8500, batch size: 71, loss[discriminator_loss=2.535, discriminator_real_loss=1.316, discriminator_fake_loss=1.22, generator_loss=31.39, generator_mel_loss=21.4, generator_kl_loss=1.987, generator_dur_loss=1.695, generator_adv_loss=2.559, generator_feat_match_loss=3.746, over 71.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.309, discriminator_fake_loss=1.282, generator_loss=30.89, generator_mel_loss=21.74, generator_kl_loss=1.965, generator_dur_loss=1.701, generator_adv_loss=2.213, generator_feat_match_loss=3.261, over 2163.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:33:39,018 INFO [train.py:811] (3/4) Start epoch 231
2023-11-13 10:37:09,218 INFO [train.py:811] (3/4) Start epoch 232
2023-11-13 10:37:40,333 INFO [train.py:467] (3/4) Epoch 232, batch 3, global_batch_idx: 8550, batch size: 52, loss[discriminator_loss=2.715, discriminator_real_loss=1.426, discriminator_fake_loss=1.289, generator_loss=29.19, generator_mel_loss=21.05, generator_kl_loss=1.896, generator_dur_loss=1.692, generator_adv_loss=1.952, generator_feat_match_loss=2.609, over 52.00 samples.], tot_loss[discriminator_loss=2.682, discriminator_real_loss=1.367, discriminator_fake_loss=1.314, generator_loss=30.06, generator_mel_loss=21.72, generator_kl_loss=1.919, generator_dur_loss=1.704, generator_adv_loss=1.976, generator_feat_match_loss=2.745, over 237.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:40:43,458 INFO [train.py:811] (3/4) Start epoch 233
2023-11-13 10:42:29,903 INFO [train.py:467] (3/4) Epoch 233, batch 16, global_batch_idx: 8600, batch size: 49, loss[discriminator_loss=2.361, discriminator_real_loss=1.271, discriminator_fake_loss=1.09, generator_loss=31.95, generator_mel_loss=21.62, generator_kl_loss=1.922, generator_dur_loss=1.71, generator_adv_loss=2.852, generator_feat_match_loss=3.848, over 49.00 samples.], tot_loss[discriminator_loss=2.738, discriminator_real_loss=1.499, discriminator_fake_loss=1.24, generator_loss=31.6, generator_mel_loss=21.58, generator_kl_loss=1.974, generator_dur_loss=1.705, generator_adv_loss=2.545, generator_feat_match_loss=3.802, over 1299.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:42:29,905 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 10:42:41,018 INFO [train.py:517] (3/4) Epoch 233, validation: discriminator_loss=2.891, discriminator_real_loss=1.424, discriminator_fake_loss=1.467, generator_loss=33.18, generator_mel_loss=22.47, generator_kl_loss=2.009, generator_dur_loss=1.672, generator_adv_loss=2.831, generator_feat_match_loss=4.196, over 100.00 samples.
2023-11-13 10:42:41,019 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27155MB
2023-11-13 10:44:23,092 INFO [train.py:811] (3/4) Start epoch 234
2023-11-13 10:47:18,925 INFO [train.py:467] (3/4) Epoch 234, batch 29, global_batch_idx: 8650, batch size: 58, loss[discriminator_loss=2.77, discriminator_real_loss=1.449, discriminator_fake_loss=1.321, generator_loss=29.72, generator_mel_loss=20.85, generator_kl_loss=1.944, generator_dur_loss=1.695, generator_adv_loss=2.305, generator_feat_match_loss=2.922, over 58.00 samples.], tot_loss[discriminator_loss=2.771, discriminator_real_loss=1.414, discriminator_fake_loss=1.357, generator_loss=29.47, generator_mel_loss=21.34, generator_kl_loss=1.931, generator_dur_loss=1.705, generator_adv_loss=1.954, generator_feat_match_loss=2.536, over 2221.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:47:57,096 INFO [train.py:811] (3/4) Start epoch 235
2023-11-13 10:51:26,040 INFO [train.py:811] (3/4) Start epoch 236
2023-11-13 10:52:11,290 INFO [train.py:467] (3/4) Epoch 236, batch 5, global_batch_idx: 8700, batch size: 67, loss[discriminator_loss=2.666, discriminator_real_loss=1.261, discriminator_fake_loss=1.405, generator_loss=29.45, generator_mel_loss=21, generator_kl_loss=1.938, generator_dur_loss=1.727, generator_adv_loss=1.971, generator_feat_match_loss=2.812, over 67.00 samples.], tot_loss[discriminator_loss=2.74, discriminator_real_loss=1.388, discriminator_fake_loss=1.351, generator_loss=29.82, generator_mel_loss=21.41, generator_kl_loss=1.981, generator_dur_loss=1.713, generator_adv_loss=1.985, generator_feat_match_loss=2.726, over 421.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:54:56,134 INFO [train.py:811] (3/4) Start epoch 237
2023-11-13 10:56:45,522 INFO [train.py:467] (3/4) Epoch 237, batch 18, global_batch_idx: 8750, batch size: 85, loss[discriminator_loss=2.715, discriminator_real_loss=1.169, discriminator_fake_loss=1.545, generator_loss=29.95, generator_mel_loss=21.36, generator_kl_loss=1.935, generator_dur_loss=1.701, generator_adv_loss=2.199, generator_feat_match_loss=2.75, over 85.00 samples.], tot_loss[discriminator_loss=2.683, discriminator_real_loss=1.34, discriminator_fake_loss=1.343, generator_loss=30.46, generator_mel_loss=21.78, generator_kl_loss=1.939, generator_dur_loss=1.698, generator_adv_loss=2.074, generator_feat_match_loss=2.974, over 1431.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 10:58:28,238 INFO [train.py:811] (3/4) Start epoch 238
2023-11-13 11:01:28,584 INFO [train.py:467] (3/4) Epoch 238, batch 31, global_batch_idx: 8800, batch size: 69, loss[discriminator_loss=2.715, discriminator_real_loss=1.39, discriminator_fake_loss=1.325, generator_loss=30.33, generator_mel_loss=21.78, generator_kl_loss=1.836, generator_dur_loss=1.698, generator_adv_loss=2.064, generator_feat_match_loss=2.953, over 69.00 samples.], tot_loss[discriminator_loss=2.668, discriminator_real_loss=1.351, discriminator_fake_loss=1.318, generator_loss=30.29, generator_mel_loss=21.7, generator_kl_loss=1.957, generator_dur_loss=1.707, generator_adv_loss=2.025, generator_feat_match_loss=2.908, over 2435.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 32.0
2023-11-13 11:01:28,586 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 11:01:39,345 INFO [train.py:517] (3/4) Epoch 238, validation: discriminator_loss=2.529, discriminator_real_loss=1.238, discriminator_fake_loss=1.29, generator_loss=31.39, generator_mel_loss=22.59, generator_kl_loss=1.976, generator_dur_loss=1.667, generator_adv_loss=1.969, generator_feat_match_loss=3.19, over 100.00 samples.
2023-11-13 11:01:39,346 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27155MB
2023-11-13 11:02:12,004 INFO [train.py:811] (3/4) Start epoch 239
2023-11-13 11:05:43,866 INFO [train.py:811] (3/4) Start epoch 240
2023-11-13 11:06:41,772 INFO [train.py:467] (3/4) Epoch 240, batch 7, global_batch_idx: 8850, batch size: 71, loss[discriminator_loss=2.684, discriminator_real_loss=1.141, discriminator_fake_loss=1.542, generator_loss=30.65, generator_mel_loss=21.51, generator_kl_loss=1.821, generator_dur_loss=1.684, generator_adv_loss=2.324, generator_feat_match_loss=3.312, over 71.00 samples.], tot_loss[discriminator_loss=2.695, discriminator_real_loss=1.349, discriminator_fake_loss=1.346, generator_loss=30.72, generator_mel_loss=21.88, generator_kl_loss=1.967, generator_dur_loss=1.698, generator_adv_loss=2.152, generator_feat_match_loss=3.026, over 707.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:09:14,847 INFO [train.py:811] (3/4) Start epoch 241
2023-11-13 11:11:24,376 INFO [train.py:467] (3/4) Epoch 241, batch 20, global_batch_idx: 8900, batch size: 95, loss[discriminator_loss=2.865, discriminator_real_loss=1.619, discriminator_fake_loss=1.246, generator_loss=30.22, generator_mel_loss=21.97, generator_kl_loss=2.066, generator_dur_loss=1.69, generator_adv_loss=1.805, generator_feat_match_loss=2.697, over 95.00 samples.], tot_loss[discriminator_loss=2.718, discriminator_real_loss=1.384, discriminator_fake_loss=1.334, generator_loss=30.33, generator_mel_loss=21.47, generator_kl_loss=1.936, generator_dur_loss=1.699, generator_adv_loss=2.134, generator_feat_match_loss=3.086, over 1635.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:12:43,720 INFO [train.py:811] (3/4) Start epoch 242
2023-11-13 11:15:59,081 INFO [train.py:467] (3/4) Epoch 242, batch 33, global_batch_idx: 8950, batch size: 67, loss[discriminator_loss=2.73, discriminator_real_loss=1.248, discriminator_fake_loss=1.483, generator_loss=30.58, generator_mel_loss=21.83, generator_kl_loss=1.956, generator_dur_loss=1.709, generator_adv_loss=2.102, generator_feat_match_loss=2.986, over 67.00 samples.], tot_loss[discriminator_loss=2.678, discriminator_real_loss=1.365, discriminator_fake_loss=1.313, generator_loss=30.39, generator_mel_loss=21.82, generator_kl_loss=1.956, generator_dur_loss=1.698, generator_adv_loss=2.012, generator_feat_match_loss=2.902, over 2745.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:16:14,372 INFO [train.py:811] (3/4) Start epoch 243
2023-11-13 11:19:44,042 INFO [train.py:811] (3/4) Start epoch 244
2023-11-13 11:20:50,566 INFO [train.py:467] (3/4) Epoch 244, batch 9, global_batch_idx: 9000, batch size: 101, loss[discriminator_loss=2.656, discriminator_real_loss=1.467, discriminator_fake_loss=1.188, generator_loss=30.34, generator_mel_loss=21.6, generator_kl_loss=2.024, generator_dur_loss=1.711, generator_adv_loss=2.107, generator_feat_match_loss=2.898, over 101.00 samples.], tot_loss[discriminator_loss=2.708, discriminator_real_loss=1.426, discriminator_fake_loss=1.282, generator_loss=30.32, generator_mel_loss=21.62, generator_kl_loss=1.941, generator_dur_loss=1.701, generator_adv_loss=2.157, generator_feat_match_loss=2.902, over 807.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:20:50,567 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 11:21:01,415 INFO [train.py:517] (3/4) Epoch 244, validation: discriminator_loss=2.725, discriminator_real_loss=1.361, discriminator_fake_loss=1.365, generator_loss=30.89, generator_mel_loss=22.52, generator_kl_loss=2.073, generator_dur_loss=1.669, generator_adv_loss=1.862, generator_feat_match_loss=2.76, over 100.00 samples.
2023-11-13 11:21:01,417 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27155MB
2023-11-13 11:23:20,651 INFO [train.py:811] (3/4) Start epoch 245
2023-11-13 11:25:37,307 INFO [train.py:467] (3/4) Epoch 245, batch 22, global_batch_idx: 9050, batch size: 52, loss[discriminator_loss=2.656, discriminator_real_loss=1.268, discriminator_fake_loss=1.388, generator_loss=30.47, generator_mel_loss=22.03, generator_kl_loss=1.966, generator_dur_loss=1.7, generator_adv_loss=1.988, generator_feat_match_loss=2.789, over 52.00 samples.], tot_loss[discriminator_loss=2.67, discriminator_real_loss=1.337, discriminator_fake_loss=1.332, generator_loss=30.38, generator_mel_loss=21.74, generator_kl_loss=1.935, generator_dur_loss=1.698, generator_adv_loss=2.049, generator_feat_match_loss=2.956, over 1805.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:26:54,845 INFO [train.py:811] (3/4) Start epoch 246
2023-11-13 11:30:22,502 INFO [train.py:467] (3/4) Epoch 246, batch 35, global_batch_idx: 9100, batch size: 81, loss[discriminator_loss=2.607, discriminator_real_loss=1.485, discriminator_fake_loss=1.122, generator_loss=31.83, generator_mel_loss=22.04, generator_kl_loss=1.92, generator_dur_loss=1.682, generator_adv_loss=2.377, generator_feat_match_loss=3.811, over 81.00 samples.], tot_loss[discriminator_loss=2.653, discriminator_real_loss=1.351, discriminator_fake_loss=1.302, generator_loss=30.7, generator_mel_loss=21.6, generator_kl_loss=1.931, generator_dur_loss=1.7, generator_adv_loss=2.204, generator_feat_match_loss=3.268, over 2728.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:30:31,315 INFO [train.py:811] (3/4) Start epoch 247
2023-11-13 11:34:00,833 INFO [train.py:811] (3/4) Start epoch 248
2023-11-13 11:35:20,499 INFO [train.py:467] (3/4) Epoch 248, batch 11, global_batch_idx: 9150, batch size: 126, loss[discriminator_loss=2.891, discriminator_real_loss=1.179, discriminator_fake_loss=1.713, generator_loss=29.98, generator_mel_loss=21.63, generator_kl_loss=2.07, generator_dur_loss=1.717, generator_adv_loss=1.807, generator_feat_match_loss=2.764, over 126.00 samples.], tot_loss[discriminator_loss=2.605, discriminator_real_loss=1.274, discriminator_fake_loss=1.332, generator_loss=30.82, generator_mel_loss=21.64, generator_kl_loss=1.991, generator_dur_loss=1.706, generator_adv_loss=2.173, generator_feat_match_loss=3.311, over 1027.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:37:37,095 INFO [train.py:811] (3/4) Start epoch 249
2023-11-13 11:40:01,578 INFO [train.py:467] (3/4) Epoch 249, batch 24, global_batch_idx: 9200, batch size: 153, loss[discriminator_loss=2.74, discriminator_real_loss=1.434, discriminator_fake_loss=1.307, generator_loss=30.38, generator_mel_loss=21.54, generator_kl_loss=1.912, generator_dur_loss=1.68, generator_adv_loss=2.193, generator_feat_match_loss=3.062, over 153.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.337, discriminator_fake_loss=1.28, generator_loss=30.52, generator_mel_loss=21.42, generator_kl_loss=1.937, generator_dur_loss=1.699, generator_adv_loss=2.189, generator_feat_match_loss=3.273, over 2006.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 32.0
2023-11-13 11:40:01,579 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 11:40:12,655 INFO [train.py:517] (3/4) Epoch 249, validation: discriminator_loss=2.564, discriminator_real_loss=1.379, discriminator_fake_loss=1.185, generator_loss=31.21, generator_mel_loss=22.41, generator_kl_loss=1.995, generator_dur_loss=1.668, generator_adv_loss=2.13, generator_feat_match_loss=3.014, over 100.00 samples.
2023-11-13 11:40:12,657 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27296MB
2023-11-13 11:41:14,928 INFO [train.py:811] (3/4) Start epoch 250
2023-11-13 11:44:43,341 INFO [train.py:811] (3/4) Start epoch 251
2023-11-13 11:44:59,111 INFO [train.py:467] (3/4) Epoch 251, batch 0, global_batch_idx: 9250, batch size: 90, loss[discriminator_loss=2.516, discriminator_real_loss=1.482, discriminator_fake_loss=1.034, generator_loss=31.23, generator_mel_loss=21.44, generator_kl_loss=1.925, generator_dur_loss=1.696, generator_adv_loss=2.395, generator_feat_match_loss=3.779, over 90.00 samples.], tot_loss[discriminator_loss=2.516, discriminator_real_loss=1.482, discriminator_fake_loss=1.034, generator_loss=31.23, generator_mel_loss=21.44, generator_kl_loss=1.925, generator_dur_loss=1.696, generator_adv_loss=2.395, generator_feat_match_loss=3.779, over 90.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:48:15,340 INFO [train.py:811] (3/4) Start epoch 252
2023-11-13 11:49:38,868 INFO [train.py:467] (3/4) Epoch 252, batch 13, global_batch_idx: 9300, batch size: 65, loss[discriminator_loss=2.504, discriminator_real_loss=1.304, discriminator_fake_loss=1.199, generator_loss=31.54, generator_mel_loss=21.87, generator_kl_loss=1.892, generator_dur_loss=1.693, generator_adv_loss=2.262, generator_feat_match_loss=3.82, over 65.00 samples.], tot_loss[discriminator_loss=2.631, discriminator_real_loss=1.338, discriminator_fake_loss=1.293, generator_loss=30.38, generator_mel_loss=21.63, generator_kl_loss=1.965, generator_dur_loss=1.707, generator_adv_loss=2.097, generator_feat_match_loss=2.986, over 1035.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:51:46,598 INFO [train.py:811] (3/4) Start epoch 253
2023-11-13 11:54:37,189 INFO [train.py:467] (3/4) Epoch 253, batch 26, global_batch_idx: 9350, batch size: 110, loss[discriminator_loss=2.828, discriminator_real_loss=1.235, discriminator_fake_loss=1.593, generator_loss=29.72, generator_mel_loss=21.24, generator_kl_loss=1.856, generator_dur_loss=1.676, generator_adv_loss=2.246, generator_feat_match_loss=2.705, over 110.00 samples.], tot_loss[discriminator_loss=2.589, discriminator_real_loss=1.313, discriminator_fake_loss=1.276, generator_loss=30.97, generator_mel_loss=21.54, generator_kl_loss=1.909, generator_dur_loss=1.699, generator_adv_loss=2.338, generator_feat_match_loss=3.486, over 2038.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:55:23,987 INFO [train.py:811] (3/4) Start epoch 254
2023-11-13 11:58:55,771 INFO [train.py:811] (3/4) Start epoch 255
2023-11-13 11:59:23,512 INFO [train.py:467] (3/4) Epoch 255, batch 2, global_batch_idx: 9400, batch size: 73, loss[discriminator_loss=2.754, discriminator_real_loss=1.35, discriminator_fake_loss=1.405, generator_loss=30.11, generator_mel_loss=21.79, generator_kl_loss=1.995, generator_dur_loss=1.674, generator_adv_loss=1.882, generator_feat_match_loss=2.773, over 73.00 samples.], tot_loss[discriminator_loss=2.604, discriminator_real_loss=1.292, discriminator_fake_loss=1.312, generator_loss=30.72, generator_mel_loss=21.76, generator_kl_loss=2, generator_dur_loss=1.689, generator_adv_loss=2.106, generator_feat_match_loss=3.16, over 293.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 11:59:23,512 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 11:59:35,608 INFO [train.py:517] (3/4) Epoch 255, validation: discriminator_loss=2.662, discriminator_real_loss=1.208, discriminator_fake_loss=1.454, generator_loss=31.34, generator_mel_loss=22.31, generator_kl_loss=2.044, generator_dur_loss=1.678, generator_adv_loss=2.136, generator_feat_match_loss=3.177, over 100.00 samples.
2023-11-13 11:59:35,609 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 12:02:43,972 INFO [train.py:811] (3/4) Start epoch 256
2023-11-13 12:04:14,534 INFO [train.py:467] (3/4) Epoch 256, batch 15, global_batch_idx: 9450, batch size: 56, loss[discriminator_loss=2.596, discriminator_real_loss=1.263, discriminator_fake_loss=1.333, generator_loss=30.39, generator_mel_loss=21.17, generator_kl_loss=2.017, generator_dur_loss=1.691, generator_adv_loss=2.277, generator_feat_match_loss=3.234, over 56.00 samples.], tot_loss[discriminator_loss=2.695, discriminator_real_loss=1.364, discriminator_fake_loss=1.331, generator_loss=30.34, generator_mel_loss=21.59, generator_kl_loss=1.979, generator_dur_loss=1.704, generator_adv_loss=2.074, generator_feat_match_loss=2.984, over 1006.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:06:17,412 INFO [train.py:811] (3/4) Start epoch 257
2023-11-13 12:09:11,877 INFO [train.py:467] (3/4) Epoch 257, batch 28, global_batch_idx: 9500, batch size: 110, loss[discriminator_loss=2.539, discriminator_real_loss=1.243, discriminator_fake_loss=1.295, generator_loss=31.07, generator_mel_loss=21.58, generator_kl_loss=1.98, generator_dur_loss=1.677, generator_adv_loss=2.117, generator_feat_match_loss=3.723, over 110.00 samples.], tot_loss[discriminator_loss=2.672, discriminator_real_loss=1.358, discriminator_fake_loss=1.313, generator_loss=30.73, generator_mel_loss=21.69, generator_kl_loss=1.943, generator_dur_loss=1.696, generator_adv_loss=2.176, generator_feat_match_loss=3.223, over 2300.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:09:52,354 INFO [train.py:811] (3/4) Start epoch 258
2023-11-13 12:13:23,098 INFO [train.py:811] (3/4) Start epoch 259
2023-11-13 12:14:01,632 INFO [train.py:467] (3/4) Epoch 259, batch 4, global_batch_idx: 9550, batch size: 53, loss[discriminator_loss=2.582, discriminator_real_loss=1.303, discriminator_fake_loss=1.279, generator_loss=30.25, generator_mel_loss=20.76, generator_kl_loss=1.839, generator_dur_loss=1.718, generator_adv_loss=2.402, generator_feat_match_loss=3.535, over 53.00 samples.], tot_loss[discriminator_loss=2.681, discriminator_real_loss=1.277, discriminator_fake_loss=1.403, generator_loss=30, generator_mel_loss=21.08, generator_kl_loss=1.906, generator_dur_loss=1.698, generator_adv_loss=2.1, generator_feat_match_loss=3.216, over 355.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:16:53,760 INFO [train.py:811] (3/4) Start epoch 260
2023-11-13 12:18:39,481 INFO [train.py:467] (3/4) Epoch 260, batch 17, global_batch_idx: 9600, batch size: 53, loss[discriminator_loss=2.848, discriminator_real_loss=1.539, discriminator_fake_loss=1.31, generator_loss=29.54, generator_mel_loss=21.51, generator_kl_loss=1.925, generator_dur_loss=1.691, generator_adv_loss=1.812, generator_feat_match_loss=2.604, over 53.00 samples.], tot_loss[discriminator_loss=2.647, discriminator_real_loss=1.329, discriminator_fake_loss=1.318, generator_loss=30.56, generator_mel_loss=21.83, generator_kl_loss=1.936, generator_dur_loss=1.692, generator_adv_loss=2.063, generator_feat_match_loss=3.037, over 1468.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 32.0
2023-11-13 12:18:39,483 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 12:18:51,923 INFO [train.py:517] (3/4) Epoch 260, validation: discriminator_loss=2.6, discriminator_real_loss=1.279, discriminator_fake_loss=1.322, generator_loss=31.63, generator_mel_loss=22.92, generator_kl_loss=1.995, generator_dur_loss=1.666, generator_adv_loss=1.953, generator_feat_match_loss=3.099, over 100.00 samples.
2023-11-13 12:18:51,924 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 12:20:37,779 INFO [train.py:811] (3/4) Start epoch 261
2023-11-13 12:23:39,486 INFO [train.py:467] (3/4) Epoch 261, batch 30, global_batch_idx: 9650, batch size: 73, loss[discriminator_loss=3.059, discriminator_real_loss=1.474, discriminator_fake_loss=1.586, generator_loss=29.24, generator_mel_loss=21.07, generator_kl_loss=1.914, generator_dur_loss=1.701, generator_adv_loss=1.85, generator_feat_match_loss=2.713, over 73.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.321, discriminator_fake_loss=1.303, generator_loss=30.75, generator_mel_loss=21.45, generator_kl_loss=1.947, generator_dur_loss=1.703, generator_adv_loss=2.257, generator_feat_match_loss=3.397, over 2150.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:24:11,385 INFO [train.py:811] (3/4) Start epoch 262
2023-11-13 12:27:37,967 INFO [train.py:811] (3/4) Start epoch 263
2023-11-13 12:28:24,523 INFO [train.py:467] (3/4) Epoch 263, batch 6, global_batch_idx: 9700, batch size: 53, loss[discriminator_loss=2.625, discriminator_real_loss=1.273, discriminator_fake_loss=1.351, generator_loss=30.23, generator_mel_loss=21.28, generator_kl_loss=1.889, generator_dur_loss=1.699, generator_adv_loss=2.234, generator_feat_match_loss=3.121, over 53.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.316, discriminator_fake_loss=1.31, generator_loss=30.47, generator_mel_loss=21.67, generator_kl_loss=1.986, generator_dur_loss=1.701, generator_adv_loss=2.064, generator_feat_match_loss=3.048, over 426.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:31:11,118 INFO [train.py:811] (3/4) Start epoch 264
2023-11-13 12:33:01,594 INFO [train.py:467] (3/4) Epoch 264, batch 19, global_batch_idx: 9750, batch size: 126, loss[discriminator_loss=2.613, discriminator_real_loss=1.342, discriminator_fake_loss=1.272, generator_loss=31.16, generator_mel_loss=22.03, generator_kl_loss=1.915, generator_dur_loss=1.69, generator_adv_loss=2.279, generator_feat_match_loss=3.252, over 126.00 samples.], tot_loss[discriminator_loss=2.645, discriminator_real_loss=1.332, discriminator_fake_loss=1.313, generator_loss=30.12, generator_mel_loss=21.31, generator_kl_loss=1.916, generator_dur_loss=1.695, generator_adv_loss=2.114, generator_feat_match_loss=3.082, over 1395.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:34:31,249 INFO [train.py:811] (3/4) Start epoch 265
2023-11-13 12:37:46,661 INFO [train.py:467] (3/4) Epoch 265, batch 32, global_batch_idx: 9800, batch size: 69, loss[discriminator_loss=2.328, discriminator_real_loss=1.133, discriminator_fake_loss=1.195, generator_loss=32, generator_mel_loss=21.48, generator_kl_loss=1.983, generator_dur_loss=1.713, generator_adv_loss=2.555, generator_feat_match_loss=4.27, over 69.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.317, discriminator_fake_loss=1.309, generator_loss=31.03, generator_mel_loss=21.74, generator_kl_loss=1.942, generator_dur_loss=1.699, generator_adv_loss=2.243, generator_feat_match_loss=3.404, over 2615.00 samples.], cur_lr_g: 1.94e-04, cur_lr_d: 1.94e-04, grad_scale: 16.0
2023-11-13 12:37:46,663 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 12:37:57,628 INFO [train.py:517] (3/4) Epoch 265, validation: discriminator_loss=2.392, discriminator_real_loss=1.221, discriminator_fake_loss=1.171, generator_loss=32.52, generator_mel_loss=22.43, generator_kl_loss=2.05, generator_dur_loss=1.678, generator_adv_loss=2.219, generator_feat_match_loss=4.139, over 100.00 samples.
2023-11-13 12:37:57,629 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 12:38:15,315 INFO [train.py:811] (3/4) Start epoch 266
2023-11-13 12:41:48,846 INFO [train.py:811] (3/4) Start epoch 267
2023-11-13 12:42:51,425 INFO [train.py:467] (3/4) Epoch 267, batch 8, global_batch_idx: 9850, batch size: 64, loss[discriminator_loss=2.621, discriminator_real_loss=1.335, discriminator_fake_loss=1.287, generator_loss=30.05, generator_mel_loss=21.54, generator_kl_loss=1.875, generator_dur_loss=1.704, generator_adv_loss=1.975, generator_feat_match_loss=2.959, over 64.00 samples.], tot_loss[discriminator_loss=2.646, discriminator_real_loss=1.349, discriminator_fake_loss=1.298, generator_loss=30.13, generator_mel_loss=21.6, generator_kl_loss=1.942, generator_dur_loss=1.698, generator_adv_loss=1.994, generator_feat_match_loss=2.89, over 639.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 12:45:23,314 INFO [train.py:811] (3/4) Start epoch 268
2023-11-13 12:47:39,260 INFO [train.py:467] (3/4) Epoch 268, batch 21, global_batch_idx: 9900, batch size: 76, loss[discriminator_loss=2.803, discriminator_real_loss=1.613, discriminator_fake_loss=1.189, generator_loss=30.63, generator_mel_loss=21.61, generator_kl_loss=1.911, generator_dur_loss=1.701, generator_adv_loss=2.344, generator_feat_match_loss=3.061, over 76.00 samples.], tot_loss[discriminator_loss=2.683, discriminator_real_loss=1.346, discriminator_fake_loss=1.337, generator_loss=30.55, generator_mel_loss=21.51, generator_kl_loss=1.965, generator_dur_loss=1.699, generator_adv_loss=2.175, generator_feat_match_loss=3.196, over 1478.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 12:48:59,512 INFO [train.py:811] (3/4) Start epoch 269
2023-11-13 12:52:21,589 INFO [train.py:467] (3/4) Epoch 269, batch 34, global_batch_idx: 9950, batch size: 65, loss[discriminator_loss=2.377, discriminator_real_loss=1.177, discriminator_fake_loss=1.2, generator_loss=32.21, generator_mel_loss=21.71, generator_kl_loss=1.908, generator_dur_loss=1.695, generator_adv_loss=2.498, generator_feat_match_loss=4.402, over 65.00 samples.], tot_loss[discriminator_loss=2.631, discriminator_real_loss=1.321, discriminator_fake_loss=1.31, generator_loss=30.55, generator_mel_loss=21.5, generator_kl_loss=1.94, generator_dur_loss=1.702, generator_adv_loss=2.182, generator_feat_match_loss=3.227, over 2626.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 12:52:32,445 INFO [train.py:811] (3/4) Start epoch 270
2023-11-13 12:55:58,788 INFO [train.py:811] (3/4) Start epoch 271
2023-11-13 12:57:08,100 INFO [train.py:467] (3/4) Epoch 271, batch 10, global_batch_idx: 10000, batch size: 54, loss[discriminator_loss=2.652, discriminator_real_loss=1.367, discriminator_fake_loss=1.286, generator_loss=30.46, generator_mel_loss=21.82, generator_kl_loss=1.926, generator_dur_loss=1.687, generator_adv_loss=2.143, generator_feat_match_loss=2.879, over 54.00 samples.], tot_loss[discriminator_loss=2.642, discriminator_real_loss=1.332, discriminator_fake_loss=1.31, generator_loss=29.94, generator_mel_loss=21.36, generator_kl_loss=1.942, generator_dur_loss=1.7, generator_adv_loss=2.017, generator_feat_match_loss=2.92, over 704.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 32.0
2023-11-13 12:57:08,104 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 12:57:18,945 INFO [train.py:517] (3/4) Epoch 271, validation: discriminator_loss=2.664, discriminator_real_loss=1.391, discriminator_fake_loss=1.273, generator_loss=31.42, generator_mel_loss=22.54, generator_kl_loss=1.989, generator_dur_loss=1.673, generator_adv_loss=2.122, generator_feat_match_loss=3.091, over 100.00 samples.
2023-11-13 12:57:18,946 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 12:59:41,335 INFO [train.py:811] (3/4) Start epoch 272
2023-11-13 13:01:58,846 INFO [train.py:467] (3/4) Epoch 272, batch 23, global_batch_idx: 10050, batch size: 126, loss[discriminator_loss=2.498, discriminator_real_loss=1.3, discriminator_fake_loss=1.198, generator_loss=31.45, generator_mel_loss=21.78, generator_kl_loss=2.011, generator_dur_loss=1.702, generator_adv_loss=2.312, generator_feat_match_loss=3.65, over 126.00 samples.], tot_loss[discriminator_loss=2.627, discriminator_real_loss=1.315, discriminator_fake_loss=1.312, generator_loss=30.46, generator_mel_loss=21.48, generator_kl_loss=1.926, generator_dur_loss=1.689, generator_adv_loss=2.155, generator_feat_match_loss=3.218, over 1829.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:03:12,751 INFO [train.py:811] (3/4) Start epoch 273
2023-11-13 13:06:49,414 INFO [train.py:467] (3/4) Epoch 273, batch 36, global_batch_idx: 10100, batch size: 56, loss[discriminator_loss=2.691, discriminator_real_loss=1.313, discriminator_fake_loss=1.379, generator_loss=29.92, generator_mel_loss=21.22, generator_kl_loss=1.973, generator_dur_loss=1.684, generator_adv_loss=1.942, generator_feat_match_loss=3.102, over 56.00 samples.], tot_loss[discriminator_loss=2.701, discriminator_real_loss=1.392, discriminator_fake_loss=1.309, generator_loss=30.46, generator_mel_loss=21.42, generator_kl_loss=1.918, generator_dur_loss=1.696, generator_adv_loss=2.168, generator_feat_match_loss=3.252, over 2760.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:06:50,107 INFO [train.py:811] (3/4) Start epoch 274
2023-11-13 13:10:22,981 INFO [train.py:811] (3/4) Start epoch 275
2023-11-13 13:11:43,920 INFO [train.py:467] (3/4) Epoch 275, batch 12, global_batch_idx: 10150, batch size: 69, loss[discriminator_loss=2.789, discriminator_real_loss=1.407, discriminator_fake_loss=1.382, generator_loss=29.27, generator_mel_loss=21.02, generator_kl_loss=1.944, generator_dur_loss=1.694, generator_adv_loss=1.973, generator_feat_match_loss=2.641, over 69.00 samples.], tot_loss[discriminator_loss=2.67, discriminator_real_loss=1.359, discriminator_fake_loss=1.311, generator_loss=30.5, generator_mel_loss=21.49, generator_kl_loss=1.972, generator_dur_loss=1.698, generator_adv_loss=2.113, generator_feat_match_loss=3.226, over 937.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:13:56,072 INFO [train.py:811] (3/4) Start epoch 276
2023-11-13 13:16:19,350 INFO [train.py:467] (3/4) Epoch 276, batch 25, global_batch_idx: 10200, batch size: 67, loss[discriminator_loss=2.598, discriminator_real_loss=1.41, discriminator_fake_loss=1.188, generator_loss=31.16, generator_mel_loss=21.57, generator_kl_loss=2.009, generator_dur_loss=1.715, generator_adv_loss=2.35, generator_feat_match_loss=3.52, over 67.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.331, discriminator_fake_loss=1.297, generator_loss=30.43, generator_mel_loss=21.5, generator_kl_loss=1.952, generator_dur_loss=1.695, generator_adv_loss=2.132, generator_feat_match_loss=3.153, over 1755.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:16:19,352 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 13:16:30,529 INFO [train.py:517] (3/4) Epoch 276, validation: discriminator_loss=2.581, discriminator_real_loss=1.203, discriminator_fake_loss=1.377, generator_loss=31.56, generator_mel_loss=22.84, generator_kl_loss=2.009, generator_dur_loss=1.67, generator_adv_loss=1.857, generator_feat_match_loss=3.182, over 100.00 samples.
2023-11-13 13:16:30,530 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 13:17:37,618 INFO [train.py:811] (3/4) Start epoch 277
2023-11-13 13:21:10,476 INFO [train.py:811] (3/4) Start epoch 278
2023-11-13 13:21:33,442 INFO [train.py:467] (3/4) Epoch 278, batch 1, global_batch_idx: 10250, batch size: 90, loss[discriminator_loss=2.711, discriminator_real_loss=1.299, discriminator_fake_loss=1.412, generator_loss=29.88, generator_mel_loss=21.41, generator_kl_loss=1.89, generator_dur_loss=1.681, generator_adv_loss=1.983, generator_feat_match_loss=2.918, over 90.00 samples.], tot_loss[discriminator_loss=2.674, discriminator_real_loss=1.329, discriminator_fake_loss=1.345, generator_loss=30.09, generator_mel_loss=21.49, generator_kl_loss=1.916, generator_dur_loss=1.684, generator_adv_loss=1.999, generator_feat_match_loss=3.001, over 180.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:24:40,142 INFO [train.py:811] (3/4) Start epoch 279
2023-11-13 13:26:13,091 INFO [train.py:467] (3/4) Epoch 279, batch 14, global_batch_idx: 10300, batch size: 55, loss[discriminator_loss=2.725, discriminator_real_loss=1.29, discriminator_fake_loss=1.435, generator_loss=30.62, generator_mel_loss=21.42, generator_kl_loss=1.937, generator_dur_loss=1.715, generator_adv_loss=2.406, generator_feat_match_loss=3.141, over 55.00 samples.], tot_loss[discriminator_loss=2.686, discriminator_real_loss=1.381, discriminator_fake_loss=1.305, generator_loss=30.47, generator_mel_loss=21.55, generator_kl_loss=1.918, generator_dur_loss=1.7, generator_adv_loss=2.14, generator_feat_match_loss=3.166, over 966.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:28:15,408 INFO [train.py:811] (3/4) Start epoch 280
2023-11-13 13:30:56,373 INFO [train.py:467] (3/4) Epoch 280, batch 27, global_batch_idx: 10350, batch size: 54, loss[discriminator_loss=2.758, discriminator_real_loss=1.533, discriminator_fake_loss=1.225, generator_loss=29.8, generator_mel_loss=21.27, generator_kl_loss=1.882, generator_dur_loss=1.736, generator_adv_loss=2.289, generator_feat_match_loss=2.619, over 54.00 samples.], tot_loss[discriminator_loss=2.647, discriminator_real_loss=1.361, discriminator_fake_loss=1.286, generator_loss=30.59, generator_mel_loss=21.33, generator_kl_loss=1.957, generator_dur_loss=1.696, generator_adv_loss=2.23, generator_feat_match_loss=3.371, over 1906.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:31:43,669 INFO [train.py:811] (3/4) Start epoch 281
2023-11-13 13:35:19,296 INFO [train.py:811] (3/4) Start epoch 282
2023-11-13 13:35:52,400 INFO [train.py:467] (3/4) Epoch 282, batch 3, global_batch_idx: 10400, batch size: 63, loss[discriminator_loss=2.666, discriminator_real_loss=1.413, discriminator_fake_loss=1.253, generator_loss=29.74, generator_mel_loss=20.82, generator_kl_loss=1.96, generator_dur_loss=1.717, generator_adv_loss=2.328, generator_feat_match_loss=2.918, over 63.00 samples.], tot_loss[discriminator_loss=2.611, discriminator_real_loss=1.302, discriminator_fake_loss=1.31, generator_loss=30.44, generator_mel_loss=21.2, generator_kl_loss=1.974, generator_dur_loss=1.703, generator_adv_loss=2.236, generator_feat_match_loss=3.325, over 268.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 32.0
2023-11-13 13:35:52,401 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 13:36:04,175 INFO [train.py:517] (3/4) Epoch 282, validation: discriminator_loss=2.715, discriminator_real_loss=1.545, discriminator_fake_loss=1.17, generator_loss=31.62, generator_mel_loss=22.3, generator_kl_loss=2.052, generator_dur_loss=1.661, generator_adv_loss=2.386, generator_feat_match_loss=3.221, over 100.00 samples.
2023-11-13 13:36:04,176 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 13:39:00,274 INFO [train.py:811] (3/4) Start epoch 283
2023-11-13 13:40:33,496 INFO [train.py:467] (3/4) Epoch 283, batch 16, global_batch_idx: 10450, batch size: 63, loss[discriminator_loss=2.479, discriminator_real_loss=1.207, discriminator_fake_loss=1.271, generator_loss=29.68, generator_mel_loss=20.67, generator_kl_loss=1.74, generator_dur_loss=1.676, generator_adv_loss=2.164, generator_feat_match_loss=3.432, over 63.00 samples.], tot_loss[discriminator_loss=2.599, discriminator_real_loss=1.306, discriminator_fake_loss=1.293, generator_loss=30.43, generator_mel_loss=21.4, generator_kl_loss=1.957, generator_dur_loss=1.699, generator_adv_loss=2.143, generator_feat_match_loss=3.231, over 1136.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:42:27,450 INFO [train.py:811] (3/4) Start epoch 284
2023-11-13 13:45:23,182 INFO [train.py:467] (3/4) Epoch 284, batch 29, global_batch_idx: 10500, batch size: 73, loss[discriminator_loss=2.668, discriminator_real_loss=1.394, discriminator_fake_loss=1.274, generator_loss=30.2, generator_mel_loss=21.49, generator_kl_loss=1.941, generator_dur_loss=1.685, generator_adv_loss=2.008, generator_feat_match_loss=3.076, over 73.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.34, discriminator_fake_loss=1.316, generator_loss=30.31, generator_mel_loss=21.3, generator_kl_loss=1.952, generator_dur_loss=1.696, generator_adv_loss=2.143, generator_feat_match_loss=3.214, over 2209.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:46:06,843 INFO [train.py:811] (3/4) Start epoch 285
2023-11-13 13:49:43,278 INFO [train.py:811] (3/4) Start epoch 286
2023-11-13 13:50:24,560 INFO [train.py:467] (3/4) Epoch 286, batch 5, global_batch_idx: 10550, batch size: 69, loss[discriminator_loss=2.77, discriminator_real_loss=1.305, discriminator_fake_loss=1.464, generator_loss=29.78, generator_mel_loss=21.53, generator_kl_loss=1.819, generator_dur_loss=1.692, generator_adv_loss=1.964, generator_feat_match_loss=2.773, over 69.00 samples.], tot_loss[discriminator_loss=2.735, discriminator_real_loss=1.398, discriminator_fake_loss=1.337, generator_loss=30.28, generator_mel_loss=21.43, generator_kl_loss=1.891, generator_dur_loss=1.697, generator_adv_loss=2.061, generator_feat_match_loss=3.194, over 380.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:53:17,403 INFO [train.py:811] (3/4) Start epoch 287
2023-11-13 13:55:11,104 INFO [train.py:467] (3/4) Epoch 287, batch 18, global_batch_idx: 10600, batch size: 71, loss[discriminator_loss=2.592, discriminator_real_loss=1.465, discriminator_fake_loss=1.127, generator_loss=30.84, generator_mel_loss=21.38, generator_kl_loss=1.941, generator_dur_loss=1.679, generator_adv_loss=2.41, generator_feat_match_loss=3.43, over 71.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.31, discriminator_fake_loss=1.275, generator_loss=30.64, generator_mel_loss=21.43, generator_kl_loss=1.942, generator_dur_loss=1.693, generator_adv_loss=2.212, generator_feat_match_loss=3.357, over 1250.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 13:55:11,105 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 13:55:22,043 INFO [train.py:517] (3/4) Epoch 287, validation: discriminator_loss=2.661, discriminator_real_loss=1.14, discriminator_fake_loss=1.521, generator_loss=30.62, generator_mel_loss=21.94, generator_kl_loss=2.034, generator_dur_loss=1.667, generator_adv_loss=1.799, generator_feat_match_loss=3.179, over 100.00 samples.
2023-11-13 13:55:22,044 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 13:56:58,593 INFO [train.py:811] (3/4) Start epoch 288
2023-11-13 13:59:59,278 INFO [train.py:467] (3/4) Epoch 288, batch 31, global_batch_idx: 10650, batch size: 126, loss[discriminator_loss=2.621, discriminator_real_loss=1.491, discriminator_fake_loss=1.129, generator_loss=30.38, generator_mel_loss=21.58, generator_kl_loss=1.878, generator_dur_loss=1.673, generator_adv_loss=2.082, generator_feat_match_loss=3.17, over 126.00 samples.], tot_loss[discriminator_loss=2.661, discriminator_real_loss=1.354, discriminator_fake_loss=1.306, generator_loss=30.43, generator_mel_loss=21.32, generator_kl_loss=1.922, generator_dur_loss=1.695, generator_adv_loss=2.181, generator_feat_match_loss=3.314, over 2270.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:00:30,359 INFO [train.py:811] (3/4) Start epoch 289
2023-11-13 14:04:01,517 INFO [train.py:811] (3/4) Start epoch 290
2023-11-13 14:05:00,431 INFO [train.py:467] (3/4) Epoch 290, batch 7, global_batch_idx: 10700, batch size: 90, loss[discriminator_loss=2.65, discriminator_real_loss=1.453, discriminator_fake_loss=1.197, generator_loss=30.4, generator_mel_loss=21.39, generator_kl_loss=1.995, generator_dur_loss=1.687, generator_adv_loss=2.168, generator_feat_match_loss=3.156, over 90.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.3, discriminator_fake_loss=1.267, generator_loss=30.78, generator_mel_loss=21.38, generator_kl_loss=1.928, generator_dur_loss=1.691, generator_adv_loss=2.237, generator_feat_match_loss=3.544, over 672.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:07:32,205 INFO [train.py:811] (3/4) Start epoch 291
2023-11-13 14:09:36,413 INFO [train.py:467] (3/4) Epoch 291, batch 20, global_batch_idx: 10750, batch size: 126, loss[discriminator_loss=2.693, discriminator_real_loss=1.835, discriminator_fake_loss=0.8589, generator_loss=31.53, generator_mel_loss=21.5, generator_kl_loss=1.988, generator_dur_loss=1.67, generator_adv_loss=2.33, generator_feat_match_loss=4.047, over 126.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.343, discriminator_fake_loss=1.285, generator_loss=30.95, generator_mel_loss=21.52, generator_kl_loss=1.945, generator_dur_loss=1.696, generator_adv_loss=2.276, generator_feat_match_loss=3.506, over 1532.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:11:01,112 INFO [train.py:811] (3/4) Start epoch 292
2023-11-13 14:14:14,493 INFO [train.py:467] (3/4) Epoch 292, batch 33, global_batch_idx: 10800, batch size: 71, loss[discriminator_loss=2.672, discriminator_real_loss=1.332, discriminator_fake_loss=1.341, generator_loss=29.6, generator_mel_loss=21.09, generator_kl_loss=1.936, generator_dur_loss=1.686, generator_adv_loss=2.016, generator_feat_match_loss=2.867, over 71.00 samples.], tot_loss[discriminator_loss=2.637, discriminator_real_loss=1.338, discriminator_fake_loss=1.299, generator_loss=30.38, generator_mel_loss=21.34, generator_kl_loss=1.944, generator_dur_loss=1.693, generator_adv_loss=2.135, generator_feat_match_loss=3.273, over 2386.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 32.0
2023-11-13 14:14:14,494 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 14:14:25,391 INFO [train.py:517] (3/4) Epoch 292, validation: discriminator_loss=2.609, discriminator_real_loss=1.31, discriminator_fake_loss=1.299, generator_loss=30.94, generator_mel_loss=22.17, generator_kl_loss=1.969, generator_dur_loss=1.665, generator_adv_loss=2.03, generator_feat_match_loss=3.101, over 100.00 samples.
2023-11-13 14:14:25,392 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 14:14:46,202 INFO [train.py:811] (3/4) Start epoch 293
2023-11-13 14:18:20,614 INFO [train.py:811] (3/4) Start epoch 294
2023-11-13 14:19:23,569 INFO [train.py:467] (3/4) Epoch 294, batch 9, global_batch_idx: 10850, batch size: 60, loss[discriminator_loss=2.605, discriminator_real_loss=1.287, discriminator_fake_loss=1.319, generator_loss=29.5, generator_mel_loss=20.85, generator_kl_loss=1.958, generator_dur_loss=1.671, generator_adv_loss=1.954, generator_feat_match_loss=3.07, over 60.00 samples.], tot_loss[discriminator_loss=2.638, discriminator_real_loss=1.333, discriminator_fake_loss=1.306, generator_loss=30.27, generator_mel_loss=21.4, generator_kl_loss=1.952, generator_dur_loss=1.687, generator_adv_loss=2.074, generator_feat_match_loss=3.157, over 800.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:21:52,359 INFO [train.py:811] (3/4) Start epoch 295
2023-11-13 14:23:59,386 INFO [train.py:467] (3/4) Epoch 295, batch 22, global_batch_idx: 10900, batch size: 58, loss[discriminator_loss=2.512, discriminator_real_loss=1.342, discriminator_fake_loss=1.169, generator_loss=30.16, generator_mel_loss=20.46, generator_kl_loss=1.951, generator_dur_loss=1.671, generator_adv_loss=2.316, generator_feat_match_loss=3.758, over 58.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.315, discriminator_fake_loss=1.266, generator_loss=30.71, generator_mel_loss=21.16, generator_kl_loss=1.92, generator_dur_loss=1.695, generator_adv_loss=2.306, generator_feat_match_loss=3.629, over 1539.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:25:18,340 INFO [train.py:811] (3/4) Start epoch 296
2023-11-13 14:28:45,954 INFO [train.py:467] (3/4) Epoch 296, batch 35, global_batch_idx: 10950, batch size: 59, loss[discriminator_loss=2.953, discriminator_real_loss=1.542, discriminator_fake_loss=1.41, generator_loss=30.43, generator_mel_loss=21.63, generator_kl_loss=2.027, generator_dur_loss=1.709, generator_adv_loss=2.158, generator_feat_match_loss=2.902, over 59.00 samples.], tot_loss[discriminator_loss=2.619, discriminator_real_loss=1.319, discriminator_fake_loss=1.3, generator_loss=30.66, generator_mel_loss=21.44, generator_kl_loss=1.948, generator_dur_loss=1.69, generator_adv_loss=2.201, generator_feat_match_loss=3.379, over 2772.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:28:52,601 INFO [train.py:811] (3/4) Start epoch 297
2023-11-13 14:32:25,215 INFO [train.py:811] (3/4) Start epoch 298
2023-11-13 14:33:34,830 INFO [train.py:467] (3/4) Epoch 298, batch 11, global_batch_idx: 11000, batch size: 67, loss[discriminator_loss=2.549, discriminator_real_loss=1.379, discriminator_fake_loss=1.17, generator_loss=30.29, generator_mel_loss=20.49, generator_kl_loss=1.903, generator_dur_loss=1.696, generator_adv_loss=2.434, generator_feat_match_loss=3.773, over 67.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.307, discriminator_fake_loss=1.271, generator_loss=30.51, generator_mel_loss=21.29, generator_kl_loss=1.954, generator_dur_loss=1.694, generator_adv_loss=2.21, generator_feat_match_loss=3.37, over 837.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:33:34,831 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 14:33:46,173 INFO [train.py:517] (3/4) Epoch 298, validation: discriminator_loss=2.492, discriminator_real_loss=1.252, discriminator_fake_loss=1.24, generator_loss=30.79, generator_mel_loss=21.56, generator_kl_loss=1.967, generator_dur_loss=1.664, generator_adv_loss=2.107, generator_feat_match_loss=3.486, over 100.00 samples.
2023-11-13 14:33:46,174 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 14:36:01,068 INFO [train.py:811] (3/4) Start epoch 299
2023-11-13 14:38:28,914 INFO [train.py:467] (3/4) Epoch 299, batch 24, global_batch_idx: 11050, batch size: 153, loss[discriminator_loss=2.738, discriminator_real_loss=1.44, discriminator_fake_loss=1.298, generator_loss=29.63, generator_mel_loss=21.19, generator_kl_loss=1.885, generator_dur_loss=1.67, generator_adv_loss=1.999, generator_feat_match_loss=2.887, over 153.00 samples.], tot_loss[discriminator_loss=2.602, discriminator_real_loss=1.335, discriminator_fake_loss=1.267, generator_loss=30.61, generator_mel_loss=21.15, generator_kl_loss=1.94, generator_dur_loss=1.688, generator_adv_loss=2.268, generator_feat_match_loss=3.566, over 1833.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:39:35,473 INFO [train.py:811] (3/4) Start epoch 300
2023-11-13 14:43:07,614 INFO [train.py:811] (3/4) Start epoch 301
2023-11-13 14:43:25,745 INFO [train.py:467] (3/4) Epoch 301, batch 0, global_batch_idx: 11100, batch size: 153, loss[discriminator_loss=2.725, discriminator_real_loss=1.406, discriminator_fake_loss=1.318, generator_loss=29.89, generator_mel_loss=21.27, generator_kl_loss=1.905, generator_dur_loss=1.686, generator_adv_loss=2.031, generator_feat_match_loss=3.006, over 153.00 samples.], tot_loss[discriminator_loss=2.725, discriminator_real_loss=1.406, discriminator_fake_loss=1.318, generator_loss=29.89, generator_mel_loss=21.27, generator_kl_loss=1.905, generator_dur_loss=1.686, generator_adv_loss=2.031, generator_feat_match_loss=3.006, over 153.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:46:38,697 INFO [train.py:811] (3/4) Start epoch 302
2023-11-13 14:47:59,370 INFO [train.py:467] (3/4) Epoch 302, batch 13, global_batch_idx: 11150, batch size: 53, loss[discriminator_loss=2.559, discriminator_real_loss=1.286, discriminator_fake_loss=1.273, generator_loss=30.46, generator_mel_loss=21.08, generator_kl_loss=1.86, generator_dur_loss=1.695, generator_adv_loss=2.135, generator_feat_match_loss=3.689, over 53.00 samples.], tot_loss[discriminator_loss=2.635, discriminator_real_loss=1.332, discriminator_fake_loss=1.304, generator_loss=30.66, generator_mel_loss=21.35, generator_kl_loss=1.949, generator_dur_loss=1.688, generator_adv_loss=2.22, generator_feat_match_loss=3.454, over 853.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:50:08,617 INFO [train.py:811] (3/4) Start epoch 303
2023-11-13 14:52:45,866 INFO [train.py:467] (3/4) Epoch 303, batch 26, global_batch_idx: 11200, batch size: 64, loss[discriminator_loss=2.947, discriminator_real_loss=1.639, discriminator_fake_loss=1.309, generator_loss=29.15, generator_mel_loss=20.62, generator_kl_loss=1.828, generator_dur_loss=1.672, generator_adv_loss=2.012, generator_feat_match_loss=3.012, over 64.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.337, discriminator_fake_loss=1.286, generator_loss=30.83, generator_mel_loss=21.33, generator_kl_loss=1.944, generator_dur_loss=1.69, generator_adv_loss=2.304, generator_feat_match_loss=3.564, over 1956.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 14:52:45,868 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 14:52:57,009 INFO [train.py:517] (3/4) Epoch 303, validation: discriminator_loss=2.558, discriminator_real_loss=1.27, discriminator_fake_loss=1.287, generator_loss=31.35, generator_mel_loss=22.14, generator_kl_loss=2.025, generator_dur_loss=1.664, generator_adv_loss=2.113, generator_feat_match_loss=3.409, over 100.00 samples.
2023-11-13 14:52:57,010 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 14:53:49,281 INFO [train.py:811] (3/4) Start epoch 304
2023-11-13 14:57:24,626 INFO [train.py:811] (3/4) Start epoch 305
2023-11-13 14:57:50,899 INFO [train.py:467] (3/4) Epoch 305, batch 2, global_batch_idx: 11250, batch size: 76, loss[discriminator_loss=2.541, discriminator_real_loss=1.265, discriminator_fake_loss=1.276, generator_loss=30.79, generator_mel_loss=21.22, generator_kl_loss=1.952, generator_dur_loss=1.691, generator_adv_loss=2.266, generator_feat_match_loss=3.656, over 76.00 samples.], tot_loss[discriminator_loss=2.572, discriminator_real_loss=1.293, discriminator_fake_loss=1.279, generator_loss=30.19, generator_mel_loss=21.05, generator_kl_loss=1.957, generator_dur_loss=1.689, generator_adv_loss=2.163, generator_feat_match_loss=3.335, over 220.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 15:00:51,783 INFO [train.py:811] (3/4) Start epoch 306
2023-11-13 15:02:25,445 INFO [train.py:467] (3/4) Epoch 306, batch 15, global_batch_idx: 11300, batch size: 53, loss[discriminator_loss=2.641, discriminator_real_loss=1.417, discriminator_fake_loss=1.225, generator_loss=30.23, generator_mel_loss=21.42, generator_kl_loss=1.909, generator_dur_loss=1.675, generator_adv_loss=2.24, generator_feat_match_loss=2.986, over 53.00 samples.], tot_loss[discriminator_loss=2.58, discriminator_real_loss=1.283, discriminator_fake_loss=1.297, generator_loss=30.49, generator_mel_loss=21.21, generator_kl_loss=1.932, generator_dur_loss=1.691, generator_adv_loss=2.218, generator_feat_match_loss=3.445, over 1189.00 samples.], cur_lr_g: 1.93e-04, cur_lr_d: 1.93e-04, grad_scale: 16.0
2023-11-13 15:04:23,248 INFO [train.py:811] (3/4) Start epoch 307
2023-11-13 15:07:05,905 INFO [train.py:467] (3/4) Epoch 307, batch 28, global_batch_idx: 11350, batch size: 64, loss[discriminator_loss=2.457, discriminator_real_loss=1.244, discriminator_fake_loss=1.214, generator_loss=29.61, generator_mel_loss=20.4, generator_kl_loss=1.885, generator_dur_loss=1.678, generator_adv_loss=2.287, generator_feat_match_loss=3.359, over 64.00 samples.], tot_loss[discriminator_loss=2.59, discriminator_real_loss=1.312, discriminator_fake_loss=1.278, generator_loss=30.74, generator_mel_loss=21.32, generator_kl_loss=1.922, generator_dur_loss=1.689, generator_adv_loss=2.251, generator_feat_match_loss=3.55, over 1886.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:07:52,410 INFO [train.py:811] (3/4) Start epoch 308
2023-11-13 15:11:22,212 INFO [train.py:811] (3/4) Start epoch 309
2023-11-13 15:11:56,221 INFO [train.py:467] (3/4) Epoch 309, batch 4, global_batch_idx: 11400, batch size: 90, loss[discriminator_loss=2.525, discriminator_real_loss=1.384, discriminator_fake_loss=1.142, generator_loss=31.26, generator_mel_loss=21.57, generator_kl_loss=1.986, generator_dur_loss=1.682, generator_adv_loss=2.133, generator_feat_match_loss=3.891, over 90.00 samples.], tot_loss[discriminator_loss=2.593, discriminator_real_loss=1.324, discriminator_fake_loss=1.268, generator_loss=30.64, generator_mel_loss=21.39, generator_kl_loss=1.945, generator_dur_loss=1.686, generator_adv_loss=2.174, generator_feat_match_loss=3.452, over 341.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:11:56,222 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 15:12:08,156 INFO [train.py:517] (3/4) Epoch 309, validation: discriminator_loss=2.516, discriminator_real_loss=1.074, discriminator_fake_loss=1.442, generator_loss=31.1, generator_mel_loss=22.33, generator_kl_loss=1.964, generator_dur_loss=1.66, generator_adv_loss=1.733, generator_feat_match_loss=3.416, over 100.00 samples.
2023-11-13 15:12:08,157 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 15:15:09,973 INFO [train.py:811] (3/4) Start epoch 310
2023-11-13 15:17:04,706 INFO [train.py:467] (3/4) Epoch 310, batch 17, global_batch_idx: 11450, batch size: 126, loss[discriminator_loss=2.66, discriminator_real_loss=1.323, discriminator_fake_loss=1.338, generator_loss=30.13, generator_mel_loss=21.36, generator_kl_loss=1.974, generator_dur_loss=1.652, generator_adv_loss=2.01, generator_feat_match_loss=3.143, over 126.00 samples.], tot_loss[discriminator_loss=2.65, discriminator_real_loss=1.356, discriminator_fake_loss=1.294, generator_loss=30.03, generator_mel_loss=21.18, generator_kl_loss=1.935, generator_dur_loss=1.681, generator_adv_loss=2.042, generator_feat_match_loss=3.189, over 1485.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:18:42,207 INFO [train.py:811] (3/4) Start epoch 311
2023-11-13 15:21:34,820 INFO [train.py:467] (3/4) Epoch 311, batch 30, global_batch_idx: 11500, batch size: 61, loss[discriminator_loss=2.734, discriminator_real_loss=1.309, discriminator_fake_loss=1.426, generator_loss=30.42, generator_mel_loss=21.59, generator_kl_loss=1.964, generator_dur_loss=1.692, generator_adv_loss=2.01, generator_feat_match_loss=3.164, over 61.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.345, discriminator_fake_loss=1.311, generator_loss=30.22, generator_mel_loss=21.35, generator_kl_loss=1.948, generator_dur_loss=1.685, generator_adv_loss=2.067, generator_feat_match_loss=3.172, over 2119.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:22:09,984 INFO [train.py:811] (3/4) Start epoch 312
2023-11-13 15:25:38,559 INFO [train.py:811] (3/4) Start epoch 313
2023-11-13 15:26:27,486 INFO [train.py:467] (3/4) Epoch 313, batch 6, global_batch_idx: 11550, batch size: 54, loss[discriminator_loss=2.723, discriminator_real_loss=1.367, discriminator_fake_loss=1.354, generator_loss=29.98, generator_mel_loss=21.26, generator_kl_loss=1.776, generator_dur_loss=1.667, generator_adv_loss=2.096, generator_feat_match_loss=3.18, over 54.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.346, discriminator_fake_loss=1.33, generator_loss=30.36, generator_mel_loss=21.39, generator_kl_loss=1.924, generator_dur_loss=1.681, generator_adv_loss=2.101, generator_feat_match_loss=3.265, over 474.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:29:11,580 INFO [train.py:811] (3/4) Start epoch 314
2023-11-13 15:31:19,124 INFO [train.py:467] (3/4) Epoch 314, batch 19, global_batch_idx: 11600, batch size: 55, loss[discriminator_loss=2.746, discriminator_real_loss=1.407, discriminator_fake_loss=1.338, generator_loss=29.41, generator_mel_loss=21.06, generator_kl_loss=1.946, generator_dur_loss=1.7, generator_adv_loss=1.98, generator_feat_match_loss=2.723, over 55.00 samples.], tot_loss[discriminator_loss=2.682, discriminator_real_loss=1.354, discriminator_fake_loss=1.328, generator_loss=30.51, generator_mel_loss=21.52, generator_kl_loss=1.927, generator_dur_loss=1.686, generator_adv_loss=2.138, generator_feat_match_loss=3.239, over 1484.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 32.0
2023-11-13 15:31:19,126 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 15:31:30,954 INFO [train.py:517] (3/4) Epoch 314, validation: discriminator_loss=2.669, discriminator_real_loss=1.306, discriminator_fake_loss=1.363, generator_loss=30.58, generator_mel_loss=22.17, generator_kl_loss=1.924, generator_dur_loss=1.659, generator_adv_loss=1.836, generator_feat_match_loss=2.989, over 100.00 samples.
2023-11-13 15:31:30,955 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 15:32:58,479 INFO [train.py:811] (3/4) Start epoch 315
2023-11-13 15:36:17,432 INFO [train.py:467] (3/4) Epoch 315, batch 32, global_batch_idx: 11650, batch size: 67, loss[discriminator_loss=2.389, discriminator_real_loss=1.213, discriminator_fake_loss=1.176, generator_loss=32.28, generator_mel_loss=21.37, generator_kl_loss=1.884, generator_dur_loss=1.708, generator_adv_loss=2.865, generator_feat_match_loss=4.449, over 67.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.312, discriminator_fake_loss=1.303, generator_loss=30.81, generator_mel_loss=21.46, generator_kl_loss=1.947, generator_dur_loss=1.683, generator_adv_loss=2.226, generator_feat_match_loss=3.49, over 2595.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:36:37,179 INFO [train.py:811] (3/4) Start epoch 316
2023-11-13 15:40:13,784 INFO [train.py:811] (3/4) Start epoch 317
2023-11-13 15:41:18,001 INFO [train.py:467] (3/4) Epoch 317, batch 8, global_batch_idx: 11700, batch size: 63, loss[discriminator_loss=2.742, discriminator_real_loss=1.308, discriminator_fake_loss=1.434, generator_loss=29.29, generator_mel_loss=21.2, generator_kl_loss=1.903, generator_dur_loss=1.674, generator_adv_loss=1.781, generator_feat_match_loss=2.732, over 63.00 samples.], tot_loss[discriminator_loss=2.674, discriminator_real_loss=1.31, discriminator_fake_loss=1.363, generator_loss=30.18, generator_mel_loss=21.18, generator_kl_loss=1.94, generator_dur_loss=1.686, generator_adv_loss=2.127, generator_feat_match_loss=3.249, over 827.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:43:45,660 INFO [train.py:811] (3/4) Start epoch 318
2023-11-13 15:46:01,210 INFO [train.py:467] (3/4) Epoch 318, batch 21, global_batch_idx: 11750, batch size: 110, loss[discriminator_loss=2.611, discriminator_real_loss=1.338, discriminator_fake_loss=1.273, generator_loss=30.36, generator_mel_loss=21.6, generator_kl_loss=1.981, generator_dur_loss=1.649, generator_adv_loss=2.033, generator_feat_match_loss=3.1, over 110.00 samples.], tot_loss[discriminator_loss=2.598, discriminator_real_loss=1.316, discriminator_fake_loss=1.282, generator_loss=30.24, generator_mel_loss=21.17, generator_kl_loss=1.94, generator_dur_loss=1.685, generator_adv_loss=2.14, generator_feat_match_loss=3.311, over 1558.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:47:19,269 INFO [train.py:811] (3/4) Start epoch 319
2023-11-13 15:50:47,695 INFO [train.py:467] (3/4) Epoch 319, batch 34, global_batch_idx: 11800, batch size: 76, loss[discriminator_loss=2.566, discriminator_real_loss=1.288, discriminator_fake_loss=1.279, generator_loss=30.63, generator_mel_loss=20.92, generator_kl_loss=1.869, generator_dur_loss=1.687, generator_adv_loss=2.461, generator_feat_match_loss=3.691, over 76.00 samples.], tot_loss[discriminator_loss=2.592, discriminator_real_loss=1.316, discriminator_fake_loss=1.276, generator_loss=30.68, generator_mel_loss=21.19, generator_kl_loss=1.933, generator_dur_loss=1.683, generator_adv_loss=2.291, generator_feat_match_loss=3.581, over 2569.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:50:47,697 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 15:51:00,596 INFO [train.py:517] (3/4) Epoch 319, validation: discriminator_loss=2.383, discriminator_real_loss=1.151, discriminator_fake_loss=1.232, generator_loss=31.38, generator_mel_loss=22.12, generator_kl_loss=1.926, generator_dur_loss=1.659, generator_adv_loss=2.087, generator_feat_match_loss=3.581, over 100.00 samples.
2023-11-13 15:51:00,597 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 15:51:10,377 INFO [train.py:811] (3/4) Start epoch 320
2023-11-13 15:54:48,078 INFO [train.py:811] (3/4) Start epoch 321
2023-11-13 15:55:58,420 INFO [train.py:467] (3/4) Epoch 321, batch 10, global_batch_idx: 11850, batch size: 71, loss[discriminator_loss=2.719, discriminator_real_loss=1.195, discriminator_fake_loss=1.523, generator_loss=30.99, generator_mel_loss=21.81, generator_kl_loss=1.984, generator_dur_loss=1.685, generator_adv_loss=2.049, generator_feat_match_loss=3.459, over 71.00 samples.], tot_loss[discriminator_loss=2.653, discriminator_real_loss=1.334, discriminator_fake_loss=1.32, generator_loss=30.48, generator_mel_loss=21.35, generator_kl_loss=1.961, generator_dur_loss=1.682, generator_adv_loss=2.171, generator_feat_match_loss=3.322, over 815.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 15:58:24,692 INFO [train.py:811] (3/4) Start epoch 322
2023-11-13 16:00:41,129 INFO [train.py:467] (3/4) Epoch 322, batch 23, global_batch_idx: 11900, batch size: 76, loss[discriminator_loss=2.574, discriminator_real_loss=1.315, discriminator_fake_loss=1.26, generator_loss=30.44, generator_mel_loss=21.36, generator_kl_loss=2.01, generator_dur_loss=1.679, generator_adv_loss=2.191, generator_feat_match_loss=3.195, over 76.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.328, discriminator_fake_loss=1.287, generator_loss=30.23, generator_mel_loss=21.18, generator_kl_loss=1.929, generator_dur_loss=1.686, generator_adv_loss=2.143, generator_feat_match_loss=3.294, over 1654.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:01:56,378 INFO [train.py:811] (3/4) Start epoch 323
2023-11-13 16:05:28,925 INFO [train.py:467] (3/4) Epoch 323, batch 36, global_batch_idx: 11950, batch size: 64, loss[discriminator_loss=2.902, discriminator_real_loss=1.782, discriminator_fake_loss=1.12, generator_loss=29.75, generator_mel_loss=21.04, generator_kl_loss=1.987, generator_dur_loss=1.677, generator_adv_loss=2.156, generator_feat_match_loss=2.898, over 64.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.346, discriminator_fake_loss=1.306, generator_loss=30.56, generator_mel_loss=21.35, generator_kl_loss=1.927, generator_dur_loss=1.688, generator_adv_loss=2.187, generator_feat_match_loss=3.405, over 2633.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:05:29,714 INFO [train.py:811] (3/4) Start epoch 324
2023-11-13 16:09:06,259 INFO [train.py:811] (3/4) Start epoch 325
2023-11-13 16:10:33,205 INFO [train.py:467] (3/4) Epoch 325, batch 12, global_batch_idx: 12000, batch size: 126, loss[discriminator_loss=2.688, discriminator_real_loss=1.311, discriminator_fake_loss=1.376, generator_loss=30.37, generator_mel_loss=21.6, generator_kl_loss=1.961, generator_dur_loss=1.677, generator_adv_loss=1.976, generator_feat_match_loss=3.16, over 126.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.298, discriminator_fake_loss=1.32, generator_loss=30.51, generator_mel_loss=21.37, generator_kl_loss=1.934, generator_dur_loss=1.684, generator_adv_loss=2.108, generator_feat_match_loss=3.41, over 1048.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 32.0
2023-11-13 16:10:33,207 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 16:10:45,042 INFO [train.py:517] (3/4) Epoch 325, validation: discriminator_loss=2.661, discriminator_real_loss=1.313, discriminator_fake_loss=1.347, generator_loss=31.06, generator_mel_loss=22.35, generator_kl_loss=2.023, generator_dur_loss=1.656, generator_adv_loss=1.906, generator_feat_match_loss=3.13, over 100.00 samples.
2023-11-13 16:10:45,043 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 16:12:57,373 INFO [train.py:811] (3/4) Start epoch 326
2023-11-13 16:15:34,173 INFO [train.py:467] (3/4) Epoch 326, batch 25, global_batch_idx: 12050, batch size: 67, loss[discriminator_loss=2.672, discriminator_real_loss=1.324, discriminator_fake_loss=1.348, generator_loss=31.23, generator_mel_loss=21.57, generator_kl_loss=2.025, generator_dur_loss=1.701, generator_adv_loss=2.365, generator_feat_match_loss=3.568, over 67.00 samples.], tot_loss[discriminator_loss=2.619, discriminator_real_loss=1.326, discriminator_fake_loss=1.293, generator_loss=30.54, generator_mel_loss=21.32, generator_kl_loss=1.924, generator_dur_loss=1.685, generator_adv_loss=2.179, generator_feat_match_loss=3.429, over 1741.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:16:33,799 INFO [train.py:811] (3/4) Start epoch 327
2023-11-13 16:20:13,970 INFO [train.py:811] (3/4) Start epoch 328
2023-11-13 16:20:36,005 INFO [train.py:467] (3/4) Epoch 328, batch 1, global_batch_idx: 12100, batch size: 53, loss[discriminator_loss=2.66, discriminator_real_loss=1.303, discriminator_fake_loss=1.357, generator_loss=30.36, generator_mel_loss=21.09, generator_kl_loss=1.928, generator_dur_loss=1.711, generator_adv_loss=2.1, generator_feat_match_loss=3.531, over 53.00 samples.], tot_loss[discriminator_loss=2.662, discriminator_real_loss=1.414, discriminator_fake_loss=1.249, generator_loss=30.58, generator_mel_loss=21.27, generator_kl_loss=1.916, generator_dur_loss=1.711, generator_adv_loss=2.127, generator_feat_match_loss=3.557, over 116.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:23:46,581 INFO [train.py:811] (3/4) Start epoch 329
2023-11-13 16:25:14,811 INFO [train.py:467] (3/4) Epoch 329, batch 14, global_batch_idx: 12150, batch size: 73, loss[discriminator_loss=2.621, discriminator_real_loss=1.273, discriminator_fake_loss=1.347, generator_loss=29.83, generator_mel_loss=20.94, generator_kl_loss=1.934, generator_dur_loss=1.682, generator_adv_loss=2.047, generator_feat_match_loss=3.23, over 73.00 samples.], tot_loss[discriminator_loss=2.662, discriminator_real_loss=1.352, discriminator_fake_loss=1.31, generator_loss=30.44, generator_mel_loss=21.38, generator_kl_loss=1.933, generator_dur_loss=1.681, generator_adv_loss=2.112, generator_feat_match_loss=3.333, over 1173.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:27:13,393 INFO [train.py:811] (3/4) Start epoch 330
2023-11-13 16:29:59,117 INFO [train.py:467] (3/4) Epoch 330, batch 27, global_batch_idx: 12200, batch size: 53, loss[discriminator_loss=2.74, discriminator_real_loss=1.446, discriminator_fake_loss=1.294, generator_loss=29.02, generator_mel_loss=20.74, generator_kl_loss=1.877, generator_dur_loss=1.698, generator_adv_loss=2.016, generator_feat_match_loss=2.691, over 53.00 samples.], tot_loss[discriminator_loss=2.694, discriminator_real_loss=1.401, discriminator_fake_loss=1.293, generator_loss=30.13, generator_mel_loss=21.15, generator_kl_loss=1.929, generator_dur_loss=1.683, generator_adv_loss=2.106, generator_feat_match_loss=3.268, over 2359.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:29:59,119 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 16:30:10,578 INFO [train.py:517] (3/4) Epoch 330, validation: discriminator_loss=2.728, discriminator_real_loss=1.36, discriminator_fake_loss=1.368, generator_loss=30.56, generator_mel_loss=22.03, generator_kl_loss=2.067, generator_dur_loss=1.651, generator_adv_loss=1.844, generator_feat_match_loss=2.969, over 100.00 samples.
2023-11-13 16:30:10,579 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 16:30:57,029 INFO [train.py:811] (3/4) Start epoch 331
2023-11-13 16:34:26,054 INFO [train.py:811] (3/4) Start epoch 332
2023-11-13 16:34:55,293 INFO [train.py:467] (3/4) Epoch 332, batch 3, global_batch_idx: 12250, batch size: 56, loss[discriminator_loss=2.701, discriminator_real_loss=1.425, discriminator_fake_loss=1.276, generator_loss=30.45, generator_mel_loss=21.84, generator_kl_loss=1.932, generator_dur_loss=1.677, generator_adv_loss=2.039, generator_feat_match_loss=2.961, over 56.00 samples.], tot_loss[discriminator_loss=2.715, discriminator_real_loss=1.381, discriminator_fake_loss=1.334, generator_loss=30.23, generator_mel_loss=21.5, generator_kl_loss=1.942, generator_dur_loss=1.68, generator_adv_loss=2.056, generator_feat_match_loss=3.044, over 223.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:37:57,071 INFO [train.py:811] (3/4) Start epoch 333
2023-11-13 16:39:39,122 INFO [train.py:467] (3/4) Epoch 333, batch 16, global_batch_idx: 12300, batch size: 49, loss[discriminator_loss=2.637, discriminator_real_loss=1.326, discriminator_fake_loss=1.311, generator_loss=30.54, generator_mel_loss=21.39, generator_kl_loss=1.921, generator_dur_loss=1.699, generator_adv_loss=2.283, generator_feat_match_loss=3.244, over 49.00 samples.], tot_loss[discriminator_loss=2.634, discriminator_real_loss=1.327, discriminator_fake_loss=1.307, generator_loss=30.42, generator_mel_loss=21.32, generator_kl_loss=1.918, generator_dur_loss=1.68, generator_adv_loss=2.153, generator_feat_match_loss=3.347, over 1116.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:41:36,065 INFO [train.py:811] (3/4) Start epoch 334
2023-11-13 16:44:27,079 INFO [train.py:467] (3/4) Epoch 334, batch 29, global_batch_idx: 12350, batch size: 101, loss[discriminator_loss=3.074, discriminator_real_loss=1.524, discriminator_fake_loss=1.55, generator_loss=29.54, generator_mel_loss=21.03, generator_kl_loss=1.947, generator_dur_loss=1.691, generator_adv_loss=2.035, generator_feat_match_loss=2.832, over 101.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.302, discriminator_fake_loss=1.279, generator_loss=30.69, generator_mel_loss=21.09, generator_kl_loss=1.934, generator_dur_loss=1.684, generator_adv_loss=2.295, generator_feat_match_loss=3.691, over 2274.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:45:06,662 INFO [train.py:811] (3/4) Start epoch 335
2023-11-13 16:48:44,359 INFO [train.py:811] (3/4) Start epoch 336
2023-11-13 16:49:29,999 INFO [train.py:467] (3/4) Epoch 336, batch 5, global_batch_idx: 12400, batch size: 64, loss[discriminator_loss=2.666, discriminator_real_loss=1.278, discriminator_fake_loss=1.388, generator_loss=30.19, generator_mel_loss=21.16, generator_kl_loss=1.956, generator_dur_loss=1.654, generator_adv_loss=2.217, generator_feat_match_loss=3.207, over 64.00 samples.], tot_loss[discriminator_loss=2.69, discriminator_real_loss=1.329, discriminator_fake_loss=1.361, generator_loss=30.21, generator_mel_loss=21.2, generator_kl_loss=1.934, generator_dur_loss=1.679, generator_adv_loss=2.147, generator_feat_match_loss=3.245, over 459.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 32.0
2023-11-13 16:49:30,001 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 16:49:42,794 INFO [train.py:517] (3/4) Epoch 336, validation: discriminator_loss=2.72, discriminator_real_loss=1.378, discriminator_fake_loss=1.342, generator_loss=30.43, generator_mel_loss=21.91, generator_kl_loss=2.01, generator_dur_loss=1.664, generator_adv_loss=1.865, generator_feat_match_loss=2.978, over 100.00 samples.
2023-11-13 16:49:42,797 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 16:52:32,927 INFO [train.py:811] (3/4) Start epoch 337
2023-11-13 16:54:26,078 INFO [train.py:467] (3/4) Epoch 337, batch 18, global_batch_idx: 12450, batch size: 61, loss[discriminator_loss=2.539, discriminator_real_loss=1.174, discriminator_fake_loss=1.366, generator_loss=30.08, generator_mel_loss=20.59, generator_kl_loss=1.921, generator_dur_loss=1.695, generator_adv_loss=2.182, generator_feat_match_loss=3.688, over 61.00 samples.], tot_loss[discriminator_loss=2.614, discriminator_real_loss=1.318, discriminator_fake_loss=1.297, generator_loss=30.59, generator_mel_loss=21.28, generator_kl_loss=1.923, generator_dur_loss=1.69, generator_adv_loss=2.203, generator_feat_match_loss=3.493, over 1198.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:56:04,584 INFO [train.py:811] (3/4) Start epoch 338
2023-11-13 16:59:07,629 INFO [train.py:467] (3/4) Epoch 338, batch 31, global_batch_idx: 12500, batch size: 52, loss[discriminator_loss=2.686, discriminator_real_loss=1.396, discriminator_fake_loss=1.289, generator_loss=29.56, generator_mel_loss=21.09, generator_kl_loss=1.903, generator_dur_loss=1.684, generator_adv_loss=1.928, generator_feat_match_loss=2.955, over 52.00 samples.], tot_loss[discriminator_loss=2.627, discriminator_real_loss=1.322, discriminator_fake_loss=1.305, generator_loss=30.3, generator_mel_loss=21.19, generator_kl_loss=1.953, generator_dur_loss=1.683, generator_adv_loss=2.121, generator_feat_match_loss=3.358, over 2324.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 16:59:39,622 INFO [train.py:811] (3/4) Start epoch 339
2023-11-13 17:03:14,355 INFO [train.py:811] (3/4) Start epoch 340
2023-11-13 17:04:11,818 INFO [train.py:467] (3/4) Epoch 340, batch 7, global_batch_idx: 12550, batch size: 51, loss[discriminator_loss=2.582, discriminator_real_loss=1.35, discriminator_fake_loss=1.233, generator_loss=30.51, generator_mel_loss=21.4, generator_kl_loss=1.833, generator_dur_loss=1.661, generator_adv_loss=2.014, generator_feat_match_loss=3.602, over 51.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.337, discriminator_fake_loss=1.286, generator_loss=30.57, generator_mel_loss=21.27, generator_kl_loss=1.94, generator_dur_loss=1.682, generator_adv_loss=2.151, generator_feat_match_loss=3.527, over 516.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:06:48,017 INFO [train.py:811] (3/4) Start epoch 341
2023-11-13 17:09:05,255 INFO [train.py:467] (3/4) Epoch 341, batch 20, global_batch_idx: 12600, batch size: 50, loss[discriminator_loss=2.637, discriminator_real_loss=1.428, discriminator_fake_loss=1.209, generator_loss=31.19, generator_mel_loss=21.69, generator_kl_loss=1.915, generator_dur_loss=1.71, generator_adv_loss=2.375, generator_feat_match_loss=3.5, over 50.00 samples.], tot_loss[discriminator_loss=2.658, discriminator_real_loss=1.328, discriminator_fake_loss=1.33, generator_loss=30.5, generator_mel_loss=21.29, generator_kl_loss=1.933, generator_dur_loss=1.682, generator_adv_loss=2.177, generator_feat_match_loss=3.415, over 1636.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:09:05,256 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 17:09:16,643 INFO [train.py:517] (3/4) Epoch 341, validation: discriminator_loss=2.477, discriminator_real_loss=1.291, discriminator_fake_loss=1.186, generator_loss=31.98, generator_mel_loss=22.19, generator_kl_loss=2.086, generator_dur_loss=1.654, generator_adv_loss=2.258, generator_feat_match_loss=3.797, over 100.00 samples.
2023-11-13 17:09:16,644 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27299MB
2023-11-13 17:10:39,549 INFO [train.py:811] (3/4) Start epoch 342
2023-11-13 17:13:57,602 INFO [train.py:467] (3/4) Epoch 342, batch 33, global_batch_idx: 12650, batch size: 53, loss[discriminator_loss=2.383, discriminator_real_loss=1.226, discriminator_fake_loss=1.157, generator_loss=31.09, generator_mel_loss=20.89, generator_kl_loss=1.931, generator_dur_loss=1.683, generator_adv_loss=2.369, generator_feat_match_loss=4.219, over 53.00 samples.], tot_loss[discriminator_loss=2.654, discriminator_real_loss=1.343, discriminator_fake_loss=1.31, generator_loss=30.7, generator_mel_loss=21.27, generator_kl_loss=1.955, generator_dur_loss=1.676, generator_adv_loss=2.222, generator_feat_match_loss=3.573, over 2455.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:14:14,251 INFO [train.py:811] (3/4) Start epoch 343
2023-11-13 17:17:50,677 INFO [train.py:811] (3/4) Start epoch 344
2023-11-13 17:18:55,275 INFO [train.py:467] (3/4) Epoch 344, batch 9, global_batch_idx: 12700, batch size: 126, loss[discriminator_loss=2.689, discriminator_real_loss=1.252, discriminator_fake_loss=1.438, generator_loss=30.96, generator_mel_loss=21.53, generator_kl_loss=2.02, generator_dur_loss=1.709, generator_adv_loss=2.295, generator_feat_match_loss=3.408, over 126.00 samples.], tot_loss[discriminator_loss=2.608, discriminator_real_loss=1.291, discriminator_fake_loss=1.317, generator_loss=30.74, generator_mel_loss=21.38, generator_kl_loss=1.96, generator_dur_loss=1.685, generator_adv_loss=2.182, generator_feat_match_loss=3.529, over 819.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:21:26,536 INFO [train.py:811] (3/4) Start epoch 345
2023-11-13 17:23:36,779 INFO [train.py:467] (3/4) Epoch 345, batch 22, global_batch_idx: 12750, batch size: 76, loss[discriminator_loss=2.75, discriminator_real_loss=1.387, discriminator_fake_loss=1.362, generator_loss=30.18, generator_mel_loss=21.19, generator_kl_loss=2.048, generator_dur_loss=1.692, generator_adv_loss=2.023, generator_feat_match_loss=3.229, over 76.00 samples.], tot_loss[discriminator_loss=2.645, discriminator_real_loss=1.342, discriminator_fake_loss=1.303, generator_loss=30.37, generator_mel_loss=21.2, generator_kl_loss=1.94, generator_dur_loss=1.68, generator_adv_loss=2.145, generator_feat_match_loss=3.412, over 1798.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:25:02,267 INFO [train.py:811] (3/4) Start epoch 346
2023-11-13 17:28:34,560 INFO [train.py:467] (3/4) Epoch 346, batch 35, global_batch_idx: 12800, batch size: 52, loss[discriminator_loss=2.594, discriminator_real_loss=1.19, discriminator_fake_loss=1.404, generator_loss=31.14, generator_mel_loss=21.68, generator_kl_loss=1.94, generator_dur_loss=1.66, generator_adv_loss=2.281, generator_feat_match_loss=3.578, over 52.00 samples.], tot_loss[discriminator_loss=2.653, discriminator_real_loss=1.345, discriminator_fake_loss=1.308, generator_loss=30.4, generator_mel_loss=21.18, generator_kl_loss=1.934, generator_dur_loss=1.68, generator_adv_loss=2.151, generator_feat_match_loss=3.453, over 2802.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 32.0
2023-11-13 17:28:34,562 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 17:28:46,963 INFO [train.py:517] (3/4) Epoch 346, validation: discriminator_loss=2.767, discriminator_real_loss=1.383, discriminator_fake_loss=1.384, generator_loss=30.62, generator_mel_loss=22.02, generator_kl_loss=2.037, generator_dur_loss=1.65, generator_adv_loss=1.855, generator_feat_match_loss=3.06, over 100.00 samples.
2023-11-13 17:28:46,964 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 17:28:52,282 INFO [train.py:811] (3/4) Start epoch 347
2023-11-13 17:32:24,597 INFO [train.py:811] (3/4) Start epoch 348
2023-11-13 17:33:38,052 INFO [train.py:467] (3/4) Epoch 348, batch 11, global_batch_idx: 12850, batch size: 71, loss[discriminator_loss=2.416, discriminator_real_loss=1.229, discriminator_fake_loss=1.187, generator_loss=31.34, generator_mel_loss=21.13, generator_kl_loss=1.946, generator_dur_loss=1.664, generator_adv_loss=2.395, generator_feat_match_loss=4.211, over 71.00 samples.], tot_loss[discriminator_loss=2.569, discriminator_real_loss=1.321, discriminator_fake_loss=1.248, generator_loss=30.77, generator_mel_loss=21.1, generator_kl_loss=1.963, generator_dur_loss=1.687, generator_adv_loss=2.321, generator_feat_match_loss=3.692, over 817.00 samples.], cur_lr_g: 1.92e-04, cur_lr_d: 1.92e-04, grad_scale: 16.0
2023-11-13 17:35:54,373 INFO [train.py:811] (3/4) Start epoch 349
2023-11-13 17:38:21,651 INFO [train.py:467] (3/4) Epoch 349, batch 24, global_batch_idx: 12900, batch size: 58, loss[discriminator_loss=2.703, discriminator_real_loss=1.246, discriminator_fake_loss=1.458, generator_loss=29.08, generator_mel_loss=20.65, generator_kl_loss=1.849, generator_dur_loss=1.687, generator_adv_loss=1.943, generator_feat_match_loss=2.955, over 58.00 samples.], tot_loss[discriminator_loss=2.694, discriminator_real_loss=1.372, discriminator_fake_loss=1.322, generator_loss=30.08, generator_mel_loss=21.2, generator_kl_loss=1.946, generator_dur_loss=1.678, generator_adv_loss=2.049, generator_feat_match_loss=3.201, over 1975.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 17:39:22,725 INFO [train.py:811] (3/4) Start epoch 350
2023-11-13 17:42:58,569 INFO [train.py:811] (3/4) Start epoch 351
2023-11-13 17:43:14,311 INFO [train.py:467] (3/4) Epoch 351, batch 0, global_batch_idx: 12950, batch size: 90, loss[discriminator_loss=2.752, discriminator_real_loss=1.332, discriminator_fake_loss=1.42, generator_loss=31.02, generator_mel_loss=22.08, generator_kl_loss=1.99, generator_dur_loss=1.704, generator_adv_loss=1.94, generator_feat_match_loss=3.305, over 90.00 samples.], tot_loss[discriminator_loss=2.752, discriminator_real_loss=1.332, discriminator_fake_loss=1.42, generator_loss=31.02, generator_mel_loss=22.08, generator_kl_loss=1.99, generator_dur_loss=1.704, generator_adv_loss=1.94, generator_feat_match_loss=3.305, over 90.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 17:46:33,229 INFO [train.py:811] (3/4) Start epoch 352
2023-11-13 17:47:51,774 INFO [train.py:467] (3/4) Epoch 352, batch 13, global_batch_idx: 13000, batch size: 52, loss[discriminator_loss=2.736, discriminator_real_loss=1.205, discriminator_fake_loss=1.531, generator_loss=29.8, generator_mel_loss=21.04, generator_kl_loss=1.908, generator_dur_loss=1.674, generator_adv_loss=1.853, generator_feat_match_loss=3.334, over 52.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.335, discriminator_fake_loss=1.313, generator_loss=30.46, generator_mel_loss=21.21, generator_kl_loss=1.955, generator_dur_loss=1.675, generator_adv_loss=2.161, generator_feat_match_loss=3.458, over 962.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 17:47:51,776 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 17:48:03,375 INFO [train.py:517] (3/4) Epoch 352, validation: discriminator_loss=2.671, discriminator_real_loss=1.116, discriminator_fake_loss=1.555, generator_loss=30.09, generator_mel_loss=21.66, generator_kl_loss=2.014, generator_dur_loss=1.65, generator_adv_loss=1.618, generator_feat_match_loss=3.148, over 100.00 samples.
2023-11-13 17:48:03,376 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 17:50:17,701 INFO [train.py:811] (3/4) Start epoch 353
2023-11-13 17:52:59,109 INFO [train.py:467] (3/4) Epoch 353, batch 26, global_batch_idx: 13050, batch size: 61, loss[discriminator_loss=2.906, discriminator_real_loss=1.264, discriminator_fake_loss=1.642, generator_loss=29.63, generator_mel_loss=21.21, generator_kl_loss=1.929, generator_dur_loss=1.672, generator_adv_loss=2.092, generator_feat_match_loss=2.723, over 61.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.357, discriminator_fake_loss=1.291, generator_loss=30.51, generator_mel_loss=21.16, generator_kl_loss=1.938, generator_dur_loss=1.684, generator_adv_loss=2.188, generator_feat_match_loss=3.541, over 1814.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 17:53:53,783 INFO [train.py:811] (3/4) Start epoch 354
2023-11-13 17:57:27,553 INFO [train.py:811] (3/4) Start epoch 355
2023-11-13 17:57:57,724 INFO [train.py:467] (3/4) Epoch 355, batch 2, global_batch_idx: 13100, batch size: 101, loss[discriminator_loss=2.641, discriminator_real_loss=1.337, discriminator_fake_loss=1.303, generator_loss=30.48, generator_mel_loss=21.32, generator_kl_loss=1.925, generator_dur_loss=1.661, generator_adv_loss=2.035, generator_feat_match_loss=3.545, over 101.00 samples.], tot_loss[discriminator_loss=2.688, discriminator_real_loss=1.364, discriminator_fake_loss=1.324, generator_loss=30.08, generator_mel_loss=21.21, generator_kl_loss=1.954, generator_dur_loss=1.674, generator_adv_loss=1.966, generator_feat_match_loss=3.276, over 209.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:00:58,652 INFO [train.py:811] (3/4) Start epoch 356
2023-11-13 18:02:38,822 INFO [train.py:467] (3/4) Epoch 356, batch 15, global_batch_idx: 13150, batch size: 51, loss[discriminator_loss=2.615, discriminator_real_loss=1.394, discriminator_fake_loss=1.222, generator_loss=30.15, generator_mel_loss=21.05, generator_kl_loss=1.969, generator_dur_loss=1.686, generator_adv_loss=2.197, generator_feat_match_loss=3.256, over 51.00 samples.], tot_loss[discriminator_loss=2.638, discriminator_real_loss=1.347, discriminator_fake_loss=1.292, generator_loss=30.27, generator_mel_loss=21.23, generator_kl_loss=1.934, generator_dur_loss=1.677, generator_adv_loss=2.1, generator_feat_match_loss=3.332, over 1117.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:04:30,433 INFO [train.py:811] (3/4) Start epoch 357
2023-11-13 18:07:13,735 INFO [train.py:467] (3/4) Epoch 357, batch 28, global_batch_idx: 13200, batch size: 76, loss[discriminator_loss=2.572, discriminator_real_loss=1.199, discriminator_fake_loss=1.373, generator_loss=30.68, generator_mel_loss=21.08, generator_kl_loss=1.933, generator_dur_loss=1.677, generator_adv_loss=2.311, generator_feat_match_loss=3.68, over 76.00 samples.], tot_loss[discriminator_loss=2.659, discriminator_real_loss=1.338, discriminator_fake_loss=1.321, generator_loss=30.3, generator_mel_loss=21.17, generator_kl_loss=1.938, generator_dur_loss=1.678, generator_adv_loss=2.112, generator_feat_match_loss=3.397, over 2165.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 32.0
2023-11-13 18:07:13,737 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 18:07:25,632 INFO [train.py:517] (3/4) Epoch 357, validation: discriminator_loss=2.6, discriminator_real_loss=1.316, discriminator_fake_loss=1.283, generator_loss=31.01, generator_mel_loss=21.87, generator_kl_loss=2.01, generator_dur_loss=1.656, generator_adv_loss=1.989, generator_feat_match_loss=3.487, over 100.00 samples.
2023-11-13 18:07:25,639 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 18:08:08,189 INFO [train.py:811] (3/4) Start epoch 358
2023-11-13 18:11:43,363 INFO [train.py:811] (3/4) Start epoch 359
2023-11-13 18:12:22,368 INFO [train.py:467] (3/4) Epoch 359, batch 4, global_batch_idx: 13250, batch size: 52, loss[discriminator_loss=3.098, discriminator_real_loss=1.677, discriminator_fake_loss=1.421, generator_loss=31.31, generator_mel_loss=21.75, generator_kl_loss=1.916, generator_dur_loss=1.702, generator_adv_loss=2.465, generator_feat_match_loss=3.48, over 52.00 samples.], tot_loss[discriminator_loss=2.777, discriminator_real_loss=1.408, discriminator_fake_loss=1.369, generator_loss=31, generator_mel_loss=21.4, generator_kl_loss=1.986, generator_dur_loss=1.68, generator_adv_loss=2.371, generator_feat_match_loss=3.558, over 382.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:15:07,855 INFO [train.py:811] (3/4) Start epoch 360
2023-11-13 18:17:09,756 INFO [train.py:467] (3/4) Epoch 360, batch 17, global_batch_idx: 13300, batch size: 90, loss[discriminator_loss=2.535, discriminator_real_loss=1.299, discriminator_fake_loss=1.237, generator_loss=30.85, generator_mel_loss=21.38, generator_kl_loss=2, generator_dur_loss=1.686, generator_adv_loss=2.303, generator_feat_match_loss=3.484, over 90.00 samples.], tot_loss[discriminator_loss=2.627, discriminator_real_loss=1.34, discriminator_fake_loss=1.287, generator_loss=30.55, generator_mel_loss=21.1, generator_kl_loss=1.949, generator_dur_loss=1.684, generator_adv_loss=2.235, generator_feat_match_loss=3.587, over 1419.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:18:47,597 INFO [train.py:811] (3/4) Start epoch 361
2023-11-13 18:21:58,707 INFO [train.py:467] (3/4) Epoch 361, batch 30, global_batch_idx: 13350, batch size: 65, loss[discriminator_loss=2.604, discriminator_real_loss=1.131, discriminator_fake_loss=1.473, generator_loss=30.58, generator_mel_loss=20.92, generator_kl_loss=1.978, generator_dur_loss=1.683, generator_adv_loss=2.217, generator_feat_match_loss=3.781, over 65.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.333, discriminator_fake_loss=1.294, generator_loss=30.66, generator_mel_loss=21.22, generator_kl_loss=1.956, generator_dur_loss=1.676, generator_adv_loss=2.183, generator_feat_match_loss=3.621, over 2238.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:22:31,755 INFO [train.py:811] (3/4) Start epoch 362
2023-11-13 18:26:08,358 INFO [train.py:811] (3/4) Start epoch 363
2023-11-13 18:26:49,616 INFO [train.py:467] (3/4) Epoch 363, batch 6, global_batch_idx: 13400, batch size: 67, loss[discriminator_loss=2.711, discriminator_real_loss=1.308, discriminator_fake_loss=1.403, generator_loss=30.02, generator_mel_loss=21.09, generator_kl_loss=1.925, generator_dur_loss=1.69, generator_adv_loss=2.066, generator_feat_match_loss=3.252, over 67.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.37, discriminator_fake_loss=1.306, generator_loss=30.14, generator_mel_loss=21.19, generator_kl_loss=1.972, generator_dur_loss=1.685, generator_adv_loss=2.073, generator_feat_match_loss=3.225, over 426.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:26:49,618 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 18:27:02,178 INFO [train.py:517] (3/4) Epoch 363, validation: discriminator_loss=2.582, discriminator_real_loss=1.272, discriminator_fake_loss=1.31, generator_loss=30.95, generator_mel_loss=21.97, generator_kl_loss=2.024, generator_dur_loss=1.651, generator_adv_loss=1.956, generator_feat_match_loss=3.342, over 100.00 samples.
2023-11-13 18:27:02,180 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 18:29:58,017 INFO [train.py:811] (3/4) Start epoch 364
2023-11-13 18:32:03,097 INFO [train.py:467] (3/4) Epoch 364, batch 19, global_batch_idx: 13450, batch size: 85, loss[discriminator_loss=2.586, discriminator_real_loss=1.218, discriminator_fake_loss=1.369, generator_loss=30.24, generator_mel_loss=20.95, generator_kl_loss=1.953, generator_dur_loss=1.682, generator_adv_loss=2.004, generator_feat_match_loss=3.648, over 85.00 samples.], tot_loss[discriminator_loss=2.691, discriminator_real_loss=1.393, discriminator_fake_loss=1.298, generator_loss=30.65, generator_mel_loss=21.26, generator_kl_loss=1.949, generator_dur_loss=1.68, generator_adv_loss=2.208, generator_feat_match_loss=3.547, over 1725.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:33:36,443 INFO [train.py:811] (3/4) Start epoch 365
2023-11-13 18:36:45,205 INFO [train.py:467] (3/4) Epoch 365, batch 32, global_batch_idx: 13500, batch size: 63, loss[discriminator_loss=2.551, discriminator_real_loss=1.326, discriminator_fake_loss=1.225, generator_loss=30.93, generator_mel_loss=21.05, generator_kl_loss=1.907, generator_dur_loss=1.675, generator_adv_loss=2.369, generator_feat_match_loss=3.93, over 63.00 samples.], tot_loss[discriminator_loss=2.583, discriminator_real_loss=1.286, discriminator_fake_loss=1.297, generator_loss=30.4, generator_mel_loss=21.07, generator_kl_loss=1.954, generator_dur_loss=1.672, generator_adv_loss=2.153, generator_feat_match_loss=3.558, over 2508.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:37:07,786 INFO [train.py:811] (3/4) Start epoch 366
2023-11-13 18:40:43,280 INFO [train.py:811] (3/4) Start epoch 367
2023-11-13 18:41:44,785 INFO [train.py:467] (3/4) Epoch 367, batch 8, global_batch_idx: 13550, batch size: 79, loss[discriminator_loss=2.508, discriminator_real_loss=1.342, discriminator_fake_loss=1.166, generator_loss=30.34, generator_mel_loss=21, generator_kl_loss=1.912, generator_dur_loss=1.696, generator_adv_loss=2.109, generator_feat_match_loss=3.627, over 79.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.348, discriminator_fake_loss=1.278, generator_loss=29.95, generator_mel_loss=20.85, generator_kl_loss=1.926, generator_dur_loss=1.679, generator_adv_loss=2.087, generator_feat_match_loss=3.403, over 689.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:44:16,131 INFO [train.py:811] (3/4) Start epoch 368
2023-11-13 18:46:36,909 INFO [train.py:467] (3/4) Epoch 368, batch 21, global_batch_idx: 13600, batch size: 64, loss[discriminator_loss=2.588, discriminator_real_loss=1.32, discriminator_fake_loss=1.268, generator_loss=30.43, generator_mel_loss=20.83, generator_kl_loss=1.862, generator_dur_loss=1.658, generator_adv_loss=2.127, generator_feat_match_loss=3.949, over 64.00 samples.], tot_loss[discriminator_loss=2.651, discriminator_real_loss=1.341, discriminator_fake_loss=1.311, generator_loss=30.26, generator_mel_loss=21.1, generator_kl_loss=1.929, generator_dur_loss=1.679, generator_adv_loss=2.131, generator_feat_match_loss=3.417, over 1563.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 32.0
2023-11-13 18:46:36,910 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 18:46:48,969 INFO [train.py:517] (3/4) Epoch 368, validation: discriminator_loss=2.436, discriminator_real_loss=1.103, discriminator_fake_loss=1.333, generator_loss=31.34, generator_mel_loss=21.77, generator_kl_loss=2.029, generator_dur_loss=1.646, generator_adv_loss=1.987, generator_feat_match_loss=3.909, over 100.00 samples.
2023-11-13 18:46:48,970 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 18:48:07,745 INFO [train.py:811] (3/4) Start epoch 369
2023-11-13 18:51:27,690 INFO [train.py:467] (3/4) Epoch 369, batch 34, global_batch_idx: 13650, batch size: 110, loss[discriminator_loss=2.541, discriminator_real_loss=1.262, discriminator_fake_loss=1.279, generator_loss=31.04, generator_mel_loss=21.54, generator_kl_loss=2.003, generator_dur_loss=1.693, generator_adv_loss=2.148, generator_feat_match_loss=3.652, over 110.00 samples.], tot_loss[discriminator_loss=2.646, discriminator_real_loss=1.335, discriminator_fake_loss=1.311, generator_loss=30.34, generator_mel_loss=21.2, generator_kl_loss=1.951, generator_dur_loss=1.675, generator_adv_loss=2.108, generator_feat_match_loss=3.402, over 2804.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:51:41,812 INFO [train.py:811] (3/4) Start epoch 370
2023-11-13 18:55:16,272 INFO [train.py:811] (3/4) Start epoch 371
2023-11-13 18:56:25,869 INFO [train.py:467] (3/4) Epoch 371, batch 10, global_batch_idx: 13700, batch size: 58, loss[discriminator_loss=2.461, discriminator_real_loss=1.235, discriminator_fake_loss=1.227, generator_loss=30.99, generator_mel_loss=20.84, generator_kl_loss=1.859, generator_dur_loss=1.672, generator_adv_loss=2.307, generator_feat_match_loss=4.312, over 58.00 samples.], tot_loss[discriminator_loss=2.57, discriminator_real_loss=1.312, discriminator_fake_loss=1.257, generator_loss=30.49, generator_mel_loss=20.93, generator_kl_loss=1.936, generator_dur_loss=1.672, generator_adv_loss=2.221, generator_feat_match_loss=3.732, over 844.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 18:58:47,467 INFO [train.py:811] (3/4) Start epoch 372
2023-11-13 19:01:09,295 INFO [train.py:467] (3/4) Epoch 372, batch 23, global_batch_idx: 13750, batch size: 101, loss[discriminator_loss=2.572, discriminator_real_loss=1.266, discriminator_fake_loss=1.307, generator_loss=30.62, generator_mel_loss=21.45, generator_kl_loss=1.936, generator_dur_loss=1.651, generator_adv_loss=2.107, generator_feat_match_loss=3.482, over 101.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.297, discriminator_fake_loss=1.277, generator_loss=30.49, generator_mel_loss=21.11, generator_kl_loss=1.944, generator_dur_loss=1.677, generator_adv_loss=2.157, generator_feat_match_loss=3.598, over 1849.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 19:02:20,985 INFO [train.py:811] (3/4) Start epoch 373
2023-11-13 19:05:55,175 INFO [train.py:467] (3/4) Epoch 373, batch 36, global_batch_idx: 13800, batch size: 110, loss[discriminator_loss=2.508, discriminator_real_loss=1.33, discriminator_fake_loss=1.178, generator_loss=30.48, generator_mel_loss=20.69, generator_kl_loss=2.016, generator_dur_loss=1.646, generator_adv_loss=2.334, generator_feat_match_loss=3.787, over 110.00 samples.], tot_loss[discriminator_loss=2.621, discriminator_real_loss=1.333, discriminator_fake_loss=1.287, generator_loss=30.59, generator_mel_loss=21.02, generator_kl_loss=1.949, generator_dur_loss=1.676, generator_adv_loss=2.249, generator_feat_match_loss=3.692, over 2602.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:05:55,177 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 19:06:07,434 INFO [train.py:517] (3/4) Epoch 373, validation: discriminator_loss=2.515, discriminator_real_loss=1.122, discriminator_fake_loss=1.393, generator_loss=31.29, generator_mel_loss=21.7, generator_kl_loss=2.09, generator_dur_loss=1.652, generator_adv_loss=2.007, generator_feat_match_loss=3.836, over 100.00 samples.
2023-11-13 19:06:07,435 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 19:06:08,264 INFO [train.py:811] (3/4) Start epoch 374
2023-11-13 19:09:39,026 INFO [train.py:811] (3/4) Start epoch 375
2023-11-13 19:10:53,230 INFO [train.py:467] (3/4) Epoch 375, batch 12, global_batch_idx: 13850, batch size: 85, loss[discriminator_loss=2.586, discriminator_real_loss=1.252, discriminator_fake_loss=1.335, generator_loss=30.77, generator_mel_loss=21.15, generator_kl_loss=1.951, generator_dur_loss=1.666, generator_adv_loss=2.316, generator_feat_match_loss=3.684, over 85.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.333, discriminator_fake_loss=1.286, generator_loss=30.38, generator_mel_loss=21.15, generator_kl_loss=1.96, generator_dur_loss=1.687, generator_adv_loss=2.134, generator_feat_match_loss=3.455, over 917.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:13:09,771 INFO [train.py:811] (3/4) Start epoch 376
2023-11-13 19:15:39,001 INFO [train.py:467] (3/4) Epoch 376, batch 25, global_batch_idx: 13900, batch size: 60, loss[discriminator_loss=2.844, discriminator_real_loss=1.574, discriminator_fake_loss=1.269, generator_loss=30.73, generator_mel_loss=21.34, generator_kl_loss=2.017, generator_dur_loss=1.686, generator_adv_loss=2.248, generator_feat_match_loss=3.438, over 60.00 samples.], tot_loss[discriminator_loss=2.651, discriminator_real_loss=1.344, discriminator_fake_loss=1.307, generator_loss=30.67, generator_mel_loss=21.2, generator_kl_loss=1.959, generator_dur_loss=1.678, generator_adv_loss=2.237, generator_feat_match_loss=3.597, over 1997.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:16:40,627 INFO [train.py:811] (3/4) Start epoch 377
2023-11-13 19:20:13,327 INFO [train.py:811] (3/4) Start epoch 378
2023-11-13 19:20:36,051 INFO [train.py:467] (3/4) Epoch 378, batch 1, global_batch_idx: 13950, batch size: 52, loss[discriminator_loss=2.65, discriminator_real_loss=1.373, discriminator_fake_loss=1.277, generator_loss=30.22, generator_mel_loss=21.09, generator_kl_loss=1.916, generator_dur_loss=1.68, generator_adv_loss=2.062, generator_feat_match_loss=3.473, over 52.00 samples.], tot_loss[discriminator_loss=2.664, discriminator_real_loss=1.43, discriminator_fake_loss=1.234, generator_loss=30.2, generator_mel_loss=21.13, generator_kl_loss=1.879, generator_dur_loss=1.691, generator_adv_loss=2.068, generator_feat_match_loss=3.435, over 107.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:23:53,661 INFO [train.py:811] (3/4) Start epoch 379
2023-11-13 19:25:22,646 INFO [train.py:467] (3/4) Epoch 379, batch 14, global_batch_idx: 14000, batch size: 76, loss[discriminator_loss=2.803, discriminator_real_loss=1.328, discriminator_fake_loss=1.475, generator_loss=30.2, generator_mel_loss=21.15, generator_kl_loss=1.89, generator_dur_loss=1.683, generator_adv_loss=2.189, generator_feat_match_loss=3.285, over 76.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.318, discriminator_fake_loss=1.298, generator_loss=30.35, generator_mel_loss=21.09, generator_kl_loss=1.955, generator_dur_loss=1.681, generator_adv_loss=2.134, generator_feat_match_loss=3.489, over 1017.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 19:25:22,648 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 19:25:34,172 INFO [train.py:517] (3/4) Epoch 379, validation: discriminator_loss=2.685, discriminator_real_loss=1.545, discriminator_fake_loss=1.14, generator_loss=31.28, generator_mel_loss=21.94, generator_kl_loss=2.153, generator_dur_loss=1.648, generator_adv_loss=2.175, generator_feat_match_loss=3.365, over 100.00 samples.
2023-11-13 19:25:34,175 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 19:27:37,090 INFO [train.py:811] (3/4) Start epoch 380
2023-11-13 19:30:11,334 INFO [train.py:467] (3/4) Epoch 380, batch 27, global_batch_idx: 14050, batch size: 79, loss[discriminator_loss=2.674, discriminator_real_loss=1.462, discriminator_fake_loss=1.212, generator_loss=30.08, generator_mel_loss=20.75, generator_kl_loss=1.981, generator_dur_loss=1.69, generator_adv_loss=2.328, generator_feat_match_loss=3.334, over 79.00 samples.], tot_loss[discriminator_loss=2.677, discriminator_real_loss=1.361, discriminator_fake_loss=1.316, generator_loss=30.51, generator_mel_loss=21.25, generator_kl_loss=1.94, generator_dur_loss=1.683, generator_adv_loss=2.189, generator_feat_match_loss=3.45, over 1902.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 19:31:03,569 INFO [train.py:811] (3/4) Start epoch 381
2023-11-13 19:34:38,451 INFO [train.py:811] (3/4) Start epoch 382
2023-11-13 19:35:11,628 INFO [train.py:467] (3/4) Epoch 382, batch 3, global_batch_idx: 14100, batch size: 59, loss[discriminator_loss=2.52, discriminator_real_loss=1.326, discriminator_fake_loss=1.193, generator_loss=31.06, generator_mel_loss=21.32, generator_kl_loss=1.907, generator_dur_loss=1.682, generator_adv_loss=2.215, generator_feat_match_loss=3.939, over 59.00 samples.], tot_loss[discriminator_loss=2.641, discriminator_real_loss=1.34, discriminator_fake_loss=1.301, generator_loss=30.55, generator_mel_loss=21.25, generator_kl_loss=1.956, generator_dur_loss=1.675, generator_adv_loss=2.095, generator_feat_match_loss=3.577, over 351.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 19:38:13,079 INFO [train.py:811] (3/4) Start epoch 383
2023-11-13 19:40:00,813 INFO [train.py:467] (3/4) Epoch 383, batch 16, global_batch_idx: 14150, batch size: 63, loss[discriminator_loss=2.633, discriminator_real_loss=1.291, discriminator_fake_loss=1.343, generator_loss=30.15, generator_mel_loss=20.45, generator_kl_loss=1.985, generator_dur_loss=1.703, generator_adv_loss=2.174, generator_feat_match_loss=3.84, over 63.00 samples.], tot_loss[discriminator_loss=2.639, discriminator_real_loss=1.326, discriminator_fake_loss=1.312, generator_loss=30.45, generator_mel_loss=20.99, generator_kl_loss=1.96, generator_dur_loss=1.672, generator_adv_loss=2.213, generator_feat_match_loss=3.617, over 1273.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:41:50,402 INFO [train.py:811] (3/4) Start epoch 384
2023-11-13 19:44:47,587 INFO [train.py:467] (3/4) Epoch 384, batch 29, global_batch_idx: 14200, batch size: 73, loss[discriminator_loss=2.699, discriminator_real_loss=1.176, discriminator_fake_loss=1.523, generator_loss=29.99, generator_mel_loss=20.7, generator_kl_loss=1.959, generator_dur_loss=1.694, generator_adv_loss=1.923, generator_feat_match_loss=3.711, over 73.00 samples.], tot_loss[discriminator_loss=2.602, discriminator_real_loss=1.327, discriminator_fake_loss=1.275, generator_loss=30.44, generator_mel_loss=20.88, generator_kl_loss=1.944, generator_dur_loss=1.676, generator_adv_loss=2.201, generator_feat_match_loss=3.74, over 2290.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:44:47,589 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 19:44:59,377 INFO [train.py:517] (3/4) Epoch 384, validation: discriminator_loss=2.806, discriminator_real_loss=1.156, discriminator_fake_loss=1.65, generator_loss=30.87, generator_mel_loss=22.28, generator_kl_loss=1.928, generator_dur_loss=1.651, generator_adv_loss=1.578, generator_feat_match_loss=3.435, over 100.00 samples.
2023-11-13 19:44:59,378 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 19:45:34,896 INFO [train.py:811] (3/4) Start epoch 385
2023-11-13 19:49:10,612 INFO [train.py:811] (3/4) Start epoch 386
2023-11-13 19:49:54,828 INFO [train.py:467] (3/4) Epoch 386, batch 5, global_batch_idx: 14250, batch size: 58, loss[discriminator_loss=2.883, discriminator_real_loss=1.634, discriminator_fake_loss=1.248, generator_loss=30.44, generator_mel_loss=20.83, generator_kl_loss=1.844, generator_dur_loss=1.687, generator_adv_loss=2.512, generator_feat_match_loss=3.564, over 58.00 samples.], tot_loss[discriminator_loss=2.657, discriminator_real_loss=1.383, discriminator_fake_loss=1.273, generator_loss=31.01, generator_mel_loss=21.12, generator_kl_loss=1.956, generator_dur_loss=1.669, generator_adv_loss=2.335, generator_feat_match_loss=3.929, over 458.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:52:48,423 INFO [train.py:811] (3/4) Start epoch 387
2023-11-13 19:54:42,922 INFO [train.py:467] (3/4) Epoch 387, batch 18, global_batch_idx: 14300, batch size: 54, loss[discriminator_loss=2.645, discriminator_real_loss=1.293, discriminator_fake_loss=1.351, generator_loss=29.65, generator_mel_loss=20.7, generator_kl_loss=1.914, generator_dur_loss=1.667, generator_adv_loss=2.223, generator_feat_match_loss=3.148, over 54.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.335, discriminator_fake_loss=1.317, generator_loss=30.04, generator_mel_loss=21.06, generator_kl_loss=1.959, generator_dur_loss=1.675, generator_adv_loss=2.07, generator_feat_match_loss=3.276, over 1252.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:56:22,312 INFO [train.py:811] (3/4) Start epoch 388
2023-11-13 19:59:33,651 INFO [train.py:467] (3/4) Epoch 388, batch 31, global_batch_idx: 14350, batch size: 52, loss[discriminator_loss=2.473, discriminator_real_loss=1.371, discriminator_fake_loss=1.101, generator_loss=31.04, generator_mel_loss=20.94, generator_kl_loss=1.846, generator_dur_loss=1.644, generator_adv_loss=2.426, generator_feat_match_loss=4.184, over 52.00 samples.], tot_loss[discriminator_loss=2.643, discriminator_real_loss=1.355, discriminator_fake_loss=1.288, generator_loss=30.47, generator_mel_loss=20.98, generator_kl_loss=1.958, generator_dur_loss=1.675, generator_adv_loss=2.217, generator_feat_match_loss=3.647, over 2384.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 8.0
2023-11-13 19:59:56,134 INFO [train.py:811] (3/4) Start epoch 389
2023-11-13 20:03:27,351 INFO [train.py:811] (3/4) Start epoch 390
2023-11-13 20:04:17,209 INFO [train.py:467] (3/4) Epoch 390, batch 7, global_batch_idx: 14400, batch size: 53, loss[discriminator_loss=2.734, discriminator_real_loss=1.537, discriminator_fake_loss=1.196, generator_loss=29.59, generator_mel_loss=20.87, generator_kl_loss=1.946, generator_dur_loss=1.673, generator_adv_loss=2.1, generator_feat_match_loss=2.996, over 53.00 samples.], tot_loss[discriminator_loss=2.699, discriminator_real_loss=1.368, discriminator_fake_loss=1.33, generator_loss=30.16, generator_mel_loss=21.11, generator_kl_loss=1.947, generator_dur_loss=1.68, generator_adv_loss=2.118, generator_feat_match_loss=3.297, over 567.00 samples.], cur_lr_g: 1.91e-04, cur_lr_d: 1.91e-04, grad_scale: 16.0
2023-11-13 20:04:17,211 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 20:04:28,951 INFO [train.py:517] (3/4) Epoch 390, validation: discriminator_loss=2.847, discriminator_real_loss=1.415, discriminator_fake_loss=1.432, generator_loss=30.63, generator_mel_loss=21.9, generator_kl_loss=2.055, generator_dur_loss=1.644, generator_adv_loss=1.844, generator_feat_match_loss=3.191, over 100.00 samples.
2023-11-13 20:04:28,952 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 20:07:11,195 INFO [train.py:811] (3/4) Start epoch 391
2023-11-13 20:09:14,886 INFO [train.py:467] (3/4) Epoch 391, batch 20, global_batch_idx: 14450, batch size: 49, loss[discriminator_loss=2.639, discriminator_real_loss=1.395, discriminator_fake_loss=1.244, generator_loss=30.95, generator_mel_loss=21.53, generator_kl_loss=2.048, generator_dur_loss=1.678, generator_adv_loss=2.129, generator_feat_match_loss=3.564, over 49.00 samples.], tot_loss[discriminator_loss=2.664, discriminator_real_loss=1.347, discriminator_fake_loss=1.317, generator_loss=30.16, generator_mel_loss=20.99, generator_kl_loss=1.95, generator_dur_loss=1.674, generator_adv_loss=2.121, generator_feat_match_loss=3.43, over 1377.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:10:43,975 INFO [train.py:811] (3/4) Start epoch 392
2023-11-13 20:13:52,963 INFO [train.py:467] (3/4) Epoch 392, batch 33, global_batch_idx: 14500, batch size: 101, loss[discriminator_loss=2.68, discriminator_real_loss=1.542, discriminator_fake_loss=1.137, generator_loss=30.55, generator_mel_loss=21.23, generator_kl_loss=1.979, generator_dur_loss=1.66, generator_adv_loss=1.908, generator_feat_match_loss=3.773, over 101.00 samples.], tot_loss[discriminator_loss=2.673, discriminator_real_loss=1.356, discriminator_fake_loss=1.316, generator_loss=30.41, generator_mel_loss=21.28, generator_kl_loss=1.942, generator_dur_loss=1.674, generator_adv_loss=2.114, generator_feat_match_loss=3.406, over 2702.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:14:14,856 INFO [train.py:811] (3/4) Start epoch 393
2023-11-13 20:17:40,874 INFO [train.py:811] (3/4) Start epoch 394
2023-11-13 20:18:50,347 INFO [train.py:467] (3/4) Epoch 394, batch 9, global_batch_idx: 14550, batch size: 95, loss[discriminator_loss=2.582, discriminator_real_loss=1.242, discriminator_fake_loss=1.34, generator_loss=30.53, generator_mel_loss=21.21, generator_kl_loss=2.018, generator_dur_loss=1.654, generator_adv_loss=2.15, generator_feat_match_loss=3.502, over 95.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.33, discriminator_fake_loss=1.288, generator_loss=30.27, generator_mel_loss=21.03, generator_kl_loss=1.945, generator_dur_loss=1.667, generator_adv_loss=2.125, generator_feat_match_loss=3.512, over 772.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:21:12,306 INFO [train.py:811] (3/4) Start epoch 395
2023-11-13 20:23:27,800 INFO [train.py:467] (3/4) Epoch 395, batch 22, global_batch_idx: 14600, batch size: 61, loss[discriminator_loss=2.527, discriminator_real_loss=1.263, discriminator_fake_loss=1.264, generator_loss=30.56, generator_mel_loss=20.95, generator_kl_loss=1.963, generator_dur_loss=1.666, generator_adv_loss=2.307, generator_feat_match_loss=3.674, over 61.00 samples.], tot_loss[discriminator_loss=2.572, discriminator_real_loss=1.292, discriminator_fake_loss=1.28, generator_loss=30.51, generator_mel_loss=21.03, generator_kl_loss=1.969, generator_dur_loss=1.675, generator_adv_loss=2.19, generator_feat_match_loss=3.645, over 1596.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:23:27,802 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 20:23:39,072 INFO [train.py:517] (3/4) Epoch 395, validation: discriminator_loss=2.607, discriminator_real_loss=1.199, discriminator_fake_loss=1.408, generator_loss=31.02, generator_mel_loss=21.93, generator_kl_loss=2.106, generator_dur_loss=1.653, generator_adv_loss=1.804, generator_feat_match_loss=3.531, over 100.00 samples.
2023-11-13 20:23:39,073 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 20:24:57,090 INFO [train.py:811] (3/4) Start epoch 396
2023-11-13 20:28:20,214 INFO [train.py:467] (3/4) Epoch 396, batch 35, global_batch_idx: 14650, batch size: 79, loss[discriminator_loss=2.609, discriminator_real_loss=1.472, discriminator_fake_loss=1.137, generator_loss=30.03, generator_mel_loss=21.06, generator_kl_loss=1.911, generator_dur_loss=1.652, generator_adv_loss=2.031, generator_feat_match_loss=3.377, over 79.00 samples.], tot_loss[discriminator_loss=2.62, discriminator_real_loss=1.329, discriminator_fake_loss=1.29, generator_loss=30.45, generator_mel_loss=20.95, generator_kl_loss=1.947, generator_dur_loss=1.673, generator_adv_loss=2.209, generator_feat_match_loss=3.68, over 2609.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:28:27,859 INFO [train.py:811] (3/4) Start epoch 397
2023-11-13 20:32:03,281 INFO [train.py:811] (3/4) Start epoch 398
2023-11-13 20:33:20,264 INFO [train.py:467] (3/4) Epoch 398, batch 11, global_batch_idx: 14700, batch size: 126, loss[discriminator_loss=2.715, discriminator_real_loss=1.312, discriminator_fake_loss=1.404, generator_loss=30.91, generator_mel_loss=21.77, generator_kl_loss=2.013, generator_dur_loss=1.63, generator_adv_loss=1.974, generator_feat_match_loss=3.521, over 126.00 samples.], tot_loss[discriminator_loss=2.631, discriminator_real_loss=1.318, discriminator_fake_loss=1.313, generator_loss=30.45, generator_mel_loss=21.19, generator_kl_loss=1.965, generator_dur_loss=1.664, generator_adv_loss=2.113, generator_feat_match_loss=3.519, over 922.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:35:36,050 INFO [train.py:811] (3/4) Start epoch 399
2023-11-13 20:38:01,190 INFO [train.py:467] (3/4) Epoch 399, batch 24, global_batch_idx: 14750, batch size: 52, loss[discriminator_loss=2.648, discriminator_real_loss=1.271, discriminator_fake_loss=1.378, generator_loss=30.43, generator_mel_loss=21.03, generator_kl_loss=1.943, generator_dur_loss=1.667, generator_adv_loss=2.146, generator_feat_match_loss=3.65, over 52.00 samples.], tot_loss[discriminator_loss=2.631, discriminator_real_loss=1.332, discriminator_fake_loss=1.3, generator_loss=30.35, generator_mel_loss=21.12, generator_kl_loss=1.978, generator_dur_loss=1.671, generator_adv_loss=2.116, generator_feat_match_loss=3.462, over 1754.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:39:10,566 INFO [train.py:811] (3/4) Start epoch 400
2023-11-13 20:42:42,878 INFO [train.py:811] (3/4) Start epoch 401
2023-11-13 20:42:59,069 INFO [train.py:467] (3/4) Epoch 401, batch 0, global_batch_idx: 14800, batch size: 52, loss[discriminator_loss=2.635, discriminator_real_loss=1.414, discriminator_fake_loss=1.221, generator_loss=30.12, generator_mel_loss=20.89, generator_kl_loss=2.036, generator_dur_loss=1.663, generator_adv_loss=1.912, generator_feat_match_loss=3.615, over 52.00 samples.], tot_loss[discriminator_loss=2.635, discriminator_real_loss=1.414, discriminator_fake_loss=1.221, generator_loss=30.12, generator_mel_loss=20.89, generator_kl_loss=2.036, generator_dur_loss=1.663, generator_adv_loss=1.912, generator_feat_match_loss=3.615, over 52.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:42:59,070 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 20:43:10,796 INFO [train.py:517] (3/4) Epoch 401, validation: discriminator_loss=2.685, discriminator_real_loss=1.134, discriminator_fake_loss=1.55, generator_loss=30.75, generator_mel_loss=21.67, generator_kl_loss=2.09, generator_dur_loss=1.65, generator_adv_loss=1.705, generator_feat_match_loss=3.639, over 100.00 samples.
2023-11-13 20:43:10,798 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 20:46:21,657 INFO [train.py:811] (3/4) Start epoch 402
2023-11-13 20:47:50,200 INFO [train.py:467] (3/4) Epoch 402, batch 13, global_batch_idx: 14850, batch size: 69, loss[discriminator_loss=2.434, discriminator_real_loss=1.139, discriminator_fake_loss=1.295, generator_loss=31.38, generator_mel_loss=21.07, generator_kl_loss=1.922, generator_dur_loss=1.679, generator_adv_loss=2.385, generator_feat_match_loss=4.324, over 69.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.355, discriminator_fake_loss=1.261, generator_loss=30.71, generator_mel_loss=20.97, generator_kl_loss=1.945, generator_dur_loss=1.677, generator_adv_loss=2.252, generator_feat_match_loss=3.873, over 980.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:49:53,136 INFO [train.py:811] (3/4) Start epoch 403
2023-11-13 20:52:22,019 INFO [train.py:467] (3/4) Epoch 403, batch 26, global_batch_idx: 14900, batch size: 79, loss[discriminator_loss=2.648, discriminator_real_loss=1.447, discriminator_fake_loss=1.201, generator_loss=29.83, generator_mel_loss=20.93, generator_kl_loss=1.918, generator_dur_loss=1.68, generator_adv_loss=1.998, generator_feat_match_loss=3.301, over 79.00 samples.], tot_loss[discriminator_loss=2.645, discriminator_real_loss=1.335, discriminator_fake_loss=1.31, generator_loss=30.37, generator_mel_loss=21.09, generator_kl_loss=1.962, generator_dur_loss=1.671, generator_adv_loss=2.129, generator_feat_match_loss=3.52, over 2097.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 20:53:20,766 INFO [train.py:811] (3/4) Start epoch 404
2023-11-13 20:56:47,760 INFO [train.py:811] (3/4) Start epoch 405
2023-11-13 20:57:17,956 INFO [train.py:467] (3/4) Epoch 405, batch 2, global_batch_idx: 14950, batch size: 95, loss[discriminator_loss=2.51, discriminator_real_loss=1.276, discriminator_fake_loss=1.233, generator_loss=31.03, generator_mel_loss=21.2, generator_kl_loss=1.942, generator_dur_loss=1.641, generator_adv_loss=2.424, generator_feat_match_loss=3.82, over 95.00 samples.], tot_loss[discriminator_loss=2.502, discriminator_real_loss=1.277, discriminator_fake_loss=1.226, generator_loss=31.18, generator_mel_loss=21.3, generator_kl_loss=1.931, generator_dur_loss=1.67, generator_adv_loss=2.355, generator_feat_match_loss=3.919, over 201.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:00:20,929 INFO [train.py:811] (3/4) Start epoch 406
2023-11-13 21:01:56,666 INFO [train.py:467] (3/4) Epoch 406, batch 15, global_batch_idx: 15000, batch size: 55, loss[discriminator_loss=2.662, discriminator_real_loss=1.135, discriminator_fake_loss=1.527, generator_loss=30.8, generator_mel_loss=21.35, generator_kl_loss=1.973, generator_dur_loss=1.687, generator_adv_loss=2.189, generator_feat_match_loss=3.605, over 55.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.335, discriminator_fake_loss=1.283, generator_loss=30.66, generator_mel_loss=20.96, generator_kl_loss=1.975, generator_dur_loss=1.677, generator_adv_loss=2.275, generator_feat_match_loss=3.774, over 942.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:01:56,668 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 21:02:07,737 INFO [train.py:517] (3/4) Epoch 406, validation: discriminator_loss=2.92, discriminator_real_loss=1.335, discriminator_fake_loss=1.585, generator_loss=30.66, generator_mel_loss=21.84, generator_kl_loss=2.085, generator_dur_loss=1.638, generator_adv_loss=1.74, generator_feat_match_loss=3.348, over 100.00 samples.
2023-11-13 21:02:07,738 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 21:04:08,534 INFO [train.py:811] (3/4) Start epoch 407
2023-11-13 21:06:57,053 INFO [train.py:467] (3/4) Epoch 407, batch 28, global_batch_idx: 15050, batch size: 81, loss[discriminator_loss=2.746, discriminator_real_loss=1.462, discriminator_fake_loss=1.284, generator_loss=29.67, generator_mel_loss=20.83, generator_kl_loss=1.919, generator_dur_loss=1.655, generator_adv_loss=1.889, generator_feat_match_loss=3.381, over 81.00 samples.], tot_loss[discriminator_loss=2.621, discriminator_real_loss=1.331, discriminator_fake_loss=1.29, generator_loss=30.25, generator_mel_loss=20.87, generator_kl_loss=1.956, generator_dur_loss=1.667, generator_adv_loss=2.145, generator_feat_match_loss=3.616, over 2229.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:07:41,665 INFO [train.py:811] (3/4) Start epoch 408
2023-11-13 21:11:10,048 INFO [train.py:811] (3/4) Start epoch 409
2023-11-13 21:11:44,647 INFO [train.py:467] (3/4) Epoch 409, batch 4, global_batch_idx: 15100, batch size: 58, loss[discriminator_loss=2.703, discriminator_real_loss=1.33, discriminator_fake_loss=1.373, generator_loss=29.95, generator_mel_loss=20.86, generator_kl_loss=1.995, generator_dur_loss=1.68, generator_adv_loss=2.104, generator_feat_match_loss=3.316, over 58.00 samples.], tot_loss[discriminator_loss=2.673, discriminator_real_loss=1.364, discriminator_fake_loss=1.309, generator_loss=30.2, generator_mel_loss=21.09, generator_kl_loss=1.946, generator_dur_loss=1.664, generator_adv_loss=2.072, generator_feat_match_loss=3.431, over 341.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:14:40,475 INFO [train.py:811] (3/4) Start epoch 410
2023-11-13 21:16:30,389 INFO [train.py:467] (3/4) Epoch 410, batch 17, global_batch_idx: 15150, batch size: 53, loss[discriminator_loss=2.6, discriminator_real_loss=1.431, discriminator_fake_loss=1.169, generator_loss=29.68, generator_mel_loss=20.84, generator_kl_loss=1.934, generator_dur_loss=1.669, generator_adv_loss=1.964, generator_feat_match_loss=3.275, over 53.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.321, discriminator_fake_loss=1.303, generator_loss=30.35, generator_mel_loss=21.05, generator_kl_loss=1.96, generator_dur_loss=1.674, generator_adv_loss=2.157, generator_feat_match_loss=3.511, over 1280.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:18:08,741 INFO [train.py:811] (3/4) Start epoch 411
2023-11-13 21:21:10,653 INFO [train.py:467] (3/4) Epoch 411, batch 30, global_batch_idx: 15200, batch size: 49, loss[discriminator_loss=2.77, discriminator_real_loss=1.511, discriminator_fake_loss=1.259, generator_loss=30.23, generator_mel_loss=21.11, generator_kl_loss=1.91, generator_dur_loss=1.708, generator_adv_loss=2.121, generator_feat_match_loss=3.385, over 49.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.323, discriminator_fake_loss=1.295, generator_loss=30.59, generator_mel_loss=20.96, generator_kl_loss=1.959, generator_dur_loss=1.675, generator_adv_loss=2.262, generator_feat_match_loss=3.729, over 2381.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 32.0
2023-11-13 21:21:10,655 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 21:21:21,735 INFO [train.py:517] (3/4) Epoch 411, validation: discriminator_loss=2.778, discriminator_real_loss=1.515, discriminator_fake_loss=1.263, generator_loss=30.97, generator_mel_loss=21.71, generator_kl_loss=2.018, generator_dur_loss=1.643, generator_adv_loss=2.269, generator_feat_match_loss=3.328, over 100.00 samples.
2023-11-13 21:21:21,737 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 21:21:51,759 INFO [train.py:811] (3/4) Start epoch 412
2023-11-13 21:25:24,719 INFO [train.py:811] (3/4) Start epoch 413
2023-11-13 21:26:11,811 INFO [train.py:467] (3/4) Epoch 413, batch 6, global_batch_idx: 15250, batch size: 76, loss[discriminator_loss=2.676, discriminator_real_loss=1.337, discriminator_fake_loss=1.34, generator_loss=29.74, generator_mel_loss=20.66, generator_kl_loss=1.995, generator_dur_loss=1.666, generator_adv_loss=2.031, generator_feat_match_loss=3.391, over 76.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.315, discriminator_fake_loss=1.285, generator_loss=29.93, generator_mel_loss=20.75, generator_kl_loss=2.017, generator_dur_loss=1.669, generator_adv_loss=2.051, generator_feat_match_loss=3.445, over 558.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:28:57,434 INFO [train.py:811] (3/4) Start epoch 414
2023-11-13 21:30:49,754 INFO [train.py:467] (3/4) Epoch 414, batch 19, global_batch_idx: 15300, batch size: 52, loss[discriminator_loss=2.496, discriminator_real_loss=1.193, discriminator_fake_loss=1.303, generator_loss=31.08, generator_mel_loss=20.9, generator_kl_loss=1.93, generator_dur_loss=1.663, generator_adv_loss=2.367, generator_feat_match_loss=4.223, over 52.00 samples.], tot_loss[discriminator_loss=2.583, discriminator_real_loss=1.315, discriminator_fake_loss=1.268, generator_loss=30.8, generator_mel_loss=20.91, generator_kl_loss=1.996, generator_dur_loss=1.666, generator_adv_loss=2.265, generator_feat_match_loss=3.956, over 1530.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:32:26,927 INFO [train.py:811] (3/4) Start epoch 415
2023-11-13 21:35:34,924 INFO [train.py:467] (3/4) Epoch 415, batch 32, global_batch_idx: 15350, batch size: 81, loss[discriminator_loss=2.678, discriminator_real_loss=1.418, discriminator_fake_loss=1.26, generator_loss=30.02, generator_mel_loss=21.03, generator_kl_loss=1.998, generator_dur_loss=1.653, generator_adv_loss=1.851, generator_feat_match_loss=3.492, over 81.00 samples.], tot_loss[discriminator_loss=2.64, discriminator_real_loss=1.349, discriminator_fake_loss=1.291, generator_loss=30.21, generator_mel_loss=21.02, generator_kl_loss=1.983, generator_dur_loss=1.665, generator_adv_loss=2.083, generator_feat_match_loss=3.455, over 2347.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:36:01,107 INFO [train.py:811] (3/4) Start epoch 416
2023-11-13 21:39:30,011 INFO [train.py:811] (3/4) Start epoch 417
2023-11-13 21:40:26,689 INFO [train.py:467] (3/4) Epoch 417, batch 8, global_batch_idx: 15400, batch size: 59, loss[discriminator_loss=2.627, discriminator_real_loss=1.303, discriminator_fake_loss=1.324, generator_loss=29.55, generator_mel_loss=20.4, generator_kl_loss=1.871, generator_dur_loss=1.688, generator_adv_loss=2.143, generator_feat_match_loss=3.447, over 59.00 samples.], tot_loss[discriminator_loss=2.63, discriminator_real_loss=1.371, discriminator_fake_loss=1.258, generator_loss=30.13, generator_mel_loss=20.84, generator_kl_loss=1.943, generator_dur_loss=1.678, generator_adv_loss=2.142, generator_feat_match_loss=3.523, over 598.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:40:26,691 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 21:40:38,425 INFO [train.py:517] (3/4) Epoch 417, validation: discriminator_loss=2.501, discriminator_real_loss=1.233, discriminator_fake_loss=1.268, generator_loss=31.27, generator_mel_loss=21.79, generator_kl_loss=1.954, generator_dur_loss=1.645, generator_adv_loss=2.051, generator_feat_match_loss=3.831, over 100.00 samples.
2023-11-13 21:40:38,426 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 21:43:11,565 INFO [train.py:811] (3/4) Start epoch 418
2023-11-13 21:45:28,051 INFO [train.py:467] (3/4) Epoch 418, batch 21, global_batch_idx: 15450, batch size: 76, loss[discriminator_loss=2.535, discriminator_real_loss=1.224, discriminator_fake_loss=1.311, generator_loss=30.96, generator_mel_loss=21.1, generator_kl_loss=1.933, generator_dur_loss=1.683, generator_adv_loss=2.383, generator_feat_match_loss=3.859, over 76.00 samples.], tot_loss[discriminator_loss=2.641, discriminator_real_loss=1.335, discriminator_fake_loss=1.306, generator_loss=30.32, generator_mel_loss=20.94, generator_kl_loss=1.951, generator_dur_loss=1.675, generator_adv_loss=2.173, generator_feat_match_loss=3.59, over 1392.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:46:43,802 INFO [train.py:811] (3/4) Start epoch 419
2023-11-13 21:50:02,926 INFO [train.py:467] (3/4) Epoch 419, batch 34, global_batch_idx: 15500, batch size: 69, loss[discriminator_loss=2.627, discriminator_real_loss=1.303, discriminator_fake_loss=1.324, generator_loss=30, generator_mel_loss=20.92, generator_kl_loss=1.976, generator_dur_loss=1.683, generator_adv_loss=1.936, generator_feat_match_loss=3.488, over 69.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.356, discriminator_fake_loss=1.32, generator_loss=30.43, generator_mel_loss=21.17, generator_kl_loss=1.987, generator_dur_loss=1.67, generator_adv_loss=2.099, generator_feat_match_loss=3.505, over 2724.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:50:16,756 INFO [train.py:811] (3/4) Start epoch 420
2023-11-13 21:53:53,583 INFO [train.py:811] (3/4) Start epoch 421
2023-11-13 21:55:14,928 INFO [train.py:467] (3/4) Epoch 421, batch 10, global_batch_idx: 15550, batch size: 153, loss[discriminator_loss=2.578, discriminator_real_loss=1.238, discriminator_fake_loss=1.339, generator_loss=31.18, generator_mel_loss=21.31, generator_kl_loss=1.991, generator_dur_loss=1.622, generator_adv_loss=2.41, generator_feat_match_loss=3.854, over 153.00 samples.], tot_loss[discriminator_loss=2.589, discriminator_real_loss=1.309, discriminator_fake_loss=1.279, generator_loss=30.69, generator_mel_loss=21.06, generator_kl_loss=1.967, generator_dur_loss=1.662, generator_adv_loss=2.229, generator_feat_match_loss=3.773, over 804.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 21:57:33,403 INFO [train.py:811] (3/4) Start epoch 422
2023-11-13 21:59:51,658 INFO [train.py:467] (3/4) Epoch 422, batch 23, global_batch_idx: 15600, batch size: 153, loss[discriminator_loss=2.602, discriminator_real_loss=1.119, discriminator_fake_loss=1.482, generator_loss=30.64, generator_mel_loss=20.8, generator_kl_loss=1.981, generator_dur_loss=1.624, generator_adv_loss=2.285, generator_feat_match_loss=3.945, over 153.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.283, discriminator_fake_loss=1.295, generator_loss=30.55, generator_mel_loss=20.72, generator_kl_loss=1.961, generator_dur_loss=1.666, generator_adv_loss=2.248, generator_feat_match_loss=3.95, over 1794.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 32.0
2023-11-13 21:59:51,660 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 22:00:02,682 INFO [train.py:517] (3/4) Epoch 422, validation: discriminator_loss=2.598, discriminator_real_loss=1.279, discriminator_fake_loss=1.319, generator_loss=30.89, generator_mel_loss=21.57, generator_kl_loss=2.034, generator_dur_loss=1.642, generator_adv_loss=1.964, generator_feat_match_loss=3.671, over 100.00 samples.
2023-11-13 22:00:02,684 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 22:01:12,388 INFO [train.py:811] (3/4) Start epoch 423
2023-11-13 22:04:45,451 INFO [train.py:467] (3/4) Epoch 423, batch 36, global_batch_idx: 15650, batch size: 53, loss[discriminator_loss=2.508, discriminator_real_loss=1.314, discriminator_fake_loss=1.192, generator_loss=30.69, generator_mel_loss=20.76, generator_kl_loss=1.909, generator_dur_loss=1.673, generator_adv_loss=2.311, generator_feat_match_loss=4.039, over 53.00 samples.], tot_loss[discriminator_loss=2.614, discriminator_real_loss=1.319, discriminator_fake_loss=1.295, generator_loss=30.38, generator_mel_loss=20.95, generator_kl_loss=1.958, generator_dur_loss=1.672, generator_adv_loss=2.153, generator_feat_match_loss=3.642, over 2596.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:04:46,142 INFO [train.py:811] (3/4) Start epoch 424
2023-11-13 22:08:05,990 INFO [train.py:811] (3/4) Start epoch 425
2023-11-13 22:09:22,647 INFO [train.py:467] (3/4) Epoch 425, batch 12, global_batch_idx: 15700, batch size: 55, loss[discriminator_loss=2.621, discriminator_real_loss=1.25, discriminator_fake_loss=1.371, generator_loss=30.39, generator_mel_loss=21.08, generator_kl_loss=1.94, generator_dur_loss=1.706, generator_adv_loss=2.301, generator_feat_match_loss=3.361, over 55.00 samples.], tot_loss[discriminator_loss=2.659, discriminator_real_loss=1.37, discriminator_fake_loss=1.289, generator_loss=30.62, generator_mel_loss=21.11, generator_kl_loss=1.969, generator_dur_loss=1.67, generator_adv_loss=2.214, generator_feat_match_loss=3.661, over 1005.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:11:36,637 INFO [train.py:811] (3/4) Start epoch 426
2023-11-13 22:13:59,546 INFO [train.py:467] (3/4) Epoch 426, batch 25, global_batch_idx: 15750, batch size: 81, loss[discriminator_loss=2.568, discriminator_real_loss=1.352, discriminator_fake_loss=1.217, generator_loss=31.62, generator_mel_loss=21.08, generator_kl_loss=1.859, generator_dur_loss=1.669, generator_adv_loss=2.473, generator_feat_match_loss=4.535, over 81.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.332, discriminator_fake_loss=1.292, generator_loss=30.66, generator_mel_loss=21, generator_kl_loss=1.945, generator_dur_loss=1.67, generator_adv_loss=2.285, generator_feat_match_loss=3.767, over 1756.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:15:08,070 INFO [train.py:811] (3/4) Start epoch 427
2023-11-13 22:18:41,277 INFO [train.py:811] (3/4) Start epoch 428
2023-11-13 22:19:03,535 INFO [train.py:467] (3/4) Epoch 428, batch 1, global_batch_idx: 15800, batch size: 90, loss[discriminator_loss=2.607, discriminator_real_loss=1.195, discriminator_fake_loss=1.412, generator_loss=30.07, generator_mel_loss=20.91, generator_kl_loss=2.037, generator_dur_loss=1.656, generator_adv_loss=2.01, generator_feat_match_loss=3.459, over 90.00 samples.], tot_loss[discriminator_loss=2.641, discriminator_real_loss=1.205, discriminator_fake_loss=1.436, generator_loss=29.77, generator_mel_loss=20.82, generator_kl_loss=1.978, generator_dur_loss=1.662, generator_adv_loss=1.907, generator_feat_match_loss=3.411, over 154.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:19:03,537 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 22:19:15,683 INFO [train.py:517] (3/4) Epoch 428, validation: discriminator_loss=2.552, discriminator_real_loss=1.217, discriminator_fake_loss=1.334, generator_loss=30.8, generator_mel_loss=21.6, generator_kl_loss=2.089, generator_dur_loss=1.651, generator_adv_loss=1.891, generator_feat_match_loss=3.575, over 100.00 samples.
2023-11-13 22:19:15,684 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 22:22:25,603 INFO [train.py:811] (3/4) Start epoch 429
2023-11-13 22:23:54,885 INFO [train.py:467] (3/4) Epoch 429, batch 14, global_batch_idx: 15850, batch size: 60, loss[discriminator_loss=2.482, discriminator_real_loss=1.289, discriminator_fake_loss=1.193, generator_loss=30.75, generator_mel_loss=20.71, generator_kl_loss=1.962, generator_dur_loss=1.656, generator_adv_loss=2.146, generator_feat_match_loss=4.277, over 60.00 samples.], tot_loss[discriminator_loss=2.649, discriminator_real_loss=1.367, discriminator_fake_loss=1.281, generator_loss=30.63, generator_mel_loss=20.95, generator_kl_loss=1.96, generator_dur_loss=1.675, generator_adv_loss=2.251, generator_feat_match_loss=3.79, over 987.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:25:58,856 INFO [train.py:811] (3/4) Start epoch 430
2023-11-13 22:28:46,594 INFO [train.py:467] (3/4) Epoch 430, batch 27, global_batch_idx: 15900, batch size: 85, loss[discriminator_loss=2.727, discriminator_real_loss=1.517, discriminator_fake_loss=1.209, generator_loss=29.71, generator_mel_loss=20.89, generator_kl_loss=1.924, generator_dur_loss=1.681, generator_adv_loss=1.968, generator_feat_match_loss=3.246, over 85.00 samples.], tot_loss[discriminator_loss=2.637, discriminator_real_loss=1.332, discriminator_fake_loss=1.305, generator_loss=30.37, generator_mel_loss=20.93, generator_kl_loss=2.006, generator_dur_loss=1.662, generator_adv_loss=2.133, generator_feat_match_loss=3.645, over 2121.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:29:34,225 INFO [train.py:811] (3/4) Start epoch 431
2023-11-13 22:33:09,222 INFO [train.py:811] (3/4) Start epoch 432
2023-11-13 22:33:45,675 INFO [train.py:467] (3/4) Epoch 432, batch 3, global_batch_idx: 15950, batch size: 85, loss[discriminator_loss=2.629, discriminator_real_loss=1.387, discriminator_fake_loss=1.242, generator_loss=30.61, generator_mel_loss=21.16, generator_kl_loss=1.947, generator_dur_loss=1.658, generator_adv_loss=2.146, generator_feat_match_loss=3.699, over 85.00 samples.], tot_loss[discriminator_loss=2.702, discriminator_real_loss=1.39, discriminator_fake_loss=1.311, generator_loss=30.28, generator_mel_loss=21.09, generator_kl_loss=1.973, generator_dur_loss=1.658, generator_adv_loss=2.013, generator_feat_match_loss=3.549, over 377.00 samples.], cur_lr_g: 1.90e-04, cur_lr_d: 1.90e-04, grad_scale: 16.0
2023-11-13 22:36:47,675 INFO [train.py:811] (3/4) Start epoch 433
2023-11-13 22:38:23,786 INFO [train.py:467] (3/4) Epoch 433, batch 16, global_batch_idx: 16000, batch size: 55, loss[discriminator_loss=2.613, discriminator_real_loss=1.339, discriminator_fake_loss=1.274, generator_loss=30.34, generator_mel_loss=20.78, generator_kl_loss=1.899, generator_dur_loss=1.683, generator_adv_loss=2.365, generator_feat_match_loss=3.611, over 55.00 samples.], tot_loss[discriminator_loss=2.643, discriminator_real_loss=1.328, discriminator_fake_loss=1.315, generator_loss=30.49, generator_mel_loss=20.99, generator_kl_loss=1.965, generator_dur_loss=1.67, generator_adv_loss=2.189, generator_feat_match_loss=3.679, over 1165.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 32.0
2023-11-13 22:38:23,787 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 22:38:34,695 INFO [train.py:517] (3/4) Epoch 433, validation: discriminator_loss=2.623, discriminator_real_loss=1.341, discriminator_fake_loss=1.282, generator_loss=31.39, generator_mel_loss=21.95, generator_kl_loss=2.056, generator_dur_loss=1.648, generator_adv_loss=2.076, generator_feat_match_loss=3.654, over 100.00 samples.
2023-11-13 22:38:34,696 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 22:40:30,121 INFO [train.py:811] (3/4) Start epoch 434
2023-11-13 22:43:22,568 INFO [train.py:467] (3/4) Epoch 434, batch 29, global_batch_idx: 16050, batch size: 76, loss[discriminator_loss=2.613, discriminator_real_loss=1.436, discriminator_fake_loss=1.177, generator_loss=30.49, generator_mel_loss=21.21, generator_kl_loss=1.866, generator_dur_loss=1.663, generator_adv_loss=2.186, generator_feat_match_loss=3.57, over 76.00 samples.], tot_loss[discriminator_loss=2.663, discriminator_real_loss=1.364, discriminator_fake_loss=1.299, generator_loss=30.27, generator_mel_loss=21.04, generator_kl_loss=1.971, generator_dur_loss=1.67, generator_adv_loss=2.092, generator_feat_match_loss=3.494, over 2075.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 22:44:01,557 INFO [train.py:811] (3/4) Start epoch 435
2023-11-13 22:47:31,968 INFO [train.py:811] (3/4) Start epoch 436
2023-11-13 22:48:13,600 INFO [train.py:467] (3/4) Epoch 436, batch 5, global_batch_idx: 16100, batch size: 65, loss[discriminator_loss=2.723, discriminator_real_loss=1.617, discriminator_fake_loss=1.105, generator_loss=29.6, generator_mel_loss=20.35, generator_kl_loss=1.988, generator_dur_loss=1.698, generator_adv_loss=2.203, generator_feat_match_loss=3.359, over 65.00 samples.], tot_loss[discriminator_loss=2.595, discriminator_real_loss=1.353, discriminator_fake_loss=1.243, generator_loss=30.52, generator_mel_loss=20.69, generator_kl_loss=1.951, generator_dur_loss=1.674, generator_adv_loss=2.283, generator_feat_match_loss=3.925, over 426.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 22:51:05,664 INFO [train.py:811] (3/4) Start epoch 437
2023-11-13 22:52:55,803 INFO [train.py:467] (3/4) Epoch 437, batch 18, global_batch_idx: 16150, batch size: 53, loss[discriminator_loss=2.76, discriminator_real_loss=1.471, discriminator_fake_loss=1.289, generator_loss=29.76, generator_mel_loss=20.6, generator_kl_loss=2.036, generator_dur_loss=1.691, generator_adv_loss=2.16, generator_feat_match_loss=3.273, over 53.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.298, discriminator_fake_loss=1.268, generator_loss=30.6, generator_mel_loss=20.72, generator_kl_loss=1.951, generator_dur_loss=1.665, generator_adv_loss=2.288, generator_feat_match_loss=3.978, over 1376.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 22:54:32,577 INFO [train.py:811] (3/4) Start epoch 438
2023-11-13 22:57:42,664 INFO [train.py:467] (3/4) Epoch 438, batch 31, global_batch_idx: 16200, batch size: 64, loss[discriminator_loss=2.52, discriminator_real_loss=1.267, discriminator_fake_loss=1.252, generator_loss=30.61, generator_mel_loss=21.23, generator_kl_loss=1.958, generator_dur_loss=1.679, generator_adv_loss=2.1, generator_feat_match_loss=3.643, over 64.00 samples.], tot_loss[discriminator_loss=2.597, discriminator_real_loss=1.316, discriminator_fake_loss=1.281, generator_loss=30.14, generator_mel_loss=20.8, generator_kl_loss=1.979, generator_dur_loss=1.673, generator_adv_loss=2.102, generator_feat_match_loss=3.584, over 2135.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 22:57:42,666 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 22:57:53,383 INFO [train.py:517] (3/4) Epoch 438, validation: discriminator_loss=2.524, discriminator_real_loss=1.23, discriminator_fake_loss=1.294, generator_loss=31.6, generator_mel_loss=21.79, generator_kl_loss=2.175, generator_dur_loss=1.647, generator_adv_loss=2.047, generator_feat_match_loss=3.945, over 100.00 samples.
2023-11-13 22:57:53,385 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 22:58:17,362 INFO [train.py:811] (3/4) Start epoch 439
2023-11-13 23:01:50,611 INFO [train.py:811] (3/4) Start epoch 440
2023-11-13 23:02:41,576 INFO [train.py:467] (3/4) Epoch 440, batch 7, global_batch_idx: 16250, batch size: 90, loss[discriminator_loss=2.617, discriminator_real_loss=1.363, discriminator_fake_loss=1.254, generator_loss=30.08, generator_mel_loss=20.99, generator_kl_loss=1.921, generator_dur_loss=1.659, generator_adv_loss=2.047, generator_feat_match_loss=3.457, over 90.00 samples.], tot_loss[discriminator_loss=2.57, discriminator_real_loss=1.253, discriminator_fake_loss=1.317, generator_loss=30.04, generator_mel_loss=20.76, generator_kl_loss=1.943, generator_dur_loss=1.665, generator_adv_loss=2.106, generator_feat_match_loss=3.564, over 606.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:05:22,486 INFO [train.py:811] (3/4) Start epoch 441
2023-11-13 23:07:31,672 INFO [train.py:467] (3/4) Epoch 441, batch 20, global_batch_idx: 16300, batch size: 90, loss[discriminator_loss=2.734, discriminator_real_loss=1.192, discriminator_fake_loss=1.541, generator_loss=30.03, generator_mel_loss=21.08, generator_kl_loss=1.941, generator_dur_loss=1.669, generator_adv_loss=1.993, generator_feat_match_loss=3.344, over 90.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.314, discriminator_fake_loss=1.299, generator_loss=30.62, generator_mel_loss=20.93, generator_kl_loss=1.993, generator_dur_loss=1.667, generator_adv_loss=2.218, generator_feat_match_loss=3.812, over 1527.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:08:52,385 INFO [train.py:811] (3/4) Start epoch 442
2023-11-13 23:12:10,652 INFO [train.py:467] (3/4) Epoch 442, batch 33, global_batch_idx: 16350, batch size: 56, loss[discriminator_loss=2.465, discriminator_real_loss=1.316, discriminator_fake_loss=1.147, generator_loss=30.59, generator_mel_loss=20.57, generator_kl_loss=2.04, generator_dur_loss=1.68, generator_adv_loss=2.279, generator_feat_match_loss=4.02, over 56.00 samples.], tot_loss[discriminator_loss=2.522, discriminator_real_loss=1.269, discriminator_fake_loss=1.254, generator_loss=30.86, generator_mel_loss=20.91, generator_kl_loss=1.967, generator_dur_loss=1.664, generator_adv_loss=2.291, generator_feat_match_loss=4.025, over 2780.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:12:32,545 INFO [train.py:811] (3/4) Start epoch 443
2023-11-13 23:16:04,899 INFO [train.py:811] (3/4) Start epoch 444
2023-11-13 23:17:12,826 INFO [train.py:467] (3/4) Epoch 444, batch 9, global_batch_idx: 16400, batch size: 71, loss[discriminator_loss=2.682, discriminator_real_loss=1.274, discriminator_fake_loss=1.407, generator_loss=29.99, generator_mel_loss=20.51, generator_kl_loss=1.941, generator_dur_loss=1.657, generator_adv_loss=2.174, generator_feat_match_loss=3.705, over 71.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.293, discriminator_fake_loss=1.292, generator_loss=30.17, generator_mel_loss=20.5, generator_kl_loss=1.971, generator_dur_loss=1.673, generator_adv_loss=2.235, generator_feat_match_loss=3.791, over 681.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 32.0
2023-11-13 23:17:12,828 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 23:17:23,924 INFO [train.py:517] (3/4) Epoch 444, validation: discriminator_loss=2.579, discriminator_real_loss=1.265, discriminator_fake_loss=1.314, generator_loss=30.75, generator_mel_loss=21.29, generator_kl_loss=2.088, generator_dur_loss=1.643, generator_adv_loss=2.073, generator_feat_match_loss=3.658, over 100.00 samples.
2023-11-13 23:17:23,925 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 23:19:49,915 INFO [train.py:811] (3/4) Start epoch 445
2023-11-13 23:22:03,118 INFO [train.py:467] (3/4) Epoch 445, batch 22, global_batch_idx: 16450, batch size: 64, loss[discriminator_loss=2.598, discriminator_real_loss=1.346, discriminator_fake_loss=1.252, generator_loss=31.14, generator_mel_loss=21.01, generator_kl_loss=2.067, generator_dur_loss=1.678, generator_adv_loss=2.211, generator_feat_match_loss=4.172, over 64.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.319, discriminator_fake_loss=1.272, generator_loss=30.3, generator_mel_loss=20.74, generator_kl_loss=1.996, generator_dur_loss=1.674, generator_adv_loss=2.186, generator_feat_match_loss=3.699, over 1732.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:23:20,639 INFO [train.py:811] (3/4) Start epoch 446
2023-11-13 23:26:47,953 INFO [train.py:467] (3/4) Epoch 446, batch 35, global_batch_idx: 16500, batch size: 95, loss[discriminator_loss=2.576, discriminator_real_loss=1.276, discriminator_fake_loss=1.3, generator_loss=30.26, generator_mel_loss=20.81, generator_kl_loss=2.012, generator_dur_loss=1.654, generator_adv_loss=2.229, generator_feat_match_loss=3.547, over 95.00 samples.], tot_loss[discriminator_loss=2.593, discriminator_real_loss=1.307, discriminator_fake_loss=1.286, generator_loss=30.28, generator_mel_loss=20.7, generator_kl_loss=1.994, generator_dur_loss=1.663, generator_adv_loss=2.171, generator_feat_match_loss=3.756, over 2706.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:26:53,747 INFO [train.py:811] (3/4) Start epoch 447
2023-11-13 23:30:19,559 INFO [train.py:811] (3/4) Start epoch 448
2023-11-13 23:31:37,819 INFO [train.py:467] (3/4) Epoch 448, batch 11, global_batch_idx: 16550, batch size: 90, loss[discriminator_loss=2.658, discriminator_real_loss=1.297, discriminator_fake_loss=1.361, generator_loss=29.78, generator_mel_loss=20.71, generator_kl_loss=1.843, generator_dur_loss=1.636, generator_adv_loss=2.133, generator_feat_match_loss=3.465, over 90.00 samples.], tot_loss[discriminator_loss=2.655, discriminator_real_loss=1.332, discriminator_fake_loss=1.323, generator_loss=30.31, generator_mel_loss=21.02, generator_kl_loss=1.98, generator_dur_loss=1.665, generator_adv_loss=2.069, generator_feat_match_loss=3.583, over 1011.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:33:54,487 INFO [train.py:811] (3/4) Start epoch 449
2023-11-13 23:36:23,410 INFO [train.py:467] (3/4) Epoch 449, batch 24, global_batch_idx: 16600, batch size: 56, loss[discriminator_loss=2.645, discriminator_real_loss=1.312, discriminator_fake_loss=1.331, generator_loss=30.43, generator_mel_loss=20.99, generator_kl_loss=1.968, generator_dur_loss=1.653, generator_adv_loss=2.25, generator_feat_match_loss=3.562, over 56.00 samples.], tot_loss[discriminator_loss=2.632, discriminator_real_loss=1.328, discriminator_fake_loss=1.304, generator_loss=30.15, generator_mel_loss=20.83, generator_kl_loss=1.951, generator_dur_loss=1.668, generator_adv_loss=2.114, generator_feat_match_loss=3.591, over 1647.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:36:23,412 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 23:36:34,679 INFO [train.py:517] (3/4) Epoch 449, validation: discriminator_loss=2.759, discriminator_real_loss=1.395, discriminator_fake_loss=1.364, generator_loss=31.01, generator_mel_loss=21.7, generator_kl_loss=2.131, generator_dur_loss=1.646, generator_adv_loss=1.918, generator_feat_match_loss=3.617, over 100.00 samples.
2023-11-13 23:36:34,680 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 23:37:40,659 INFO [train.py:811] (3/4) Start epoch 450
2023-11-13 23:41:16,061 INFO [train.py:811] (3/4) Start epoch 451
2023-11-13 23:41:32,075 INFO [train.py:467] (3/4) Epoch 451, batch 0, global_batch_idx: 16650, batch size: 52, loss[discriminator_loss=2.652, discriminator_real_loss=1.186, discriminator_fake_loss=1.467, generator_loss=30.34, generator_mel_loss=21.23, generator_kl_loss=1.934, generator_dur_loss=1.676, generator_adv_loss=1.917, generator_feat_match_loss=3.582, over 52.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.186, discriminator_fake_loss=1.467, generator_loss=30.34, generator_mel_loss=21.23, generator_kl_loss=1.934, generator_dur_loss=1.676, generator_adv_loss=1.917, generator_feat_match_loss=3.582, over 52.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:44:48,675 INFO [train.py:811] (3/4) Start epoch 452
2023-11-13 23:46:12,802 INFO [train.py:467] (3/4) Epoch 452, batch 13, global_batch_idx: 16700, batch size: 73, loss[discriminator_loss=2.652, discriminator_real_loss=1.483, discriminator_fake_loss=1.169, generator_loss=30.86, generator_mel_loss=21.03, generator_kl_loss=1.999, generator_dur_loss=1.642, generator_adv_loss=2.127, generator_feat_match_loss=4.062, over 73.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.328, discriminator_fake_loss=1.29, generator_loss=30.4, generator_mel_loss=20.86, generator_kl_loss=1.981, generator_dur_loss=1.661, generator_adv_loss=2.198, generator_feat_match_loss=3.7, over 1039.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:48:14,969 INFO [train.py:811] (3/4) Start epoch 453
2023-11-13 23:50:45,730 INFO [train.py:467] (3/4) Epoch 453, batch 26, global_batch_idx: 16750, batch size: 81, loss[discriminator_loss=2.559, discriminator_real_loss=1.349, discriminator_fake_loss=1.211, generator_loss=29.88, generator_mel_loss=20.9, generator_kl_loss=2.008, generator_dur_loss=1.672, generator_adv_loss=1.798, generator_feat_match_loss=3.504, over 81.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.308, discriminator_fake_loss=1.277, generator_loss=30.27, generator_mel_loss=20.77, generator_kl_loss=1.994, generator_dur_loss=1.666, generator_adv_loss=2.126, generator_feat_match_loss=3.715, over 1915.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-13 23:51:41,664 INFO [train.py:811] (3/4) Start epoch 454
2023-11-13 23:55:12,437 INFO [train.py:811] (3/4) Start epoch 455
2023-11-13 23:55:36,287 INFO [train.py:467] (3/4) Epoch 455, batch 2, global_batch_idx: 16800, batch size: 85, loss[discriminator_loss=2.631, discriminator_real_loss=1.323, discriminator_fake_loss=1.308, generator_loss=30.84, generator_mel_loss=21.39, generator_kl_loss=1.991, generator_dur_loss=1.665, generator_adv_loss=2.113, generator_feat_match_loss=3.678, over 85.00 samples.], tot_loss[discriminator_loss=2.64, discriminator_real_loss=1.319, discriminator_fake_loss=1.32, generator_loss=30.55, generator_mel_loss=21.2, generator_kl_loss=1.987, generator_dur_loss=1.678, generator_adv_loss=2.119, generator_feat_match_loss=3.569, over 201.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 32.0
2023-11-13 23:55:36,288 INFO [train.py:508] (3/4) Computing validation loss
2023-11-13 23:55:47,918 INFO [train.py:517] (3/4) Epoch 455, validation: discriminator_loss=2.686, discriminator_real_loss=1.338, discriminator_fake_loss=1.347, generator_loss=30.89, generator_mel_loss=21.51, generator_kl_loss=2.185, generator_dur_loss=1.65, generator_adv_loss=1.963, generator_feat_match_loss=3.585, over 100.00 samples.
2023-11-13 23:55:47,919 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-13 23:58:54,442 INFO [train.py:811] (3/4) Start epoch 456
2023-11-14 00:00:36,747 INFO [train.py:467] (3/4) Epoch 456, batch 15, global_batch_idx: 16850, batch size: 71, loss[discriminator_loss=2.531, discriminator_real_loss=1.303, discriminator_fake_loss=1.229, generator_loss=30.13, generator_mel_loss=20.58, generator_kl_loss=1.922, generator_dur_loss=1.656, generator_adv_loss=2.182, generator_feat_match_loss=3.795, over 71.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.328, discriminator_fake_loss=1.328, generator_loss=30.11, generator_mel_loss=20.64, generator_kl_loss=1.971, generator_dur_loss=1.662, generator_adv_loss=2.156, generator_feat_match_loss=3.674, over 1211.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:02:31,851 INFO [train.py:811] (3/4) Start epoch 457
2023-11-14 00:05:07,819 INFO [train.py:467] (3/4) Epoch 457, batch 28, global_batch_idx: 16900, batch size: 76, loss[discriminator_loss=2.727, discriminator_real_loss=1.426, discriminator_fake_loss=1.3, generator_loss=29.99, generator_mel_loss=20.73, generator_kl_loss=1.893, generator_dur_loss=1.662, generator_adv_loss=2.168, generator_feat_match_loss=3.533, over 76.00 samples.], tot_loss[discriminator_loss=2.632, discriminator_real_loss=1.339, discriminator_fake_loss=1.293, generator_loss=30.38, generator_mel_loss=20.89, generator_kl_loss=1.968, generator_dur_loss=1.665, generator_adv_loss=2.173, generator_feat_match_loss=3.688, over 1949.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:05:59,146 INFO [train.py:811] (3/4) Start epoch 458
2023-11-14 00:09:25,600 INFO [train.py:811] (3/4) Start epoch 459
2023-11-14 00:10:04,100 INFO [train.py:467] (3/4) Epoch 459, batch 4, global_batch_idx: 16950, batch size: 65, loss[discriminator_loss=2.623, discriminator_real_loss=1.406, discriminator_fake_loss=1.217, generator_loss=30.97, generator_mel_loss=21.32, generator_kl_loss=1.952, generator_dur_loss=1.655, generator_adv_loss=2.264, generator_feat_match_loss=3.777, over 65.00 samples.], tot_loss[discriminator_loss=2.632, discriminator_real_loss=1.34, discriminator_fake_loss=1.292, generator_loss=30.45, generator_mel_loss=21.05, generator_kl_loss=1.984, generator_dur_loss=1.666, generator_adv_loss=2.126, generator_feat_match_loss=3.623, over 324.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:12:58,592 INFO [train.py:811] (3/4) Start epoch 460
2023-11-14 00:14:48,881 INFO [train.py:467] (3/4) Epoch 460, batch 17, global_batch_idx: 17000, batch size: 153, loss[discriminator_loss=2.936, discriminator_real_loss=1.349, discriminator_fake_loss=1.587, generator_loss=30.38, generator_mel_loss=21.47, generator_kl_loss=1.96, generator_dur_loss=1.634, generator_adv_loss=1.945, generator_feat_match_loss=3.371, over 153.00 samples.], tot_loss[discriminator_loss=2.682, discriminator_real_loss=1.337, discriminator_fake_loss=1.345, generator_loss=30.84, generator_mel_loss=20.94, generator_kl_loss=1.976, generator_dur_loss=1.674, generator_adv_loss=2.283, generator_feat_match_loss=3.974, over 1309.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:14:48,882 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 00:15:00,454 INFO [train.py:517] (3/4) Epoch 460, validation: discriminator_loss=2.665, discriminator_real_loss=1.441, discriminator_fake_loss=1.223, generator_loss=30.68, generator_mel_loss=21.55, generator_kl_loss=2.03, generator_dur_loss=1.65, generator_adv_loss=2.015, generator_feat_match_loss=3.434, over 100.00 samples.
2023-11-14 00:15:00,455 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 00:16:50,806 INFO [train.py:811] (3/4) Start epoch 461
2023-11-14 00:19:53,257 INFO [train.py:467] (3/4) Epoch 461, batch 30, global_batch_idx: 17050, batch size: 82, loss[discriminator_loss=2.605, discriminator_real_loss=1.334, discriminator_fake_loss=1.272, generator_loss=30.24, generator_mel_loss=20.9, generator_kl_loss=1.967, generator_dur_loss=1.659, generator_adv_loss=2.08, generator_feat_match_loss=3.633, over 82.00 samples.], tot_loss[discriminator_loss=2.629, discriminator_real_loss=1.33, discriminator_fake_loss=1.299, generator_loss=30.26, generator_mel_loss=20.9, generator_kl_loss=1.99, generator_dur_loss=1.668, generator_adv_loss=2.078, generator_feat_match_loss=3.62, over 2369.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:20:26,939 INFO [train.py:811] (3/4) Start epoch 462
2023-11-14 00:23:57,280 INFO [train.py:811] (3/4) Start epoch 463
2023-11-14 00:24:49,984 INFO [train.py:467] (3/4) Epoch 463, batch 6, global_batch_idx: 17100, batch size: 71, loss[discriminator_loss=2.441, discriminator_real_loss=1.169, discriminator_fake_loss=1.273, generator_loss=31.33, generator_mel_loss=21.03, generator_kl_loss=2.045, generator_dur_loss=1.662, generator_adv_loss=2.297, generator_feat_match_loss=4.301, over 71.00 samples.], tot_loss[discriminator_loss=2.491, discriminator_real_loss=1.248, discriminator_fake_loss=1.244, generator_loss=30.86, generator_mel_loss=20.84, generator_kl_loss=1.995, generator_dur_loss=1.666, generator_adv_loss=2.239, generator_feat_match_loss=4.119, over 563.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:27:31,284 INFO [train.py:811] (3/4) Start epoch 464
2023-11-14 00:29:25,879 INFO [train.py:467] (3/4) Epoch 464, batch 19, global_batch_idx: 17150, batch size: 71, loss[discriminator_loss=2.586, discriminator_real_loss=1.271, discriminator_fake_loss=1.315, generator_loss=30.04, generator_mel_loss=20.63, generator_kl_loss=1.813, generator_dur_loss=1.657, generator_adv_loss=2.164, generator_feat_match_loss=3.779, over 71.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.305, discriminator_fake_loss=1.276, generator_loss=30.51, generator_mel_loss=20.85, generator_kl_loss=1.976, generator_dur_loss=1.667, generator_adv_loss=2.202, generator_feat_match_loss=3.814, over 1331.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:30:59,245 INFO [train.py:811] (3/4) Start epoch 465
2023-11-14 00:34:09,585 INFO [train.py:467] (3/4) Epoch 465, batch 32, global_batch_idx: 17200, batch size: 126, loss[discriminator_loss=2.742, discriminator_real_loss=1.302, discriminator_fake_loss=1.439, generator_loss=30.29, generator_mel_loss=21.04, generator_kl_loss=2.053, generator_dur_loss=1.662, generator_adv_loss=2.062, generator_feat_match_loss=3.475, over 126.00 samples.], tot_loss[discriminator_loss=2.614, discriminator_real_loss=1.327, discriminator_fake_loss=1.287, generator_loss=30.59, generator_mel_loss=20.82, generator_kl_loss=1.989, generator_dur_loss=1.663, generator_adv_loss=2.227, generator_feat_match_loss=3.891, over 2540.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 32.0
2023-11-14 00:34:09,587 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 00:34:20,938 INFO [train.py:517] (3/4) Epoch 465, validation: discriminator_loss=2.597, discriminator_real_loss=1.297, discriminator_fake_loss=1.3, generator_loss=31.33, generator_mel_loss=21.9, generator_kl_loss=2.121, generator_dur_loss=1.642, generator_adv_loss=2.012, generator_feat_match_loss=3.655, over 100.00 samples.
2023-11-14 00:34:20,939 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 00:34:43,295 INFO [train.py:811] (3/4) Start epoch 466
2023-11-14 00:38:13,489 INFO [train.py:811] (3/4) Start epoch 467
2023-11-14 00:39:09,823 INFO [train.py:467] (3/4) Epoch 467, batch 8, global_batch_idx: 17250, batch size: 60, loss[discriminator_loss=2.551, discriminator_real_loss=1.285, discriminator_fake_loss=1.267, generator_loss=30.54, generator_mel_loss=21.01, generator_kl_loss=2.097, generator_dur_loss=1.677, generator_adv_loss=2.002, generator_feat_match_loss=3.746, over 60.00 samples.], tot_loss[discriminator_loss=2.595, discriminator_real_loss=1.331, discriminator_fake_loss=1.264, generator_loss=30.41, generator_mel_loss=20.95, generator_kl_loss=2.028, generator_dur_loss=1.668, generator_adv_loss=2.095, generator_feat_match_loss=3.66, over 620.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:41:42,607 INFO [train.py:811] (3/4) Start epoch 468
2023-11-14 00:43:57,897 INFO [train.py:467] (3/4) Epoch 468, batch 21, global_batch_idx: 17300, batch size: 69, loss[discriminator_loss=2.566, discriminator_real_loss=1.182, discriminator_fake_loss=1.385, generator_loss=29.97, generator_mel_loss=20.26, generator_kl_loss=1.893, generator_dur_loss=1.665, generator_adv_loss=2.166, generator_feat_match_loss=3.99, over 69.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.331, discriminator_fake_loss=1.295, generator_loss=30.46, generator_mel_loss=20.67, generator_kl_loss=1.968, generator_dur_loss=1.661, generator_adv_loss=2.241, generator_feat_match_loss=3.917, over 1790.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:45:19,363 INFO [train.py:811] (3/4) Start epoch 469
2023-11-14 00:48:44,773 INFO [train.py:467] (3/4) Epoch 469, batch 34, global_batch_idx: 17350, batch size: 65, loss[discriminator_loss=2.549, discriminator_real_loss=1.254, discriminator_fake_loss=1.295, generator_loss=30.71, generator_mel_loss=20.81, generator_kl_loss=2.095, generator_dur_loss=1.66, generator_adv_loss=2.221, generator_feat_match_loss=3.918, over 65.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.332, discriminator_fake_loss=1.282, generator_loss=30.26, generator_mel_loss=20.85, generator_kl_loss=1.991, generator_dur_loss=1.665, generator_adv_loss=2.096, generator_feat_match_loss=3.657, over 2519.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:48:54,021 INFO [train.py:811] (3/4) Start epoch 470
2023-11-14 00:52:27,322 INFO [train.py:811] (3/4) Start epoch 471
2023-11-14 00:53:43,770 INFO [train.py:467] (3/4) Epoch 471, batch 10, global_batch_idx: 17400, batch size: 59, loss[discriminator_loss=2.609, discriminator_real_loss=1.266, discriminator_fake_loss=1.344, generator_loss=31.27, generator_mel_loss=21.29, generator_kl_loss=1.967, generator_dur_loss=1.696, generator_adv_loss=2.238, generator_feat_match_loss=4.074, over 59.00 samples.], tot_loss[discriminator_loss=2.633, discriminator_real_loss=1.342, discriminator_fake_loss=1.291, generator_loss=30.75, generator_mel_loss=20.99, generator_kl_loss=2.006, generator_dur_loss=1.665, generator_adv_loss=2.217, generator_feat_match_loss=3.866, over 853.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:53:43,772 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 00:53:55,627 INFO [train.py:517] (3/4) Epoch 471, validation: discriminator_loss=2.586, discriminator_real_loss=1.263, discriminator_fake_loss=1.323, generator_loss=31.27, generator_mel_loss=21.78, generator_kl_loss=2.146, generator_dur_loss=1.636, generator_adv_loss=1.985, generator_feat_match_loss=3.723, over 100.00 samples.
2023-11-14 00:53:55,629 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 00:56:15,801 INFO [train.py:811] (3/4) Start epoch 472
2023-11-14 00:58:31,895 INFO [train.py:467] (3/4) Epoch 472, batch 23, global_batch_idx: 17450, batch size: 50, loss[discriminator_loss=2.645, discriminator_real_loss=1.281, discriminator_fake_loss=1.362, generator_loss=30.36, generator_mel_loss=21.03, generator_kl_loss=1.994, generator_dur_loss=1.647, generator_adv_loss=2.055, generator_feat_match_loss=3.633, over 50.00 samples.], tot_loss[discriminator_loss=2.659, discriminator_real_loss=1.347, discriminator_fake_loss=1.311, generator_loss=30.33, generator_mel_loss=20.93, generator_kl_loss=1.96, generator_dur_loss=1.667, generator_adv_loss=2.14, generator_feat_match_loss=3.635, over 1602.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 00:59:52,778 INFO [train.py:811] (3/4) Start epoch 473
2023-11-14 01:03:24,353 INFO [train.py:467] (3/4) Epoch 473, batch 36, global_batch_idx: 17500, batch size: 59, loss[discriminator_loss=2.648, discriminator_real_loss=1.32, discriminator_fake_loss=1.329, generator_loss=30.15, generator_mel_loss=20.83, generator_kl_loss=2.021, generator_dur_loss=1.66, generator_adv_loss=2.117, generator_feat_match_loss=3.527, over 59.00 samples.], tot_loss[discriminator_loss=2.636, discriminator_real_loss=1.34, discriminator_fake_loss=1.296, generator_loss=30.47, generator_mel_loss=20.92, generator_kl_loss=1.971, generator_dur_loss=1.663, generator_adv_loss=2.187, generator_feat_match_loss=3.734, over 2776.00 samples.], cur_lr_g: 1.89e-04, cur_lr_d: 1.89e-04, grad_scale: 16.0
2023-11-14 01:03:24,997 INFO [train.py:811] (3/4) Start epoch 474
2023-11-14 01:06:51,615 INFO [train.py:811] (3/4) Start epoch 475
2023-11-14 01:08:17,280 INFO [train.py:467] (3/4) Epoch 475, batch 12, global_batch_idx: 17550, batch size: 67, loss[discriminator_loss=2.73, discriminator_real_loss=1.296, discriminator_fake_loss=1.434, generator_loss=29.61, generator_mel_loss=20.7, generator_kl_loss=1.927, generator_dur_loss=1.685, generator_adv_loss=2.102, generator_feat_match_loss=3.199, over 67.00 samples.], tot_loss[discriminator_loss=2.639, discriminator_real_loss=1.338, discriminator_fake_loss=1.301, generator_loss=30.37, generator_mel_loss=20.66, generator_kl_loss=1.93, generator_dur_loss=1.666, generator_adv_loss=2.215, generator_feat_match_loss=3.9, over 975.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:10:24,914 INFO [train.py:811] (3/4) Start epoch 476
2023-11-14 01:13:03,747 INFO [train.py:467] (3/4) Epoch 476, batch 25, global_batch_idx: 17600, batch size: 153, loss[discriminator_loss=2.584, discriminator_real_loss=1.279, discriminator_fake_loss=1.305, generator_loss=30.67, generator_mel_loss=21.09, generator_kl_loss=2.02, generator_dur_loss=1.659, generator_adv_loss=2.008, generator_feat_match_loss=3.895, over 153.00 samples.], tot_loss[discriminator_loss=2.603, discriminator_real_loss=1.32, discriminator_fake_loss=1.282, generator_loss=30.34, generator_mel_loss=20.91, generator_kl_loss=2.008, generator_dur_loss=1.659, generator_adv_loss=2.074, generator_feat_match_loss=3.693, over 1927.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 32.0
2023-11-14 01:13:03,748 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 01:13:14,841 INFO [train.py:517] (3/4) Epoch 476, validation: discriminator_loss=2.622, discriminator_real_loss=1.184, discriminator_fake_loss=1.438, generator_loss=31.07, generator_mel_loss=21.51, generator_kl_loss=2.189, generator_dur_loss=1.645, generator_adv_loss=1.809, generator_feat_match_loss=3.92, over 100.00 samples.
2023-11-14 01:13:14,843 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 01:14:08,246 INFO [train.py:811] (3/4) Start epoch 477
2023-11-14 01:17:42,966 INFO [train.py:811] (3/4) Start epoch 478
2023-11-14 01:18:05,250 INFO [train.py:467] (3/4) Epoch 478, batch 1, global_batch_idx: 17650, batch size: 61, loss[discriminator_loss=2.516, discriminator_real_loss=1.271, discriminator_fake_loss=1.244, generator_loss=30.86, generator_mel_loss=20.81, generator_kl_loss=2.167, generator_dur_loss=1.658, generator_adv_loss=2.186, generator_feat_match_loss=4.039, over 61.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.331, discriminator_fake_loss=1.249, generator_loss=30.84, generator_mel_loss=21.11, generator_kl_loss=2.077, generator_dur_loss=1.667, generator_adv_loss=2.103, generator_feat_match_loss=3.879, over 187.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 32.0
2023-11-14 01:21:20,129 INFO [train.py:811] (3/4) Start epoch 479
2023-11-14 01:22:52,515 INFO [train.py:467] (3/4) Epoch 479, batch 14, global_batch_idx: 17700, batch size: 50, loss[discriminator_loss=2.377, discriminator_real_loss=1.238, discriminator_fake_loss=1.139, generator_loss=30.96, generator_mel_loss=20.18, generator_kl_loss=2.018, generator_dur_loss=1.688, generator_adv_loss=2.48, generator_feat_match_loss=4.59, over 50.00 samples.], tot_loss[discriminator_loss=2.594, discriminator_real_loss=1.307, discriminator_fake_loss=1.287, generator_loss=30.65, generator_mel_loss=20.74, generator_kl_loss=1.982, generator_dur_loss=1.662, generator_adv_loss=2.254, generator_feat_match_loss=4.012, over 1028.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:24:53,246 INFO [train.py:811] (3/4) Start epoch 480
2023-11-14 01:27:37,058 INFO [train.py:467] (3/4) Epoch 480, batch 27, global_batch_idx: 17750, batch size: 153, loss[discriminator_loss=2.57, discriminator_real_loss=1.264, discriminator_fake_loss=1.307, generator_loss=30.45, generator_mel_loss=20.76, generator_kl_loss=1.937, generator_dur_loss=1.66, generator_adv_loss=2.068, generator_feat_match_loss=4.027, over 153.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.304, discriminator_fake_loss=1.287, generator_loss=30.28, generator_mel_loss=20.69, generator_kl_loss=1.978, generator_dur_loss=1.662, generator_adv_loss=2.145, generator_feat_match_loss=3.803, over 2261.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:28:22,278 INFO [train.py:811] (3/4) Start epoch 481
2023-11-14 01:31:54,827 INFO [train.py:811] (3/4) Start epoch 482
2023-11-14 01:32:24,363 INFO [train.py:467] (3/4) Epoch 482, batch 3, global_batch_idx: 17800, batch size: 76, loss[discriminator_loss=2.559, discriminator_real_loss=1.275, discriminator_fake_loss=1.283, generator_loss=30.98, generator_mel_loss=20.75, generator_kl_loss=1.951, generator_dur_loss=1.656, generator_adv_loss=2.664, generator_feat_match_loss=3.955, over 76.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.298, discriminator_fake_loss=1.275, generator_loss=30.77, generator_mel_loss=20.8, generator_kl_loss=1.969, generator_dur_loss=1.675, generator_adv_loss=2.33, generator_feat_match_loss=3.992, over 250.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:32:24,365 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 01:32:36,718 INFO [train.py:517] (3/4) Epoch 482, validation: discriminator_loss=2.663, discriminator_real_loss=1.5, discriminator_fake_loss=1.163, generator_loss=31.29, generator_mel_loss=21.22, generator_kl_loss=2.11, generator_dur_loss=1.642, generator_adv_loss=2.364, generator_feat_match_loss=3.958, over 100.00 samples.
2023-11-14 01:32:36,719 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 01:35:40,367 INFO [train.py:811] (3/4) Start epoch 483
2023-11-14 01:37:24,203 INFO [train.py:467] (3/4) Epoch 483, batch 16, global_batch_idx: 17850, batch size: 55, loss[discriminator_loss=2.699, discriminator_real_loss=1.356, discriminator_fake_loss=1.344, generator_loss=29.81, generator_mel_loss=20.49, generator_kl_loss=2.102, generator_dur_loss=1.708, generator_adv_loss=2.123, generator_feat_match_loss=3.396, over 55.00 samples.], tot_loss[discriminator_loss=2.694, discriminator_real_loss=1.378, discriminator_fake_loss=1.316, generator_loss=30.37, generator_mel_loss=20.95, generator_kl_loss=1.999, generator_dur_loss=1.669, generator_adv_loss=2.117, generator_feat_match_loss=3.641, over 1153.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:39:16,043 INFO [train.py:811] (3/4) Start epoch 484
2023-11-14 01:41:58,328 INFO [train.py:467] (3/4) Epoch 484, batch 29, global_batch_idx: 17900, batch size: 50, loss[discriminator_loss=2.459, discriminator_real_loss=1.322, discriminator_fake_loss=1.137, generator_loss=30.77, generator_mel_loss=20.49, generator_kl_loss=1.933, generator_dur_loss=1.672, generator_adv_loss=2.285, generator_feat_match_loss=4.395, over 50.00 samples.], tot_loss[discriminator_loss=2.647, discriminator_real_loss=1.342, discriminator_fake_loss=1.305, generator_loss=30.53, generator_mel_loss=20.89, generator_kl_loss=2.002, generator_dur_loss=1.654, generator_adv_loss=2.205, generator_feat_match_loss=3.779, over 2255.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:42:42,894 INFO [train.py:811] (3/4) Start epoch 485
2023-11-14 01:46:21,848 INFO [train.py:811] (3/4) Start epoch 486
2023-11-14 01:47:01,166 INFO [train.py:467] (3/4) Epoch 486, batch 5, global_batch_idx: 17950, batch size: 95, loss[discriminator_loss=2.418, discriminator_real_loss=1.145, discriminator_fake_loss=1.272, generator_loss=32.02, generator_mel_loss=21.06, generator_kl_loss=2.081, generator_dur_loss=1.644, generator_adv_loss=2.582, generator_feat_match_loss=4.648, over 95.00 samples.], tot_loss[discriminator_loss=2.555, discriminator_real_loss=1.266, discriminator_fake_loss=1.289, generator_loss=30.59, generator_mel_loss=20.61, generator_kl_loss=1.996, generator_dur_loss=1.66, generator_adv_loss=2.274, generator_feat_match_loss=4.048, over 429.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:49:57,252 INFO [train.py:811] (3/4) Start epoch 487
2023-11-14 01:51:46,751 INFO [train.py:467] (3/4) Epoch 487, batch 18, global_batch_idx: 18000, batch size: 52, loss[discriminator_loss=2.574, discriminator_real_loss=1.341, discriminator_fake_loss=1.234, generator_loss=30.26, generator_mel_loss=20.43, generator_kl_loss=1.959, generator_dur_loss=1.673, generator_adv_loss=2.305, generator_feat_match_loss=3.891, over 52.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.305, discriminator_fake_loss=1.259, generator_loss=30.43, generator_mel_loss=20.7, generator_kl_loss=2.024, generator_dur_loss=1.656, generator_adv_loss=2.185, generator_feat_match_loss=3.872, over 1370.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:51:46,753 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 01:51:57,850 INFO [train.py:517] (3/4) Epoch 487, validation: discriminator_loss=2.564, discriminator_real_loss=1.254, discriminator_fake_loss=1.31, generator_loss=31.05, generator_mel_loss=21.43, generator_kl_loss=2.126, generator_dur_loss=1.647, generator_adv_loss=1.936, generator_feat_match_loss=3.913, over 100.00 samples.
2023-11-14 01:51:57,851 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 01:53:34,670 INFO [train.py:811] (3/4) Start epoch 488
2023-11-14 01:56:43,699 INFO [train.py:467] (3/4) Epoch 488, batch 31, global_batch_idx: 18050, batch size: 60, loss[discriminator_loss=2.598, discriminator_real_loss=1.288, discriminator_fake_loss=1.311, generator_loss=30.37, generator_mel_loss=20.78, generator_kl_loss=1.955, generator_dur_loss=1.687, generator_adv_loss=2.166, generator_feat_match_loss=3.781, over 60.00 samples.], tot_loss[discriminator_loss=2.576, discriminator_real_loss=1.308, discriminator_fake_loss=1.268, generator_loss=30.32, generator_mel_loss=20.5, generator_kl_loss=1.986, generator_dur_loss=1.663, generator_adv_loss=2.225, generator_feat_match_loss=3.944, over 2212.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 01:57:11,633 INFO [train.py:811] (3/4) Start epoch 489
2023-11-14 02:00:41,268 INFO [train.py:811] (3/4) Start epoch 490
2023-11-14 02:01:41,353 INFO [train.py:467] (3/4) Epoch 490, batch 7, global_batch_idx: 18100, batch size: 79, loss[discriminator_loss=2.715, discriminator_real_loss=1.504, discriminator_fake_loss=1.211, generator_loss=30.64, generator_mel_loss=21.12, generator_kl_loss=2.002, generator_dur_loss=1.635, generator_adv_loss=2.008, generator_feat_match_loss=3.875, over 79.00 samples.], tot_loss[discriminator_loss=2.672, discriminator_real_loss=1.365, discriminator_fake_loss=1.307, generator_loss=30.76, generator_mel_loss=21.08, generator_kl_loss=1.984, generator_dur_loss=1.662, generator_adv_loss=2.192, generator_feat_match_loss=3.842, over 570.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:04:13,321 INFO [train.py:811] (3/4) Start epoch 491
2023-11-14 02:06:06,663 INFO [train.py:467] (3/4) Epoch 491, batch 20, global_batch_idx: 18150, batch size: 59, loss[discriminator_loss=2.582, discriminator_real_loss=1.309, discriminator_fake_loss=1.272, generator_loss=30.51, generator_mel_loss=20.99, generator_kl_loss=1.879, generator_dur_loss=1.681, generator_adv_loss=2.162, generator_feat_match_loss=3.795, over 59.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.341, discriminator_fake_loss=1.306, generator_loss=30.2, generator_mel_loss=20.76, generator_kl_loss=1.968, generator_dur_loss=1.663, generator_adv_loss=2.137, generator_feat_match_loss=3.679, over 1492.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:07:44,494 INFO [train.py:811] (3/4) Start epoch 492
2023-11-14 02:11:00,681 INFO [train.py:467] (3/4) Epoch 492, batch 33, global_batch_idx: 18200, batch size: 59, loss[discriminator_loss=2.629, discriminator_real_loss=1.309, discriminator_fake_loss=1.321, generator_loss=30.62, generator_mel_loss=20.75, generator_kl_loss=1.972, generator_dur_loss=1.649, generator_adv_loss=2.391, generator_feat_match_loss=3.857, over 59.00 samples.], tot_loss[discriminator_loss=2.629, discriminator_real_loss=1.327, discriminator_fake_loss=1.301, generator_loss=30.54, generator_mel_loss=20.93, generator_kl_loss=1.988, generator_dur_loss=1.656, generator_adv_loss=2.168, generator_feat_match_loss=3.799, over 2727.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:11:00,682 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 02:11:12,291 INFO [train.py:517] (3/4) Epoch 492, validation: discriminator_loss=2.553, discriminator_real_loss=1.265, discriminator_fake_loss=1.288, generator_loss=31.02, generator_mel_loss=21.31, generator_kl_loss=2.021, generator_dur_loss=1.642, generator_adv_loss=2.181, generator_feat_match_loss=3.864, over 100.00 samples.
2023-11-14 02:11:12,292 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 02:11:30,513 INFO [train.py:811] (3/4) Start epoch 493
2023-11-14 02:15:04,376 INFO [train.py:811] (3/4) Start epoch 494
2023-11-14 02:16:06,185 INFO [train.py:467] (3/4) Epoch 494, batch 9, global_batch_idx: 18250, batch size: 153, loss[discriminator_loss=2.508, discriminator_real_loss=1.294, discriminator_fake_loss=1.215, generator_loss=30.94, generator_mel_loss=20.8, generator_kl_loss=2.042, generator_dur_loss=1.633, generator_adv_loss=2.205, generator_feat_match_loss=4.262, over 153.00 samples.], tot_loss[discriminator_loss=2.601, discriminator_real_loss=1.323, discriminator_fake_loss=1.278, generator_loss=30.4, generator_mel_loss=20.72, generator_kl_loss=1.989, generator_dur_loss=1.654, generator_adv_loss=2.149, generator_feat_match_loss=3.895, over 808.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:18:35,776 INFO [train.py:811] (3/4) Start epoch 495
2023-11-14 02:20:43,499 INFO [train.py:467] (3/4) Epoch 495, batch 22, global_batch_idx: 18300, batch size: 85, loss[discriminator_loss=2.705, discriminator_real_loss=1.295, discriminator_fake_loss=1.41, generator_loss=30.4, generator_mel_loss=21.14, generator_kl_loss=2.049, generator_dur_loss=1.631, generator_adv_loss=2.08, generator_feat_match_loss=3.5, over 85.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.338, discriminator_fake_loss=1.29, generator_loss=30.35, generator_mel_loss=20.79, generator_kl_loss=2.016, generator_dur_loss=1.661, generator_adv_loss=2.116, generator_feat_match_loss=3.768, over 1693.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:22:06,622 INFO [train.py:811] (3/4) Start epoch 496
2023-11-14 02:25:33,414 INFO [train.py:467] (3/4) Epoch 496, batch 35, global_batch_idx: 18350, batch size: 63, loss[discriminator_loss=2.711, discriminator_real_loss=1.485, discriminator_fake_loss=1.227, generator_loss=29.77, generator_mel_loss=20.87, generator_kl_loss=1.87, generator_dur_loss=1.696, generator_adv_loss=1.91, generator_feat_match_loss=3.42, over 63.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.352, discriminator_fake_loss=1.3, generator_loss=30.25, generator_mel_loss=20.78, generator_kl_loss=1.973, generator_dur_loss=1.664, generator_adv_loss=2.126, generator_feat_match_loss=3.714, over 2389.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:25:41,707 INFO [train.py:811] (3/4) Start epoch 497
2023-11-14 02:29:15,811 INFO [train.py:811] (3/4) Start epoch 498
2023-11-14 02:30:30,786 INFO [train.py:467] (3/4) Epoch 498, batch 11, global_batch_idx: 18400, batch size: 54, loss[discriminator_loss=2.715, discriminator_real_loss=1.542, discriminator_fake_loss=1.172, generator_loss=30.64, generator_mel_loss=20.96, generator_kl_loss=2.07, generator_dur_loss=1.674, generator_adv_loss=2.422, generator_feat_match_loss=3.512, over 54.00 samples.], tot_loss[discriminator_loss=2.686, discriminator_real_loss=1.371, discriminator_fake_loss=1.315, generator_loss=30.49, generator_mel_loss=20.88, generator_kl_loss=1.972, generator_dur_loss=1.667, generator_adv_loss=2.217, generator_feat_match_loss=3.754, over 799.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 32.0
2023-11-14 02:30:30,788 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 02:30:41,822 INFO [train.py:517] (3/4) Epoch 498, validation: discriminator_loss=2.681, discriminator_real_loss=1.591, discriminator_fake_loss=1.09, generator_loss=31.44, generator_mel_loss=21.5, generator_kl_loss=2.11, generator_dur_loss=1.651, generator_adv_loss=2.465, generator_feat_match_loss=3.706, over 100.00 samples.
2023-11-14 02:30:41,823 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 02:33:01,550 INFO [train.py:811] (3/4) Start epoch 499
2023-11-14 02:35:21,428 INFO [train.py:467] (3/4) Epoch 499, batch 24, global_batch_idx: 18450, batch size: 69, loss[discriminator_loss=2.555, discriminator_real_loss=1.294, discriminator_fake_loss=1.262, generator_loss=30.05, generator_mel_loss=20.56, generator_kl_loss=1.988, generator_dur_loss=1.665, generator_adv_loss=2.098, generator_feat_match_loss=3.736, over 69.00 samples.], tot_loss[discriminator_loss=2.619, discriminator_real_loss=1.332, discriminator_fake_loss=1.286, generator_loss=30.11, generator_mel_loss=20.67, generator_kl_loss=1.994, generator_dur_loss=1.664, generator_adv_loss=2.089, generator_feat_match_loss=3.694, over 1708.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:36:31,813 INFO [train.py:811] (3/4) Start epoch 500
2023-11-14 02:40:01,862 INFO [train.py:811] (3/4) Start epoch 501
2023-11-14 02:40:20,083 INFO [train.py:467] (3/4) Epoch 501, batch 0, global_batch_idx: 18500, batch size: 60, loss[discriminator_loss=2.676, discriminator_real_loss=1.356, discriminator_fake_loss=1.32, generator_loss=30.96, generator_mel_loss=21.54, generator_kl_loss=2.035, generator_dur_loss=1.672, generator_adv_loss=1.96, generator_feat_match_loss=3.746, over 60.00 samples.], tot_loss[discriminator_loss=2.676, discriminator_real_loss=1.356, discriminator_fake_loss=1.32, generator_loss=30.96, generator_mel_loss=21.54, generator_kl_loss=2.035, generator_dur_loss=1.672, generator_adv_loss=1.96, generator_feat_match_loss=3.746, over 60.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:43:35,518 INFO [train.py:811] (3/4) Start epoch 502
2023-11-14 02:45:08,607 INFO [train.py:467] (3/4) Epoch 502, batch 13, global_batch_idx: 18550, batch size: 153, loss[discriminator_loss=2.594, discriminator_real_loss=1.322, discriminator_fake_loss=1.272, generator_loss=30.39, generator_mel_loss=20.82, generator_kl_loss=1.911, generator_dur_loss=1.65, generator_adv_loss=2.09, generator_feat_match_loss=3.918, over 153.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.331, discriminator_fake_loss=1.297, generator_loss=30.21, generator_mel_loss=20.78, generator_kl_loss=1.962, generator_dur_loss=1.657, generator_adv_loss=2.133, generator_feat_match_loss=3.676, over 1049.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:47:09,935 INFO [train.py:811] (3/4) Start epoch 503
2023-11-14 02:49:51,326 INFO [train.py:467] (3/4) Epoch 503, batch 26, global_batch_idx: 18600, batch size: 59, loss[discriminator_loss=2.629, discriminator_real_loss=1.393, discriminator_fake_loss=1.236, generator_loss=30.47, generator_mel_loss=20.81, generator_kl_loss=2.071, generator_dur_loss=1.672, generator_adv_loss=2.25, generator_feat_match_loss=3.672, over 59.00 samples.], tot_loss[discriminator_loss=2.598, discriminator_real_loss=1.323, discriminator_fake_loss=1.275, generator_loss=30.58, generator_mel_loss=20.79, generator_kl_loss=1.987, generator_dur_loss=1.66, generator_adv_loss=2.219, generator_feat_match_loss=3.932, over 1917.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:49:51,328 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 02:50:03,296 INFO [train.py:517] (3/4) Epoch 503, validation: discriminator_loss=2.597, discriminator_real_loss=1.343, discriminator_fake_loss=1.253, generator_loss=31.87, generator_mel_loss=21.7, generator_kl_loss=2.231, generator_dur_loss=1.642, generator_adv_loss=2.25, generator_feat_match_loss=4.045, over 100.00 samples.
2023-11-14 02:50:03,297 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 02:50:51,864 INFO [train.py:811] (3/4) Start epoch 504
2023-11-14 02:54:22,118 INFO [train.py:811] (3/4) Start epoch 505
2023-11-14 02:54:50,913 INFO [train.py:467] (3/4) Epoch 505, batch 2, global_batch_idx: 18650, batch size: 90, loss[discriminator_loss=2.674, discriminator_real_loss=1.325, discriminator_fake_loss=1.349, generator_loss=30.41, generator_mel_loss=20.98, generator_kl_loss=2.086, generator_dur_loss=1.679, generator_adv_loss=2.02, generator_feat_match_loss=3.648, over 90.00 samples.], tot_loss[discriminator_loss=2.639, discriminator_real_loss=1.341, discriminator_fake_loss=1.298, generator_loss=30.41, generator_mel_loss=20.91, generator_kl_loss=2.01, generator_dur_loss=1.673, generator_adv_loss=2.068, generator_feat_match_loss=3.743, over 213.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 02:57:54,675 INFO [train.py:811] (3/4) Start epoch 506
2023-11-14 02:59:26,038 INFO [train.py:467] (3/4) Epoch 506, batch 15, global_batch_idx: 18700, batch size: 110, loss[discriminator_loss=2.629, discriminator_real_loss=1.062, discriminator_fake_loss=1.566, generator_loss=30.03, generator_mel_loss=20.38, generator_kl_loss=1.911, generator_dur_loss=1.682, generator_adv_loss=2.197, generator_feat_match_loss=3.854, over 110.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.271, discriminator_fake_loss=1.32, generator_loss=30.5, generator_mel_loss=20.6, generator_kl_loss=1.982, generator_dur_loss=1.661, generator_adv_loss=2.233, generator_feat_match_loss=4.025, over 1181.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:01:26,300 INFO [train.py:811] (3/4) Start epoch 507
2023-11-14 03:04:12,558 INFO [train.py:467] (3/4) Epoch 507, batch 28, global_batch_idx: 18750, batch size: 81, loss[discriminator_loss=2.973, discriminator_real_loss=1.456, discriminator_fake_loss=1.516, generator_loss=29.74, generator_mel_loss=20.6, generator_kl_loss=2.024, generator_dur_loss=1.676, generator_adv_loss=1.974, generator_feat_match_loss=3.469, over 81.00 samples.], tot_loss[discriminator_loss=2.53, discriminator_real_loss=1.273, discriminator_fake_loss=1.257, generator_loss=30.69, generator_mel_loss=20.52, generator_kl_loss=1.966, generator_dur_loss=1.661, generator_adv_loss=2.331, generator_feat_match_loss=4.216, over 2040.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 8.0
2023-11-14 03:04:57,880 INFO [train.py:811] (3/4) Start epoch 508
2023-11-14 03:08:34,571 INFO [train.py:811] (3/4) Start epoch 509
2023-11-14 03:09:08,975 INFO [train.py:467] (3/4) Epoch 509, batch 4, global_batch_idx: 18800, batch size: 59, loss[discriminator_loss=2.574, discriminator_real_loss=1.244, discriminator_fake_loss=1.331, generator_loss=29.69, generator_mel_loss=20.46, generator_kl_loss=1.958, generator_dur_loss=1.674, generator_adv_loss=2.045, generator_feat_match_loss=3.551, over 59.00 samples.], tot_loss[discriminator_loss=2.608, discriminator_real_loss=1.359, discriminator_fake_loss=1.248, generator_loss=29.8, generator_mel_loss=20.38, generator_kl_loss=1.958, generator_dur_loss=1.675, generator_adv_loss=2.124, generator_feat_match_loss=3.668, over 318.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:09:08,977 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 03:09:20,575 INFO [train.py:517] (3/4) Epoch 509, validation: discriminator_loss=2.57, discriminator_real_loss=1.201, discriminator_fake_loss=1.368, generator_loss=30.81, generator_mel_loss=21.53, generator_kl_loss=2.057, generator_dur_loss=1.642, generator_adv_loss=1.856, generator_feat_match_loss=3.719, over 100.00 samples.
2023-11-14 03:09:20,576 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 03:12:16,413 INFO [train.py:811] (3/4) Start epoch 510
2023-11-14 03:14:01,917 INFO [train.py:467] (3/4) Epoch 510, batch 17, global_batch_idx: 18850, batch size: 126, loss[discriminator_loss=2.598, discriminator_real_loss=1.215, discriminator_fake_loss=1.383, generator_loss=30.82, generator_mel_loss=20.85, generator_kl_loss=1.954, generator_dur_loss=1.655, generator_adv_loss=2.449, generator_feat_match_loss=3.906, over 126.00 samples.], tot_loss[discriminator_loss=2.608, discriminator_real_loss=1.321, discriminator_fake_loss=1.287, generator_loss=30.58, generator_mel_loss=20.89, generator_kl_loss=1.983, generator_dur_loss=1.661, generator_adv_loss=2.185, generator_feat_match_loss=3.861, over 1369.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:15:45,570 INFO [train.py:811] (3/4) Start epoch 511
2023-11-14 03:18:44,818 INFO [train.py:467] (3/4) Epoch 511, batch 30, global_batch_idx: 18900, batch size: 53, loss[discriminator_loss=2.789, discriminator_real_loss=1.482, discriminator_fake_loss=1.306, generator_loss=30.63, generator_mel_loss=21.21, generator_kl_loss=2.014, generator_dur_loss=1.666, generator_adv_loss=2.051, generator_feat_match_loss=3.686, over 53.00 samples.], tot_loss[discriminator_loss=2.638, discriminator_real_loss=1.337, discriminator_fake_loss=1.301, generator_loss=30.43, generator_mel_loss=20.83, generator_kl_loss=1.99, generator_dur_loss=1.664, generator_adv_loss=2.162, generator_feat_match_loss=3.783, over 2095.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:19:21,584 INFO [train.py:811] (3/4) Start epoch 512
2023-11-14 03:22:51,614 INFO [train.py:811] (3/4) Start epoch 513
2023-11-14 03:23:44,076 INFO [train.py:467] (3/4) Epoch 513, batch 6, global_batch_idx: 18950, batch size: 154, loss[discriminator_loss=2.645, discriminator_real_loss=1.276, discriminator_fake_loss=1.369, generator_loss=31.39, generator_mel_loss=21.23, generator_kl_loss=2.007, generator_dur_loss=1.647, generator_adv_loss=2.377, generator_feat_match_loss=4.125, over 154.00 samples.], tot_loss[discriminator_loss=2.646, discriminator_real_loss=1.322, discriminator_fake_loss=1.324, generator_loss=30.88, generator_mel_loss=21.03, generator_kl_loss=2.018, generator_dur_loss=1.646, generator_adv_loss=2.213, generator_feat_match_loss=3.973, over 655.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:26:27,625 INFO [train.py:811] (3/4) Start epoch 514
2023-11-14 03:28:28,757 INFO [train.py:467] (3/4) Epoch 514, batch 19, global_batch_idx: 19000, batch size: 58, loss[discriminator_loss=2.564, discriminator_real_loss=1.332, discriminator_fake_loss=1.232, generator_loss=30.25, generator_mel_loss=20.57, generator_kl_loss=1.965, generator_dur_loss=1.662, generator_adv_loss=2.223, generator_feat_match_loss=3.836, over 58.00 samples.], tot_loss[discriminator_loss=2.586, discriminator_real_loss=1.287, discriminator_fake_loss=1.298, generator_loss=30.22, generator_mel_loss=20.64, generator_kl_loss=1.988, generator_dur_loss=1.668, generator_adv_loss=2.139, generator_feat_match_loss=3.789, over 1228.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 16.0
2023-11-14 03:28:28,758 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 03:28:39,817 INFO [train.py:517] (3/4) Epoch 514, validation: discriminator_loss=2.611, discriminator_real_loss=1.291, discriminator_fake_loss=1.319, generator_loss=30.79, generator_mel_loss=21.2, generator_kl_loss=2.047, generator_dur_loss=1.646, generator_adv_loss=1.98, generator_feat_match_loss=3.917, over 100.00 samples.
2023-11-14 03:28:39,819 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 03:30:10,293 INFO [train.py:811] (3/4) Start epoch 515
2023-11-14 03:33:17,810 INFO [train.py:467] (3/4) Epoch 515, batch 32, global_batch_idx: 19050, batch size: 52, loss[discriminator_loss=2.496, discriminator_real_loss=1.315, discriminator_fake_loss=1.18, generator_loss=31.48, generator_mel_loss=20.47, generator_kl_loss=2.03, generator_dur_loss=1.657, generator_adv_loss=2.584, generator_feat_match_loss=4.738, over 52.00 samples.], tot_loss[discriminator_loss=2.58, discriminator_real_loss=1.317, discriminator_fake_loss=1.263, generator_loss=30.6, generator_mel_loss=20.71, generator_kl_loss=1.992, generator_dur_loss=1.657, generator_adv_loss=2.247, generator_feat_match_loss=3.996, over 2381.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 8.0
2023-11-14 03:33:45,479 INFO [train.py:811] (3/4) Start epoch 516
2023-11-14 03:37:17,796 INFO [train.py:811] (3/4) Start epoch 517
2023-11-14 03:38:13,516 INFO [train.py:467] (3/4) Epoch 517, batch 8, global_batch_idx: 19100, batch size: 59, loss[discriminator_loss=2.588, discriminator_real_loss=1.309, discriminator_fake_loss=1.279, generator_loss=30.37, generator_mel_loss=20.65, generator_kl_loss=2.041, generator_dur_loss=1.658, generator_adv_loss=2.158, generator_feat_match_loss=3.867, over 59.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.327, discriminator_fake_loss=1.287, generator_loss=30.16, generator_mel_loss=20.64, generator_kl_loss=1.983, generator_dur_loss=1.658, generator_adv_loss=2.097, generator_feat_match_loss=3.783, over 655.00 samples.], cur_lr_g: 1.88e-04, cur_lr_d: 1.88e-04, grad_scale: 8.0
2023-11-14 03:40:52,703 INFO [train.py:811] (3/4) Start epoch 518
2023-11-14 03:42:59,096 INFO [train.py:467] (3/4) Epoch 518, batch 21, global_batch_idx: 19150, batch size: 65, loss[discriminator_loss=2.578, discriminator_real_loss=1.302, discriminator_fake_loss=1.275, generator_loss=30.71, generator_mel_loss=21.23, generator_kl_loss=1.917, generator_dur_loss=1.677, generator_adv_loss=2.07, generator_feat_match_loss=3.824, over 65.00 samples.], tot_loss[discriminator_loss=2.662, discriminator_real_loss=1.348, discriminator_fake_loss=1.314, generator_loss=30.3, generator_mel_loss=20.78, generator_kl_loss=2.002, generator_dur_loss=1.66, generator_adv_loss=2.099, generator_feat_match_loss=3.751, over 1455.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 03:44:30,012 INFO [train.py:811] (3/4) Start epoch 519
2023-11-14 03:47:46,554 INFO [train.py:467] (3/4) Epoch 519, batch 34, global_batch_idx: 19200, batch size: 55, loss[discriminator_loss=2.641, discriminator_real_loss=1.313, discriminator_fake_loss=1.328, generator_loss=30.26, generator_mel_loss=20.68, generator_kl_loss=2.094, generator_dur_loss=1.69, generator_adv_loss=2.014, generator_feat_match_loss=3.785, over 55.00 samples.], tot_loss[discriminator_loss=2.609, discriminator_real_loss=1.326, discriminator_fake_loss=1.283, generator_loss=30.56, generator_mel_loss=20.85, generator_kl_loss=1.967, generator_dur_loss=1.657, generator_adv_loss=2.179, generator_feat_match_loss=3.91, over 2437.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 03:47:46,555 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 03:47:57,491 INFO [train.py:517] (3/4) Epoch 519, validation: discriminator_loss=2.64, discriminator_real_loss=1.206, discriminator_fake_loss=1.434, generator_loss=30.82, generator_mel_loss=21.42, generator_kl_loss=2.125, generator_dur_loss=1.645, generator_adv_loss=1.822, generator_feat_match_loss=3.804, over 100.00 samples.
2023-11-14 03:47:57,492 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 03:48:06,438 INFO [train.py:811] (3/4) Start epoch 520
2023-11-14 03:51:42,262 INFO [train.py:811] (3/4) Start epoch 521
2023-11-14 03:52:53,618 INFO [train.py:467] (3/4) Epoch 521, batch 10, global_batch_idx: 19250, batch size: 61, loss[discriminator_loss=2.627, discriminator_real_loss=1.262, discriminator_fake_loss=1.365, generator_loss=30, generator_mel_loss=20.64, generator_kl_loss=1.987, generator_dur_loss=1.688, generator_adv_loss=2.127, generator_feat_match_loss=3.557, over 61.00 samples.], tot_loss[discriminator_loss=2.67, discriminator_real_loss=1.379, discriminator_fake_loss=1.291, generator_loss=29.99, generator_mel_loss=20.52, generator_kl_loss=1.984, generator_dur_loss=1.664, generator_adv_loss=2.111, generator_feat_match_loss=3.72, over 783.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 03:55:13,820 INFO [train.py:811] (3/4) Start epoch 522
2023-11-14 03:57:33,717 INFO [train.py:467] (3/4) Epoch 522, batch 23, global_batch_idx: 19300, batch size: 59, loss[discriminator_loss=2.836, discriminator_real_loss=1.445, discriminator_fake_loss=1.391, generator_loss=29.21, generator_mel_loss=20.34, generator_kl_loss=1.906, generator_dur_loss=1.66, generator_adv_loss=1.889, generator_feat_match_loss=3.41, over 59.00 samples.], tot_loss[discriminator_loss=2.568, discriminator_real_loss=1.298, discriminator_fake_loss=1.269, generator_loss=30.76, generator_mel_loss=20.63, generator_kl_loss=1.989, generator_dur_loss=1.655, generator_adv_loss=2.305, generator_feat_match_loss=4.175, over 1841.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 03:58:39,222 INFO [train.py:811] (3/4) Start epoch 523
2023-11-14 04:02:15,326 INFO [train.py:467] (3/4) Epoch 523, batch 36, global_batch_idx: 19350, batch size: 79, loss[discriminator_loss=2.635, discriminator_real_loss=1.445, discriminator_fake_loss=1.189, generator_loss=30.02, generator_mel_loss=20.49, generator_kl_loss=1.999, generator_dur_loss=1.67, generator_adv_loss=2.186, generator_feat_match_loss=3.676, over 79.00 samples.], tot_loss[discriminator_loss=2.627, discriminator_real_loss=1.329, discriminator_fake_loss=1.297, generator_loss=30.21, generator_mel_loss=20.6, generator_kl_loss=1.988, generator_dur_loss=1.654, generator_adv_loss=2.139, generator_feat_match_loss=3.833, over 2862.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 04:02:15,934 INFO [train.py:811] (3/4) Start epoch 524
2023-11-14 04:05:43,633 INFO [train.py:811] (3/4) Start epoch 525
2023-11-14 04:07:04,516 INFO [train.py:467] (3/4) Epoch 525, batch 12, global_batch_idx: 19400, batch size: 69, loss[discriminator_loss=2.811, discriminator_real_loss=1.378, discriminator_fake_loss=1.433, generator_loss=30.21, generator_mel_loss=21.11, generator_kl_loss=1.919, generator_dur_loss=1.658, generator_adv_loss=2.008, generator_feat_match_loss=3.52, over 69.00 samples.], tot_loss[discriminator_loss=2.64, discriminator_real_loss=1.346, discriminator_fake_loss=1.294, generator_loss=30.27, generator_mel_loss=20.73, generator_kl_loss=1.956, generator_dur_loss=1.664, generator_adv_loss=2.132, generator_feat_match_loss=3.795, over 848.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 04:07:04,517 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 04:07:15,969 INFO [train.py:517] (3/4) Epoch 525, validation: discriminator_loss=2.768, discriminator_real_loss=1.453, discriminator_fake_loss=1.315, generator_loss=31.06, generator_mel_loss=21.68, generator_kl_loss=2.127, generator_dur_loss=1.643, generator_adv_loss=1.993, generator_feat_match_loss=3.615, over 100.00 samples.
2023-11-14 04:07:15,971 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 04:09:28,235 INFO [train.py:811] (3/4) Start epoch 526
2023-11-14 04:12:00,399 INFO [train.py:467] (3/4) Epoch 526, batch 25, global_batch_idx: 19450, batch size: 60, loss[discriminator_loss=2.66, discriminator_real_loss=1.249, discriminator_fake_loss=1.41, generator_loss=30.54, generator_mel_loss=21.03, generator_kl_loss=2.005, generator_dur_loss=1.675, generator_adv_loss=2.109, generator_feat_match_loss=3.723, over 60.00 samples.], tot_loss[discriminator_loss=2.651, discriminator_real_loss=1.351, discriminator_fake_loss=1.301, generator_loss=30.39, generator_mel_loss=20.86, generator_kl_loss=1.981, generator_dur_loss=1.66, generator_adv_loss=2.128, generator_feat_match_loss=3.76, over 1750.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 04:13:01,135 INFO [train.py:811] (3/4) Start epoch 527
2023-11-14 04:16:32,603 INFO [train.py:811] (3/4) Start epoch 528
2023-11-14 04:16:50,686 INFO [train.py:467] (3/4) Epoch 528, batch 1, global_batch_idx: 19500, batch size: 49, loss[discriminator_loss=2.438, discriminator_real_loss=1.244, discriminator_fake_loss=1.193, generator_loss=30.39, generator_mel_loss=20.13, generator_kl_loss=1.872, generator_dur_loss=1.671, generator_adv_loss=2.428, generator_feat_match_loss=4.289, over 49.00 samples.], tot_loss[discriminator_loss=2.44, discriminator_real_loss=1.263, discriminator_fake_loss=1.178, generator_loss=30.36, generator_mel_loss=20.15, generator_kl_loss=1.938, generator_dur_loss=1.664, generator_adv_loss=2.36, generator_feat_match_loss=4.247, over 122.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 04:20:04,684 INFO [train.py:811] (3/4) Start epoch 529
2023-11-14 04:21:42,282 INFO [train.py:467] (3/4) Epoch 529, batch 14, global_batch_idx: 19550, batch size: 73, loss[discriminator_loss=2.598, discriminator_real_loss=1.319, discriminator_fake_loss=1.278, generator_loss=30.06, generator_mel_loss=20.51, generator_kl_loss=1.957, generator_dur_loss=1.658, generator_adv_loss=2.096, generator_feat_match_loss=3.836, over 73.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.316, discriminator_fake_loss=1.307, generator_loss=30.13, generator_mel_loss=20.64, generator_kl_loss=1.962, generator_dur_loss=1.659, generator_adv_loss=2.086, generator_feat_match_loss=3.786, over 1024.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 04:23:43,115 INFO [train.py:811] (3/4) Start epoch 530
2023-11-14 04:26:23,994 INFO [train.py:467] (3/4) Epoch 530, batch 27, global_batch_idx: 19600, batch size: 101, loss[discriminator_loss=2.469, discriminator_real_loss=1.188, discriminator_fake_loss=1.281, generator_loss=30.27, generator_mel_loss=20.13, generator_kl_loss=1.94, generator_dur_loss=1.648, generator_adv_loss=2.268, generator_feat_match_loss=4.285, over 101.00 samples.], tot_loss[discriminator_loss=2.575, discriminator_real_loss=1.308, discriminator_fake_loss=1.267, generator_loss=30.59, generator_mel_loss=20.45, generator_kl_loss=1.975, generator_dur_loss=1.654, generator_adv_loss=2.271, generator_feat_match_loss=4.239, over 2151.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 32.0
2023-11-14 04:26:23,995 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 04:26:34,812 INFO [train.py:517] (3/4) Epoch 530, validation: discriminator_loss=2.489, discriminator_real_loss=1.078, discriminator_fake_loss=1.411, generator_loss=30.36, generator_mel_loss=20.86, generator_kl_loss=2.071, generator_dur_loss=1.644, generator_adv_loss=1.811, generator_feat_match_loss=3.977, over 100.00 samples.
2023-11-14 04:26:34,813 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 04:27:25,624 INFO [train.py:811] (3/4) Start epoch 531
2023-11-14 04:30:54,654 INFO [train.py:811] (3/4) Start epoch 532
2023-11-14 04:31:25,649 INFO [train.py:467] (3/4) Epoch 532, batch 3, global_batch_idx: 19650, batch size: 101, loss[discriminator_loss=2.418, discriminator_real_loss=1.157, discriminator_fake_loss=1.261, generator_loss=31.11, generator_mel_loss=20.43, generator_kl_loss=2.03, generator_dur_loss=1.648, generator_adv_loss=2.229, generator_feat_match_loss=4.777, over 101.00 samples.], tot_loss[discriminator_loss=2.708, discriminator_real_loss=1.479, discriminator_fake_loss=1.23, generator_loss=30.31, generator_mel_loss=20.44, generator_kl_loss=2, generator_dur_loss=1.645, generator_adv_loss=2.184, generator_feat_match_loss=4.036, over 323.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:34:24,253 INFO [train.py:811] (3/4) Start epoch 533
2023-11-14 04:36:10,392 INFO [train.py:467] (3/4) Epoch 533, batch 16, global_batch_idx: 19700, batch size: 54, loss[discriminator_loss=2.627, discriminator_real_loss=1.237, discriminator_fake_loss=1.39, generator_loss=30.42, generator_mel_loss=20.62, generator_kl_loss=1.984, generator_dur_loss=1.687, generator_adv_loss=2.211, generator_feat_match_loss=3.914, over 54.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.31, discriminator_fake_loss=1.271, generator_loss=30.41, generator_mel_loss=20.68, generator_kl_loss=1.992, generator_dur_loss=1.649, generator_adv_loss=2.164, generator_feat_match_loss=3.93, over 1263.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:37:59,036 INFO [train.py:811] (3/4) Start epoch 534
2023-11-14 04:40:49,992 INFO [train.py:467] (3/4) Epoch 534, batch 29, global_batch_idx: 19750, batch size: 50, loss[discriminator_loss=2.59, discriminator_real_loss=1.18, discriminator_fake_loss=1.411, generator_loss=30.43, generator_mel_loss=20.7, generator_kl_loss=2.004, generator_dur_loss=1.663, generator_adv_loss=2.32, generator_feat_match_loss=3.742, over 50.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.351, discriminator_fake_loss=1.305, generator_loss=30.27, generator_mel_loss=20.71, generator_kl_loss=1.977, generator_dur_loss=1.659, generator_adv_loss=2.141, generator_feat_match_loss=3.789, over 1968.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:41:29,085 INFO [train.py:811] (3/4) Start epoch 535
2023-11-14 04:45:03,754 INFO [train.py:811] (3/4) Start epoch 536
2023-11-14 04:45:44,562 INFO [train.py:467] (3/4) Epoch 536, batch 5, global_batch_idx: 19800, batch size: 50, loss[discriminator_loss=2.57, discriminator_real_loss=1.274, discriminator_fake_loss=1.297, generator_loss=30.51, generator_mel_loss=20.79, generator_kl_loss=1.962, generator_dur_loss=1.687, generator_adv_loss=2.057, generator_feat_match_loss=4.012, over 50.00 samples.], tot_loss[discriminator_loss=2.593, discriminator_real_loss=1.31, discriminator_fake_loss=1.284, generator_loss=30.24, generator_mel_loss=20.44, generator_kl_loss=1.956, generator_dur_loss=1.661, generator_adv_loss=2.193, generator_feat_match_loss=3.991, over 341.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:45:44,564 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 04:45:56,195 INFO [train.py:517] (3/4) Epoch 536, validation: discriminator_loss=2.601, discriminator_real_loss=1.14, discriminator_fake_loss=1.461, generator_loss=30.86, generator_mel_loss=21.6, generator_kl_loss=2.091, generator_dur_loss=1.634, generator_adv_loss=1.735, generator_feat_match_loss=3.803, over 100.00 samples.
2023-11-14 04:45:56,196 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 04:48:52,199 INFO [train.py:811] (3/4) Start epoch 537
2023-11-14 04:50:42,109 INFO [train.py:467] (3/4) Epoch 537, batch 18, global_batch_idx: 19850, batch size: 85, loss[discriminator_loss=2.555, discriminator_real_loss=1.296, discriminator_fake_loss=1.26, generator_loss=30.65, generator_mel_loss=20.93, generator_kl_loss=1.979, generator_dur_loss=1.661, generator_adv_loss=2.107, generator_feat_match_loss=3.977, over 85.00 samples.], tot_loss[discriminator_loss=2.547, discriminator_real_loss=1.287, discriminator_fake_loss=1.26, generator_loss=30.59, generator_mel_loss=20.68, generator_kl_loss=2.014, generator_dur_loss=1.656, generator_adv_loss=2.21, generator_feat_match_loss=4.033, over 1518.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:52:20,228 INFO [train.py:811] (3/4) Start epoch 538
2023-11-14 04:55:22,351 INFO [train.py:467] (3/4) Epoch 538, batch 31, global_batch_idx: 19900, batch size: 110, loss[discriminator_loss=2.5, discriminator_real_loss=1.235, discriminator_fake_loss=1.266, generator_loss=31.34, generator_mel_loss=20.7, generator_kl_loss=1.923, generator_dur_loss=1.634, generator_adv_loss=2.285, generator_feat_match_loss=4.805, over 110.00 samples.], tot_loss[discriminator_loss=2.625, discriminator_real_loss=1.339, discriminator_fake_loss=1.286, generator_loss=30.65, generator_mel_loss=20.78, generator_kl_loss=2.025, generator_dur_loss=1.656, generator_adv_loss=2.205, generator_feat_match_loss=3.983, over 2588.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 04:55:55,123 INFO [train.py:811] (3/4) Start epoch 539
2023-11-14 04:59:27,910 INFO [train.py:811] (3/4) Start epoch 540
2023-11-14 05:00:28,226 INFO [train.py:467] (3/4) Epoch 540, batch 7, global_batch_idx: 19950, batch size: 56, loss[discriminator_loss=2.586, discriminator_real_loss=1.293, discriminator_fake_loss=1.293, generator_loss=30.18, generator_mel_loss=20.74, generator_kl_loss=2.009, generator_dur_loss=1.631, generator_adv_loss=2.082, generator_feat_match_loss=3.723, over 56.00 samples.], tot_loss[discriminator_loss=2.558, discriminator_real_loss=1.294, discriminator_fake_loss=1.264, generator_loss=30.24, generator_mel_loss=20.55, generator_kl_loss=2.008, generator_dur_loss=1.655, generator_adv_loss=2.124, generator_feat_match_loss=3.898, over 594.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:02:57,566 INFO [train.py:811] (3/4) Start epoch 541
2023-11-14 05:05:03,643 INFO [train.py:467] (3/4) Epoch 541, batch 20, global_batch_idx: 20000, batch size: 73, loss[discriminator_loss=2.645, discriminator_real_loss=1.271, discriminator_fake_loss=1.375, generator_loss=30.58, generator_mel_loss=20.86, generator_kl_loss=1.98, generator_dur_loss=1.661, generator_adv_loss=2.234, generator_feat_match_loss=3.84, over 73.00 samples.], tot_loss[discriminator_loss=2.62, discriminator_real_loss=1.33, discriminator_fake_loss=1.29, generator_loss=30.3, generator_mel_loss=20.75, generator_kl_loss=1.968, generator_dur_loss=1.658, generator_adv_loss=2.126, generator_feat_match_loss=3.8, over 1578.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:05:03,645 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 05:05:14,429 INFO [train.py:517] (3/4) Epoch 541, validation: discriminator_loss=2.567, discriminator_real_loss=1.315, discriminator_fake_loss=1.253, generator_loss=30.51, generator_mel_loss=20.92, generator_kl_loss=2.065, generator_dur_loss=1.654, generator_adv_loss=2.058, generator_feat_match_loss=3.816, over 100.00 samples.
2023-11-14 05:05:14,430 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 05:06:45,417 INFO [train.py:811] (3/4) Start epoch 542
2023-11-14 05:10:00,792 INFO [train.py:467] (3/4) Epoch 542, batch 33, global_batch_idx: 20050, batch size: 67, loss[discriminator_loss=2.523, discriminator_real_loss=1.338, discriminator_fake_loss=1.186, generator_loss=30.13, generator_mel_loss=20.36, generator_kl_loss=1.9, generator_dur_loss=1.664, generator_adv_loss=2.367, generator_feat_match_loss=3.84, over 67.00 samples.], tot_loss[discriminator_loss=2.664, discriminator_real_loss=1.356, discriminator_fake_loss=1.308, generator_loss=30.33, generator_mel_loss=20.79, generator_kl_loss=1.955, generator_dur_loss=1.652, generator_adv_loss=2.147, generator_feat_match_loss=3.79, over 2431.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:10:17,145 INFO [train.py:811] (3/4) Start epoch 543
2023-11-14 05:13:50,263 INFO [train.py:811] (3/4) Start epoch 544
2023-11-14 05:14:54,326 INFO [train.py:467] (3/4) Epoch 544, batch 9, global_batch_idx: 20100, batch size: 73, loss[discriminator_loss=2.734, discriminator_real_loss=1.265, discriminator_fake_loss=1.469, generator_loss=30.5, generator_mel_loss=20.71, generator_kl_loss=2.056, generator_dur_loss=1.637, generator_adv_loss=2.293, generator_feat_match_loss=3.805, over 73.00 samples.], tot_loss[discriminator_loss=2.648, discriminator_real_loss=1.326, discriminator_fake_loss=1.322, generator_loss=30.51, generator_mel_loss=20.85, generator_kl_loss=1.98, generator_dur_loss=1.653, generator_adv_loss=2.19, generator_feat_match_loss=3.84, over 673.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:17:26,942 INFO [train.py:811] (3/4) Start epoch 545
2023-11-14 05:19:42,815 INFO [train.py:467] (3/4) Epoch 545, batch 22, global_batch_idx: 20150, batch size: 55, loss[discriminator_loss=2.398, discriminator_real_loss=1.158, discriminator_fake_loss=1.241, generator_loss=31.95, generator_mel_loss=20.43, generator_kl_loss=2.007, generator_dur_loss=1.7, generator_adv_loss=2.709, generator_feat_match_loss=5.098, over 55.00 samples.], tot_loss[discriminator_loss=2.563, discriminator_real_loss=1.286, discriminator_fake_loss=1.277, generator_loss=30.81, generator_mel_loss=20.71, generator_kl_loss=1.978, generator_dur_loss=1.652, generator_adv_loss=2.291, generator_feat_match_loss=4.185, over 1687.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:21:02,841 INFO [train.py:811] (3/4) Start epoch 546
2023-11-14 05:24:30,252 INFO [train.py:467] (3/4) Epoch 546, batch 35, global_batch_idx: 20200, batch size: 64, loss[discriminator_loss=2.648, discriminator_real_loss=1.318, discriminator_fake_loss=1.331, generator_loss=30.01, generator_mel_loss=20.4, generator_kl_loss=1.938, generator_dur_loss=1.648, generator_adv_loss=2.139, generator_feat_match_loss=3.891, over 64.00 samples.], tot_loss[discriminator_loss=2.626, discriminator_real_loss=1.333, discriminator_fake_loss=1.293, generator_loss=30.26, generator_mel_loss=20.54, generator_kl_loss=1.966, generator_dur_loss=1.656, generator_adv_loss=2.158, generator_feat_match_loss=3.941, over 2788.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:24:30,254 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 05:24:41,469 INFO [train.py:517] (3/4) Epoch 546, validation: discriminator_loss=2.577, discriminator_real_loss=1.25, discriminator_fake_loss=1.327, generator_loss=30.97, generator_mel_loss=21.25, generator_kl_loss=2.117, generator_dur_loss=1.65, generator_adv_loss=1.969, generator_feat_match_loss=3.982, over 100.00 samples.
2023-11-14 05:24:41,470 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 05:24:46,215 INFO [train.py:811] (3/4) Start epoch 547
2023-11-14 05:28:22,842 INFO [train.py:811] (3/4) Start epoch 548
2023-11-14 05:29:36,683 INFO [train.py:467] (3/4) Epoch 548, batch 11, global_batch_idx: 20250, batch size: 61, loss[discriminator_loss=2.592, discriminator_real_loss=1.289, discriminator_fake_loss=1.303, generator_loss=30.1, generator_mel_loss=20.28, generator_kl_loss=1.955, generator_dur_loss=1.672, generator_adv_loss=2.172, generator_feat_match_loss=4.023, over 61.00 samples.], tot_loss[discriminator_loss=2.681, discriminator_real_loss=1.352, discriminator_fake_loss=1.33, generator_loss=30.36, generator_mel_loss=20.76, generator_kl_loss=2.025, generator_dur_loss=1.661, generator_adv_loss=2.097, generator_feat_match_loss=3.821, over 880.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:31:58,320 INFO [train.py:811] (3/4) Start epoch 549
2023-11-14 05:34:19,715 INFO [train.py:467] (3/4) Epoch 549, batch 24, global_batch_idx: 20300, batch size: 64, loss[discriminator_loss=2.688, discriminator_real_loss=1.42, discriminator_fake_loss=1.268, generator_loss=30.15, generator_mel_loss=20.61, generator_kl_loss=1.994, generator_dur_loss=1.641, generator_adv_loss=2.084, generator_feat_match_loss=3.826, over 64.00 samples.], tot_loss[discriminator_loss=2.603, discriminator_real_loss=1.315, discriminator_fake_loss=1.288, generator_loss=30.51, generator_mel_loss=20.7, generator_kl_loss=1.99, generator_dur_loss=1.656, generator_adv_loss=2.2, generator_feat_match_loss=3.965, over 1600.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:35:25,979 INFO [train.py:811] (3/4) Start epoch 550
2023-11-14 05:38:56,485 INFO [train.py:811] (3/4) Start epoch 551
2023-11-14 05:39:12,531 INFO [train.py:467] (3/4) Epoch 551, batch 0, global_batch_idx: 20350, batch size: 85, loss[discriminator_loss=2.547, discriminator_real_loss=1.209, discriminator_fake_loss=1.339, generator_loss=30.52, generator_mel_loss=20.53, generator_kl_loss=1.948, generator_dur_loss=1.651, generator_adv_loss=2.467, generator_feat_match_loss=3.93, over 85.00 samples.], tot_loss[discriminator_loss=2.547, discriminator_real_loss=1.209, discriminator_fake_loss=1.339, generator_loss=30.52, generator_mel_loss=20.53, generator_kl_loss=1.948, generator_dur_loss=1.651, generator_adv_loss=2.467, generator_feat_match_loss=3.93, over 85.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 05:42:33,855 INFO [train.py:811] (3/4) Start epoch 552
2023-11-14 05:43:59,696 INFO [train.py:467] (3/4) Epoch 552, batch 13, global_batch_idx: 20400, batch size: 79, loss[discriminator_loss=2.668, discriminator_real_loss=1.248, discriminator_fake_loss=1.42, generator_loss=30.23, generator_mel_loss=20.76, generator_kl_loss=2.003, generator_dur_loss=1.65, generator_adv_loss=2.082, generator_feat_match_loss=3.732, over 79.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.317, discriminator_fake_loss=1.3, generator_loss=30.25, generator_mel_loss=20.62, generator_kl_loss=2.006, generator_dur_loss=1.662, generator_adv_loss=2.121, generator_feat_match_loss=3.841, over 943.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:43:59,698 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 05:44:10,831 INFO [train.py:517] (3/4) Epoch 552, validation: discriminator_loss=2.596, discriminator_real_loss=1.278, discriminator_fake_loss=1.318, generator_loss=31.21, generator_mel_loss=21.3, generator_kl_loss=2.196, generator_dur_loss=1.634, generator_adv_loss=2.004, generator_feat_match_loss=4.074, over 100.00 samples.
2023-11-14 05:44:10,832 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 05:46:20,233 INFO [train.py:811] (3/4) Start epoch 553
2023-11-14 05:48:58,518 INFO [train.py:467] (3/4) Epoch 553, batch 26, global_batch_idx: 20450, batch size: 63, loss[discriminator_loss=2.557, discriminator_real_loss=1.316, discriminator_fake_loss=1.24, generator_loss=29.3, generator_mel_loss=20.01, generator_kl_loss=1.881, generator_dur_loss=1.655, generator_adv_loss=2.078, generator_feat_match_loss=3.68, over 63.00 samples.], tot_loss[discriminator_loss=2.668, discriminator_real_loss=1.361, discriminator_fake_loss=1.307, generator_loss=30.25, generator_mel_loss=20.56, generator_kl_loss=1.99, generator_dur_loss=1.655, generator_adv_loss=2.163, generator_feat_match_loss=3.879, over 1922.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:49:54,751 INFO [train.py:811] (3/4) Start epoch 554
2023-11-14 05:53:23,243 INFO [train.py:811] (3/4) Start epoch 555
2023-11-14 05:53:51,142 INFO [train.py:467] (3/4) Epoch 555, batch 2, global_batch_idx: 20500, batch size: 153, loss[discriminator_loss=2.672, discriminator_real_loss=1.252, discriminator_fake_loss=1.42, generator_loss=30.74, generator_mel_loss=20.92, generator_kl_loss=1.949, generator_dur_loss=1.632, generator_adv_loss=1.951, generator_feat_match_loss=4.293, over 153.00 samples.], tot_loss[discriminator_loss=2.655, discriminator_real_loss=1.318, discriminator_fake_loss=1.337, generator_loss=30.58, generator_mel_loss=20.71, generator_kl_loss=1.952, generator_dur_loss=1.645, generator_adv_loss=2.045, generator_feat_match_loss=4.23, over 273.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 05:56:53,191 INFO [train.py:811] (3/4) Start epoch 556
2023-11-14 05:58:30,022 INFO [train.py:467] (3/4) Epoch 556, batch 15, global_batch_idx: 20550, batch size: 95, loss[discriminator_loss=2.436, discriminator_real_loss=1.202, discriminator_fake_loss=1.233, generator_loss=30.59, generator_mel_loss=20.15, generator_kl_loss=2.025, generator_dur_loss=1.627, generator_adv_loss=2.246, generator_feat_match_loss=4.535, over 95.00 samples.], tot_loss[discriminator_loss=2.512, discriminator_real_loss=1.272, discriminator_fake_loss=1.239, generator_loss=30.6, generator_mel_loss=20.46, generator_kl_loss=1.964, generator_dur_loss=1.648, generator_adv_loss=2.29, generator_feat_match_loss=4.238, over 1284.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 06:00:26,809 INFO [train.py:811] (3/4) Start epoch 557
2023-11-14 06:03:10,806 INFO [train.py:467] (3/4) Epoch 557, batch 28, global_batch_idx: 20600, batch size: 101, loss[discriminator_loss=2.789, discriminator_real_loss=1.57, discriminator_fake_loss=1.219, generator_loss=30.22, generator_mel_loss=20.73, generator_kl_loss=1.966, generator_dur_loss=1.659, generator_adv_loss=2.295, generator_feat_match_loss=3.566, over 101.00 samples.], tot_loss[discriminator_loss=2.57, discriminator_real_loss=1.292, discriminator_fake_loss=1.277, generator_loss=30.68, generator_mel_loss=20.51, generator_kl_loss=1.984, generator_dur_loss=1.652, generator_adv_loss=2.313, generator_feat_match_loss=4.22, over 2109.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 06:03:10,808 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 06:03:21,704 INFO [train.py:517] (3/4) Epoch 557, validation: discriminator_loss=2.709, discriminator_real_loss=1.569, discriminator_fake_loss=1.14, generator_loss=31.36, generator_mel_loss=21.37, generator_kl_loss=2.118, generator_dur_loss=1.641, generator_adv_loss=2.378, generator_feat_match_loss=3.859, over 100.00 samples.
2023-11-14 06:03:21,705 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 06:04:05,005 INFO [train.py:811] (3/4) Start epoch 558
2023-11-14 06:07:34,769 INFO [train.py:811] (3/4) Start epoch 559
2023-11-14 06:08:16,280 INFO [train.py:467] (3/4) Epoch 559, batch 4, global_batch_idx: 20650, batch size: 153, loss[discriminator_loss=2.598, discriminator_real_loss=1.311, discriminator_fake_loss=1.288, generator_loss=30.75, generator_mel_loss=20.84, generator_kl_loss=2.04, generator_dur_loss=1.63, generator_adv_loss=2.238, generator_feat_match_loss=4.004, over 153.00 samples.], tot_loss[discriminator_loss=2.579, discriminator_real_loss=1.311, discriminator_fake_loss=1.268, generator_loss=30.31, generator_mel_loss=20.6, generator_kl_loss=1.985, generator_dur_loss=1.643, generator_adv_loss=2.169, generator_feat_match_loss=3.922, over 472.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 16.0
2023-11-14 06:11:05,012 INFO [train.py:811] (3/4) Start epoch 560
2023-11-14 06:12:51,399 INFO [train.py:467] (3/4) Epoch 560, batch 17, global_batch_idx: 20700, batch size: 54, loss[discriminator_loss=2.523, discriminator_real_loss=1.223, discriminator_fake_loss=1.3, generator_loss=31.06, generator_mel_loss=20.45, generator_kl_loss=1.953, generator_dur_loss=1.653, generator_adv_loss=2.502, generator_feat_match_loss=4.5, over 54.00 samples.], tot_loss[discriminator_loss=2.611, discriminator_real_loss=1.325, discriminator_fake_loss=1.286, generator_loss=30.71, generator_mel_loss=20.64, generator_kl_loss=1.996, generator_dur_loss=1.659, generator_adv_loss=2.268, generator_feat_match_loss=4.142, over 1233.00 samples.], cur_lr_g: 1.87e-04, cur_lr_d: 1.87e-04, grad_scale: 8.0
2023-11-14 06:14:39,624 INFO [train.py:811] (3/4) Start epoch 561
2023-11-14 06:17:38,138 INFO [train.py:467] (3/4) Epoch 561, batch 30, global_batch_idx: 20750, batch size: 85, loss[discriminator_loss=2.572, discriminator_real_loss=1.406, discriminator_fake_loss=1.166, generator_loss=29.99, generator_mel_loss=20.57, generator_kl_loss=1.968, generator_dur_loss=1.645, generator_adv_loss=2.01, generator_feat_match_loss=3.795, over 85.00 samples.], tot_loss[discriminator_loss=2.551, discriminator_real_loss=1.296, discriminator_fake_loss=1.256, generator_loss=30.39, generator_mel_loss=20.48, generator_kl_loss=2.001, generator_dur_loss=1.657, generator_adv_loss=2.201, generator_feat_match_loss=4.049, over 2225.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 06:18:15,678 INFO [train.py:811] (3/4) Start epoch 562
2023-11-14 06:21:45,206 INFO [train.py:811] (3/4) Start epoch 563
2023-11-14 06:22:31,594 INFO [train.py:467] (3/4) Epoch 563, batch 6, global_batch_idx: 20800, batch size: 59, loss[discriminator_loss=2.805, discriminator_real_loss=1.616, discriminator_fake_loss=1.188, generator_loss=30.49, generator_mel_loss=20.93, generator_kl_loss=1.97, generator_dur_loss=1.666, generator_adv_loss=2.244, generator_feat_match_loss=3.684, over 59.00 samples.], tot_loss[discriminator_loss=2.673, discriminator_real_loss=1.33, discriminator_fake_loss=1.343, generator_loss=31.04, generator_mel_loss=20.86, generator_kl_loss=1.997, generator_dur_loss=1.661, generator_adv_loss=2.36, generator_feat_match_loss=4.161, over 475.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:22:31,596 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 06:22:43,365 INFO [train.py:517] (3/4) Epoch 563, validation: discriminator_loss=2.729, discriminator_real_loss=1.466, discriminator_fake_loss=1.263, generator_loss=30.51, generator_mel_loss=21, generator_kl_loss=2.106, generator_dur_loss=1.645, generator_adv_loss=2.2, generator_feat_match_loss=3.556, over 100.00 samples.
2023-11-14 06:22:43,366 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 06:25:33,029 INFO [train.py:811] (3/4) Start epoch 564
2023-11-14 06:27:34,786 INFO [train.py:467] (3/4) Epoch 564, batch 19, global_batch_idx: 20850, batch size: 110, loss[discriminator_loss=2.547, discriminator_real_loss=1.244, discriminator_fake_loss=1.302, generator_loss=30.83, generator_mel_loss=21.04, generator_kl_loss=2.013, generator_dur_loss=1.641, generator_adv_loss=2.174, generator_feat_match_loss=3.959, over 110.00 samples.], tot_loss[discriminator_loss=2.61, discriminator_real_loss=1.339, discriminator_fake_loss=1.272, generator_loss=30.49, generator_mel_loss=20.76, generator_kl_loss=2.015, generator_dur_loss=1.648, generator_adv_loss=2.152, generator_feat_match_loss=3.912, over 1494.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:29:06,290 INFO [train.py:811] (3/4) Start epoch 565
2023-11-14 06:32:14,080 INFO [train.py:467] (3/4) Epoch 565, batch 32, global_batch_idx: 20900, batch size: 73, loss[discriminator_loss=2.689, discriminator_real_loss=1.499, discriminator_fake_loss=1.19, generator_loss=30.68, generator_mel_loss=20.51, generator_kl_loss=2.021, generator_dur_loss=1.638, generator_adv_loss=2.395, generator_feat_match_loss=4.117, over 73.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.33, discriminator_fake_loss=1.27, generator_loss=30.64, generator_mel_loss=20.45, generator_kl_loss=1.992, generator_dur_loss=1.657, generator_adv_loss=2.308, generator_feat_match_loss=4.228, over 2381.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:32:37,808 INFO [train.py:811] (3/4) Start epoch 566
2023-11-14 06:36:06,466 INFO [train.py:811] (3/4) Start epoch 567
2023-11-14 06:37:03,288 INFO [train.py:467] (3/4) Epoch 567, batch 8, global_batch_idx: 20950, batch size: 69, loss[discriminator_loss=2.67, discriminator_real_loss=1.341, discriminator_fake_loss=1.329, generator_loss=29.91, generator_mel_loss=20.37, generator_kl_loss=2.008, generator_dur_loss=1.653, generator_adv_loss=2.148, generator_feat_match_loss=3.73, over 69.00 samples.], tot_loss[discriminator_loss=2.65, discriminator_real_loss=1.327, discriminator_fake_loss=1.323, generator_loss=30.28, generator_mel_loss=20.69, generator_kl_loss=1.981, generator_dur_loss=1.644, generator_adv_loss=2.104, generator_feat_match_loss=3.864, over 615.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:39:37,637 INFO [train.py:811] (3/4) Start epoch 568
2023-11-14 06:41:45,532 INFO [train.py:467] (3/4) Epoch 568, batch 21, global_batch_idx: 21000, batch size: 73, loss[discriminator_loss=2.559, discriminator_real_loss=1.273, discriminator_fake_loss=1.284, generator_loss=30.11, generator_mel_loss=20.25, generator_kl_loss=1.944, generator_dur_loss=1.65, generator_adv_loss=2.188, generator_feat_match_loss=4.086, over 73.00 samples.], tot_loss[discriminator_loss=2.632, discriminator_real_loss=1.332, discriminator_fake_loss=1.3, generator_loss=30.4, generator_mel_loss=20.73, generator_kl_loss=1.997, generator_dur_loss=1.652, generator_adv_loss=2.13, generator_feat_match_loss=3.889, over 1668.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:41:45,533 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 06:41:56,463 INFO [train.py:517] (3/4) Epoch 568, validation: discriminator_loss=2.506, discriminator_real_loss=1.235, discriminator_fake_loss=1.271, generator_loss=30.96, generator_mel_loss=21.07, generator_kl_loss=2.146, generator_dur_loss=1.636, generator_adv_loss=2.12, generator_feat_match_loss=3.986, over 100.00 samples.
2023-11-14 06:41:56,464 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 06:43:21,026 INFO [train.py:811] (3/4) Start epoch 569
2023-11-14 06:46:45,332 INFO [train.py:467] (3/4) Epoch 569, batch 34, global_batch_idx: 21050, batch size: 76, loss[discriminator_loss=2.643, discriminator_real_loss=1.251, discriminator_fake_loss=1.392, generator_loss=30.3, generator_mel_loss=20.86, generator_kl_loss=1.906, generator_dur_loss=1.674, generator_adv_loss=1.983, generator_feat_match_loss=3.877, over 76.00 samples.], tot_loss[discriminator_loss=2.68, discriminator_real_loss=1.358, discriminator_fake_loss=1.322, generator_loss=30.38, generator_mel_loss=20.74, generator_kl_loss=2.01, generator_dur_loss=1.654, generator_adv_loss=2.147, generator_feat_match_loss=3.832, over 2671.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 06:46:55,541 INFO [train.py:811] (3/4) Start epoch 570
2023-11-14 06:50:27,766 INFO [train.py:811] (3/4) Start epoch 571
2023-11-14 06:51:46,406 INFO [train.py:467] (3/4) Epoch 571, batch 10, global_batch_idx: 21100, batch size: 65, loss[discriminator_loss=2.375, discriminator_real_loss=1.287, discriminator_fake_loss=1.088, generator_loss=31.61, generator_mel_loss=20.56, generator_kl_loss=2.087, generator_dur_loss=1.642, generator_adv_loss=2.535, generator_feat_match_loss=4.785, over 65.00 samples.], tot_loss[discriminator_loss=2.521, discriminator_real_loss=1.289, discriminator_fake_loss=1.232, generator_loss=31.14, generator_mel_loss=20.49, generator_kl_loss=2.009, generator_dur_loss=1.651, generator_adv_loss=2.439, generator_feat_match_loss=4.557, over 850.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 06:53:56,984 INFO [train.py:811] (3/4) Start epoch 572
2023-11-14 06:56:21,886 INFO [train.py:467] (3/4) Epoch 572, batch 23, global_batch_idx: 21150, batch size: 153, loss[discriminator_loss=2.361, discriminator_real_loss=1.236, discriminator_fake_loss=1.125, generator_loss=31.32, generator_mel_loss=20.25, generator_kl_loss=2.047, generator_dur_loss=1.663, generator_adv_loss=2.531, generator_feat_match_loss=4.836, over 153.00 samples.], tot_loss[discriminator_loss=2.499, discriminator_real_loss=1.255, discriminator_fake_loss=1.244, generator_loss=30.9, generator_mel_loss=20.31, generator_kl_loss=1.984, generator_dur_loss=1.651, generator_adv_loss=2.394, generator_feat_match_loss=4.559, over 2086.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 06:57:31,630 INFO [train.py:811] (3/4) Start epoch 573
2023-11-14 07:00:53,126 INFO [train.py:467] (3/4) Epoch 573, batch 36, global_batch_idx: 21200, batch size: 85, loss[discriminator_loss=2.684, discriminator_real_loss=1.34, discriminator_fake_loss=1.344, generator_loss=30.35, generator_mel_loss=20.54, generator_kl_loss=2.117, generator_dur_loss=1.638, generator_adv_loss=2.199, generator_feat_match_loss=3.852, over 85.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.315, discriminator_fake_loss=1.299, generator_loss=30.23, generator_mel_loss=20.52, generator_kl_loss=2, generator_dur_loss=1.655, generator_adv_loss=2.125, generator_feat_match_loss=3.937, over 2519.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:00:53,127 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 07:01:04,806 INFO [train.py:517] (3/4) Epoch 573, validation: discriminator_loss=2.648, discriminator_real_loss=1.389, discriminator_fake_loss=1.26, generator_loss=30.9, generator_mel_loss=21.23, generator_kl_loss=2.15, generator_dur_loss=1.635, generator_adv_loss=2.081, generator_feat_match_loss=3.801, over 100.00 samples.
2023-11-14 07:01:04,807 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 07:01:05,380 INFO [train.py:811] (3/4) Start epoch 574
2023-11-14 07:04:36,757 INFO [train.py:811] (3/4) Start epoch 575
2023-11-14 07:05:59,253 INFO [train.py:467] (3/4) Epoch 575, batch 12, global_batch_idx: 21250, batch size: 126, loss[discriminator_loss=2.494, discriminator_real_loss=1.264, discriminator_fake_loss=1.23, generator_loss=30.55, generator_mel_loss=20.42, generator_kl_loss=1.962, generator_dur_loss=1.637, generator_adv_loss=2.363, generator_feat_match_loss=4.164, over 126.00 samples.], tot_loss[discriminator_loss=2.5, discriminator_real_loss=1.268, discriminator_fake_loss=1.233, generator_loss=31.19, generator_mel_loss=20.73, generator_kl_loss=1.99, generator_dur_loss=1.646, generator_adv_loss=2.382, generator_feat_match_loss=4.443, over 1148.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:08:07,680 INFO [train.py:811] (3/4) Start epoch 576
2023-11-14 07:10:37,066 INFO [train.py:467] (3/4) Epoch 576, batch 25, global_batch_idx: 21300, batch size: 73, loss[discriminator_loss=2.656, discriminator_real_loss=1.392, discriminator_fake_loss=1.266, generator_loss=30.28, generator_mel_loss=20.92, generator_kl_loss=1.946, generator_dur_loss=1.631, generator_adv_loss=1.93, generator_feat_match_loss=3.852, over 73.00 samples.], tot_loss[discriminator_loss=2.602, discriminator_real_loss=1.317, discriminator_fake_loss=1.286, generator_loss=30.31, generator_mel_loss=20.56, generator_kl_loss=2.006, generator_dur_loss=1.651, generator_adv_loss=2.155, generator_feat_match_loss=3.94, over 2022.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:11:43,760 INFO [train.py:811] (3/4) Start epoch 577
2023-11-14 07:15:09,144 INFO [train.py:811] (3/4) Start epoch 578
2023-11-14 07:15:29,908 INFO [train.py:467] (3/4) Epoch 578, batch 1, global_batch_idx: 21350, batch size: 50, loss[discriminator_loss=2.672, discriminator_real_loss=1.384, discriminator_fake_loss=1.287, generator_loss=30.2, generator_mel_loss=20.85, generator_kl_loss=2.01, generator_dur_loss=1.681, generator_adv_loss=2.086, generator_feat_match_loss=3.57, over 50.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.294, discriminator_fake_loss=1.269, generator_loss=30.42, generator_mel_loss=20.72, generator_kl_loss=1.966, generator_dur_loss=1.673, generator_adv_loss=2.232, generator_feat_match_loss=3.828, over 160.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:18:39,427 INFO [train.py:811] (3/4) Start epoch 579
2023-11-14 07:20:15,418 INFO [train.py:467] (3/4) Epoch 579, batch 14, global_batch_idx: 21400, batch size: 79, loss[discriminator_loss=2.531, discriminator_real_loss=1.15, discriminator_fake_loss=1.381, generator_loss=29.91, generator_mel_loss=20.05, generator_kl_loss=1.919, generator_dur_loss=1.648, generator_adv_loss=2.17, generator_feat_match_loss=4.125, over 79.00 samples.], tot_loss[discriminator_loss=2.589, discriminator_real_loss=1.319, discriminator_fake_loss=1.27, generator_loss=30.33, generator_mel_loss=20.44, generator_kl_loss=1.971, generator_dur_loss=1.643, generator_adv_loss=2.195, generator_feat_match_loss=4.087, over 1279.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:20:15,420 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 07:20:26,424 INFO [train.py:517] (3/4) Epoch 579, validation: discriminator_loss=2.57, discriminator_real_loss=1.246, discriminator_fake_loss=1.323, generator_loss=30.95, generator_mel_loss=21.05, generator_kl_loss=2.034, generator_dur_loss=1.638, generator_adv_loss=1.944, generator_feat_match_loss=4.283, over 100.00 samples.
2023-11-14 07:20:26,425 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 07:22:25,631 INFO [train.py:811] (3/4) Start epoch 580
2023-11-14 07:25:00,791 INFO [train.py:467] (3/4) Epoch 580, batch 27, global_batch_idx: 21450, batch size: 90, loss[discriminator_loss=2.553, discriminator_real_loss=1.309, discriminator_fake_loss=1.244, generator_loss=30.53, generator_mel_loss=20.42, generator_kl_loss=1.968, generator_dur_loss=1.652, generator_adv_loss=2.307, generator_feat_match_loss=4.188, over 90.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.297, discriminator_fake_loss=1.282, generator_loss=30.41, generator_mel_loss=20.6, generator_kl_loss=1.998, generator_dur_loss=1.654, generator_adv_loss=2.158, generator_feat_match_loss=4.002, over 2089.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:25:54,011 INFO [train.py:811] (3/4) Start epoch 581
2023-11-14 07:29:22,358 INFO [train.py:811] (3/4) Start epoch 582
2023-11-14 07:29:59,169 INFO [train.py:467] (3/4) Epoch 582, batch 3, global_batch_idx: 21500, batch size: 126, loss[discriminator_loss=2.625, discriminator_real_loss=1.2, discriminator_fake_loss=1.425, generator_loss=30.51, generator_mel_loss=20.74, generator_kl_loss=2.027, generator_dur_loss=1.631, generator_adv_loss=2.051, generator_feat_match_loss=4.066, over 126.00 samples.], tot_loss[discriminator_loss=2.659, discriminator_real_loss=1.343, discriminator_fake_loss=1.316, generator_loss=30.1, generator_mel_loss=20.61, generator_kl_loss=1.987, generator_dur_loss=1.65, generator_adv_loss=2.005, generator_feat_match_loss=3.844, over 332.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:32:55,273 INFO [train.py:811] (3/4) Start epoch 583
2023-11-14 07:34:38,766 INFO [train.py:467] (3/4) Epoch 583, batch 16, global_batch_idx: 21550, batch size: 49, loss[discriminator_loss=2.553, discriminator_real_loss=1.22, discriminator_fake_loss=1.333, generator_loss=30.2, generator_mel_loss=20.35, generator_kl_loss=1.902, generator_dur_loss=1.665, generator_adv_loss=2.189, generator_feat_match_loss=4.09, over 49.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.329, discriminator_fake_loss=1.289, generator_loss=30.27, generator_mel_loss=20.43, generator_kl_loss=1.979, generator_dur_loss=1.65, generator_adv_loss=2.178, generator_feat_match_loss=4.035, over 1274.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:36:28,783 INFO [train.py:811] (3/4) Start epoch 584
2023-11-14 07:39:14,225 INFO [train.py:467] (3/4) Epoch 584, batch 29, global_batch_idx: 21600, batch size: 71, loss[discriminator_loss=2.852, discriminator_real_loss=1.66, discriminator_fake_loss=1.19, generator_loss=30.69, generator_mel_loss=20.76, generator_kl_loss=2.049, generator_dur_loss=1.648, generator_adv_loss=2.365, generator_feat_match_loss=3.877, over 71.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.304, discriminator_fake_loss=1.263, generator_loss=30.58, generator_mel_loss=20.46, generator_kl_loss=1.992, generator_dur_loss=1.654, generator_adv_loss=2.28, generator_feat_match_loss=4.192, over 2085.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:39:14,226 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 07:39:25,425 INFO [train.py:517] (3/4) Epoch 584, validation: discriminator_loss=2.718, discriminator_real_loss=1.526, discriminator_fake_loss=1.192, generator_loss=31.35, generator_mel_loss=21.09, generator_kl_loss=2.166, generator_dur_loss=1.646, generator_adv_loss=2.405, generator_feat_match_loss=4.042, over 100.00 samples.
2023-11-14 07:39:25,426 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 07:40:06,988 INFO [train.py:811] (3/4) Start epoch 585
2023-11-14 07:43:38,447 INFO [train.py:811] (3/4) Start epoch 586
2023-11-14 07:44:22,593 INFO [train.py:467] (3/4) Epoch 586, batch 5, global_batch_idx: 21650, batch size: 52, loss[discriminator_loss=2.572, discriminator_real_loss=1.303, discriminator_fake_loss=1.27, generator_loss=29.93, generator_mel_loss=20.36, generator_kl_loss=2.008, generator_dur_loss=1.667, generator_adv_loss=2.072, generator_feat_match_loss=3.82, over 52.00 samples.], tot_loss[discriminator_loss=2.622, discriminator_real_loss=1.331, discriminator_fake_loss=1.291, generator_loss=30.09, generator_mel_loss=20.49, generator_kl_loss=2.033, generator_dur_loss=1.651, generator_adv_loss=2.063, generator_feat_match_loss=3.847, over 448.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:47:06,598 INFO [train.py:811] (3/4) Start epoch 587
2023-11-14 07:48:54,944 INFO [train.py:467] (3/4) Epoch 587, batch 18, global_batch_idx: 21700, batch size: 65, loss[discriminator_loss=2.684, discriminator_real_loss=1.349, discriminator_fake_loss=1.335, generator_loss=30.55, generator_mel_loss=20.61, generator_kl_loss=1.996, generator_dur_loss=1.656, generator_adv_loss=2.211, generator_feat_match_loss=4.082, over 65.00 samples.], tot_loss[discriminator_loss=2.602, discriminator_real_loss=1.31, discriminator_fake_loss=1.292, generator_loss=30.62, generator_mel_loss=20.65, generator_kl_loss=2.002, generator_dur_loss=1.648, generator_adv_loss=2.202, generator_feat_match_loss=4.115, over 1437.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 07:50:37,655 INFO [train.py:811] (3/4) Start epoch 588
2023-11-14 07:53:36,481 INFO [train.py:467] (3/4) Epoch 588, batch 31, global_batch_idx: 21750, batch size: 53, loss[discriminator_loss=2.637, discriminator_real_loss=1.449, discriminator_fake_loss=1.188, generator_loss=30.42, generator_mel_loss=20.43, generator_kl_loss=2.01, generator_dur_loss=1.641, generator_adv_loss=2.238, generator_feat_match_loss=4.094, over 53.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.291, discriminator_fake_loss=1.273, generator_loss=30.61, generator_mel_loss=20.47, generator_kl_loss=2.003, generator_dur_loss=1.655, generator_adv_loss=2.279, generator_feat_match_loss=4.202, over 2072.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 07:54:08,850 INFO [train.py:811] (3/4) Start epoch 589
2023-11-14 07:57:40,797 INFO [train.py:811] (3/4) Start epoch 590
2023-11-14 07:58:33,754 INFO [train.py:467] (3/4) Epoch 590, batch 7, global_batch_idx: 21800, batch size: 67, loss[discriminator_loss=2.57, discriminator_real_loss=1.074, discriminator_fake_loss=1.497, generator_loss=30.07, generator_mel_loss=20.44, generator_kl_loss=1.948, generator_dur_loss=1.661, generator_adv_loss=2.043, generator_feat_match_loss=3.977, over 67.00 samples.], tot_loss[discriminator_loss=2.563, discriminator_real_loss=1.247, discriminator_fake_loss=1.317, generator_loss=30.71, generator_mel_loss=20.48, generator_kl_loss=1.996, generator_dur_loss=1.651, generator_adv_loss=2.27, generator_feat_match_loss=4.313, over 549.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 07:58:33,755 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 07:58:45,634 INFO [train.py:517] (3/4) Epoch 590, validation: discriminator_loss=2.695, discriminator_real_loss=1.171, discriminator_fake_loss=1.524, generator_loss=29.96, generator_mel_loss=20.78, generator_kl_loss=2.099, generator_dur_loss=1.633, generator_adv_loss=1.747, generator_feat_match_loss=3.7, over 100.00 samples.
2023-11-14 07:58:45,636 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 08:01:21,372 INFO [train.py:811] (3/4) Start epoch 591
2023-11-14 08:03:28,883 INFO [train.py:467] (3/4) Epoch 591, batch 20, global_batch_idx: 21850, batch size: 79, loss[discriminator_loss=2.385, discriminator_real_loss=1.2, discriminator_fake_loss=1.185, generator_loss=31.35, generator_mel_loss=20.1, generator_kl_loss=2.005, generator_dur_loss=1.685, generator_adv_loss=2.592, generator_feat_match_loss=4.961, over 79.00 samples.], tot_loss[discriminator_loss=2.558, discriminator_real_loss=1.303, discriminator_fake_loss=1.255, generator_loss=30.53, generator_mel_loss=20.3, generator_kl_loss=2.018, generator_dur_loss=1.654, generator_adv_loss=2.299, generator_feat_match_loss=4.263, over 1467.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 08:04:53,902 INFO [train.py:811] (3/4) Start epoch 592
2023-11-14 08:08:10,963 INFO [train.py:467] (3/4) Epoch 592, batch 33, global_batch_idx: 21900, batch size: 153, loss[discriminator_loss=2.535, discriminator_real_loss=1.346, discriminator_fake_loss=1.188, generator_loss=31.06, generator_mel_loss=20.77, generator_kl_loss=2.129, generator_dur_loss=1.648, generator_adv_loss=2.225, generator_feat_match_loss=4.293, over 153.00 samples.], tot_loss[discriminator_loss=2.599, discriminator_real_loss=1.312, discriminator_fake_loss=1.286, generator_loss=30.37, generator_mel_loss=20.37, generator_kl_loss=2.001, generator_dur_loss=1.653, generator_adv_loss=2.22, generator_feat_match_loss=4.129, over 2633.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 08:08:27,595 INFO [train.py:811] (3/4) Start epoch 593
2023-11-14 08:12:00,236 INFO [train.py:811] (3/4) Start epoch 594
2023-11-14 08:13:06,922 INFO [train.py:467] (3/4) Epoch 594, batch 9, global_batch_idx: 21950, batch size: 49, loss[discriminator_loss=2.625, discriminator_real_loss=1.171, discriminator_fake_loss=1.453, generator_loss=29.88, generator_mel_loss=20.28, generator_kl_loss=1.965, generator_dur_loss=1.67, generator_adv_loss=2.289, generator_feat_match_loss=3.674, over 49.00 samples.], tot_loss[discriminator_loss=2.4, discriminator_real_loss=1.242, discriminator_fake_loss=1.157, generator_loss=31.35, generator_mel_loss=20.34, generator_kl_loss=1.988, generator_dur_loss=1.657, generator_adv_loss=2.474, generator_feat_match_loss=4.889, over 720.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 8.0
2023-11-14 08:15:28,667 INFO [train.py:811] (3/4) Start epoch 595
2023-11-14 08:17:43,019 INFO [train.py:467] (3/4) Epoch 595, batch 22, global_batch_idx: 22000, batch size: 52, loss[discriminator_loss=2.664, discriminator_real_loss=1.315, discriminator_fake_loss=1.348, generator_loss=29.65, generator_mel_loss=20.31, generator_kl_loss=1.977, generator_dur_loss=1.669, generator_adv_loss=1.986, generator_feat_match_loss=3.705, over 52.00 samples.], tot_loss[discriminator_loss=2.647, discriminator_real_loss=1.341, discriminator_fake_loss=1.306, generator_loss=30.18, generator_mel_loss=20.56, generator_kl_loss=2.002, generator_dur_loss=1.647, generator_adv_loss=2.088, generator_feat_match_loss=3.881, over 1723.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:17:43,021 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 08:17:54,307 INFO [train.py:517] (3/4) Epoch 595, validation: discriminator_loss=2.616, discriminator_real_loss=1.26, discriminator_fake_loss=1.356, generator_loss=30.43, generator_mel_loss=21.16, generator_kl_loss=2.069, generator_dur_loss=1.636, generator_adv_loss=1.879, generator_feat_match_loss=3.686, over 100.00 samples.
2023-11-14 08:17:54,308 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 08:19:07,796 INFO [train.py:811] (3/4) Start epoch 596
2023-11-14 08:22:33,014 INFO [train.py:467] (3/4) Epoch 596, batch 35, global_batch_idx: 22050, batch size: 126, loss[discriminator_loss=2.711, discriminator_real_loss=1.352, discriminator_fake_loss=1.359, generator_loss=30.52, generator_mel_loss=20.98, generator_kl_loss=2.038, generator_dur_loss=1.637, generator_adv_loss=2.002, generator_feat_match_loss=3.857, over 126.00 samples.], tot_loss[discriminator_loss=2.632, discriminator_real_loss=1.342, discriminator_fake_loss=1.291, generator_loss=30.27, generator_mel_loss=20.55, generator_kl_loss=2.03, generator_dur_loss=1.647, generator_adv_loss=2.145, generator_feat_match_loss=3.906, over 2735.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:22:38,268 INFO [train.py:811] (3/4) Start epoch 597
2023-11-14 08:26:09,071 INFO [train.py:811] (3/4) Start epoch 598
2023-11-14 08:27:25,790 INFO [train.py:467] (3/4) Epoch 598, batch 11, global_batch_idx: 22100, batch size: 64, loss[discriminator_loss=2.719, discriminator_real_loss=1.404, discriminator_fake_loss=1.314, generator_loss=29.47, generator_mel_loss=20.39, generator_kl_loss=1.942, generator_dur_loss=1.651, generator_adv_loss=1.874, generator_feat_match_loss=3.611, over 64.00 samples.], tot_loss[discriminator_loss=2.649, discriminator_real_loss=1.339, discriminator_fake_loss=1.311, generator_loss=30.35, generator_mel_loss=20.67, generator_kl_loss=1.98, generator_dur_loss=1.645, generator_adv_loss=2.121, generator_feat_match_loss=3.933, over 907.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:29:42,226 INFO [train.py:811] (3/4) Start epoch 599
2023-11-14 08:32:02,702 INFO [train.py:467] (3/4) Epoch 599, batch 24, global_batch_idx: 22150, batch size: 55, loss[discriminator_loss=2.684, discriminator_real_loss=1.547, discriminator_fake_loss=1.137, generator_loss=30.68, generator_mel_loss=20.47, generator_kl_loss=2.005, generator_dur_loss=1.676, generator_adv_loss=2.381, generator_feat_match_loss=4.148, over 55.00 samples.], tot_loss[discriminator_loss=2.609, discriminator_real_loss=1.33, discriminator_fake_loss=1.279, generator_loss=30.3, generator_mel_loss=20.42, generator_kl_loss=1.967, generator_dur_loss=1.647, generator_adv_loss=2.231, generator_feat_match_loss=4.036, over 1670.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:33:16,249 INFO [train.py:811] (3/4) Start epoch 600
2023-11-14 08:36:52,393 INFO [train.py:811] (3/4) Start epoch 601
2023-11-14 08:37:08,992 INFO [train.py:467] (3/4) Epoch 601, batch 0, global_batch_idx: 22200, batch size: 65, loss[discriminator_loss=2.586, discriminator_real_loss=1.375, discriminator_fake_loss=1.211, generator_loss=29.58, generator_mel_loss=20.08, generator_kl_loss=2.001, generator_dur_loss=1.661, generator_adv_loss=2.076, generator_feat_match_loss=3.766, over 65.00 samples.], tot_loss[discriminator_loss=2.586, discriminator_real_loss=1.375, discriminator_fake_loss=1.211, generator_loss=29.58, generator_mel_loss=20.08, generator_kl_loss=2.001, generator_dur_loss=1.661, generator_adv_loss=2.076, generator_feat_match_loss=3.766, over 65.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:37:08,994 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 08:37:20,804 INFO [train.py:517] (3/4) Epoch 601, validation: discriminator_loss=2.585, discriminator_real_loss=1.242, discriminator_fake_loss=1.343, generator_loss=30.81, generator_mel_loss=20.9, generator_kl_loss=2.05, generator_dur_loss=1.638, generator_adv_loss=2.001, generator_feat_match_loss=4.224, over 100.00 samples.
2023-11-14 08:37:20,806 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 08:40:40,200 INFO [train.py:811] (3/4) Start epoch 602
2023-11-14 08:42:07,695 INFO [train.py:467] (3/4) Epoch 602, batch 13, global_batch_idx: 22250, batch size: 63, loss[discriminator_loss=2.65, discriminator_real_loss=1.515, discriminator_fake_loss=1.136, generator_loss=30.71, generator_mel_loss=20.72, generator_kl_loss=2.08, generator_dur_loss=1.643, generator_adv_loss=2.229, generator_feat_match_loss=4.035, over 63.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.328, discriminator_fake_loss=1.288, generator_loss=30.74, generator_mel_loss=20.82, generator_kl_loss=2.023, generator_dur_loss=1.656, generator_adv_loss=2.183, generator_feat_match_loss=4.053, over 955.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:44:13,033 INFO [train.py:811] (3/4) Start epoch 603
2023-11-14 08:46:49,396 INFO [train.py:467] (3/4) Epoch 603, batch 26, global_batch_idx: 22300, batch size: 71, loss[discriminator_loss=2.562, discriminator_real_loss=1.249, discriminator_fake_loss=1.314, generator_loss=30.46, generator_mel_loss=20.64, generator_kl_loss=2.041, generator_dur_loss=1.627, generator_adv_loss=2.031, generator_feat_match_loss=4.121, over 71.00 samples.], tot_loss[discriminator_loss=2.611, discriminator_real_loss=1.319, discriminator_fake_loss=1.292, generator_loss=30.58, generator_mel_loss=20.59, generator_kl_loss=1.988, generator_dur_loss=1.648, generator_adv_loss=2.241, generator_feat_match_loss=4.118, over 1959.00 samples.], cur_lr_g: 1.86e-04, cur_lr_d: 1.86e-04, grad_scale: 16.0
2023-11-14 08:47:48,142 INFO [train.py:811] (3/4) Start epoch 604
2023-11-14 08:51:19,148 INFO [train.py:811] (3/4) Start epoch 605
2023-11-14 08:51:43,951 INFO [train.py:467] (3/4) Epoch 605, batch 2, global_batch_idx: 22350, batch size: 54, loss[discriminator_loss=2.641, discriminator_real_loss=1.457, discriminator_fake_loss=1.183, generator_loss=29.9, generator_mel_loss=20.22, generator_kl_loss=2.032, generator_dur_loss=1.654, generator_adv_loss=2.121, generator_feat_match_loss=3.875, over 54.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.382, discriminator_fake_loss=1.243, generator_loss=30.1, generator_mel_loss=20.41, generator_kl_loss=2.026, generator_dur_loss=1.647, generator_adv_loss=2.158, generator_feat_match_loss=3.859, over 179.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 08:54:52,226 INFO [train.py:811] (3/4) Start epoch 606
2023-11-14 08:56:29,679 INFO [train.py:467] (3/4) Epoch 606, batch 15, global_batch_idx: 22400, batch size: 49, loss[discriminator_loss=2.613, discriminator_real_loss=1.348, discriminator_fake_loss=1.267, generator_loss=30.22, generator_mel_loss=20.59, generator_kl_loss=1.954, generator_dur_loss=1.657, generator_adv_loss=2.113, generator_feat_match_loss=3.91, over 49.00 samples.], tot_loss[discriminator_loss=2.656, discriminator_real_loss=1.348, discriminator_fake_loss=1.308, generator_loss=30.31, generator_mel_loss=20.59, generator_kl_loss=2.016, generator_dur_loss=1.651, generator_adv_loss=2.12, generator_feat_match_loss=3.929, over 1152.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 08:56:29,680 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 08:56:40,712 INFO [train.py:517] (3/4) Epoch 606, validation: discriminator_loss=2.547, discriminator_real_loss=1.237, discriminator_fake_loss=1.31, generator_loss=30.86, generator_mel_loss=21.06, generator_kl_loss=2.2, generator_dur_loss=1.637, generator_adv_loss=2.034, generator_feat_match_loss=3.924, over 100.00 samples.
2023-11-14 08:56:40,713 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 08:58:38,551 INFO [train.py:811] (3/4) Start epoch 607
2023-11-14 09:01:30,563 INFO [train.py:467] (3/4) Epoch 607, batch 28, global_batch_idx: 22450, batch size: 51, loss[discriminator_loss=2.582, discriminator_real_loss=1.338, discriminator_fake_loss=1.243, generator_loss=30.93, generator_mel_loss=20.85, generator_kl_loss=1.996, generator_dur_loss=1.666, generator_adv_loss=2.27, generator_feat_match_loss=4.148, over 51.00 samples.], tot_loss[discriminator_loss=2.635, discriminator_real_loss=1.33, discriminator_fake_loss=1.305, generator_loss=30.42, generator_mel_loss=20.64, generator_kl_loss=1.993, generator_dur_loss=1.644, generator_adv_loss=2.142, generator_feat_match_loss=4.002, over 2240.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 09:02:07,428 INFO [train.py:811] (3/4) Start epoch 608
2023-11-14 09:05:37,462 INFO [train.py:811] (3/4) Start epoch 609
2023-11-14 09:06:13,802 INFO [train.py:467] (3/4) Epoch 609, batch 4, global_batch_idx: 22500, batch size: 58, loss[discriminator_loss=2.645, discriminator_real_loss=1.362, discriminator_fake_loss=1.282, generator_loss=30.05, generator_mel_loss=20.42, generator_kl_loss=1.925, generator_dur_loss=1.637, generator_adv_loss=2.213, generator_feat_match_loss=3.854, over 58.00 samples.], tot_loss[discriminator_loss=2.665, discriminator_real_loss=1.336, discriminator_fake_loss=1.329, generator_loss=30.34, generator_mel_loss=20.72, generator_kl_loss=1.999, generator_dur_loss=1.65, generator_adv_loss=2.103, generator_feat_match_loss=3.864, over 343.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 09:09:12,622 INFO [train.py:811] (3/4) Start epoch 610
2023-11-14 09:10:59,122 INFO [train.py:467] (3/4) Epoch 610, batch 17, global_batch_idx: 22550, batch size: 55, loss[discriminator_loss=2.508, discriminator_real_loss=1.261, discriminator_fake_loss=1.247, generator_loss=31.79, generator_mel_loss=20.92, generator_kl_loss=1.997, generator_dur_loss=1.68, generator_adv_loss=2.377, generator_feat_match_loss=4.82, over 55.00 samples.], tot_loss[discriminator_loss=2.541, discriminator_real_loss=1.279, discriminator_fake_loss=1.262, generator_loss=30.54, generator_mel_loss=20.4, generator_kl_loss=2.001, generator_dur_loss=1.643, generator_adv_loss=2.29, generator_feat_match_loss=4.205, over 1345.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:12:50,112 INFO [train.py:811] (3/4) Start epoch 611
2023-11-14 09:15:56,616 INFO [train.py:467] (3/4) Epoch 611, batch 30, global_batch_idx: 22600, batch size: 63, loss[discriminator_loss=2.535, discriminator_real_loss=1.325, discriminator_fake_loss=1.209, generator_loss=29.75, generator_mel_loss=20.01, generator_kl_loss=2.032, generator_dur_loss=1.654, generator_adv_loss=2.16, generator_feat_match_loss=3.889, over 63.00 samples.], tot_loss[discriminator_loss=2.551, discriminator_real_loss=1.287, discriminator_fake_loss=1.265, generator_loss=30.3, generator_mel_loss=20.24, generator_kl_loss=1.979, generator_dur_loss=1.655, generator_adv_loss=2.229, generator_feat_match_loss=4.204, over 2354.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:15:56,617 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 09:16:07,666 INFO [train.py:517] (3/4) Epoch 611, validation: discriminator_loss=2.545, discriminator_real_loss=1.245, discriminator_fake_loss=1.3, generator_loss=30.97, generator_mel_loss=20.94, generator_kl_loss=2.134, generator_dur_loss=1.63, generator_adv_loss=2.076, generator_feat_match_loss=4.18, over 100.00 samples.
2023-11-14 09:16:07,667 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 09:16:34,326 INFO [train.py:811] (3/4) Start epoch 612
2023-11-14 09:20:06,578 INFO [train.py:811] (3/4) Start epoch 613
2023-11-14 09:20:51,682 INFO [train.py:467] (3/4) Epoch 613, batch 6, global_batch_idx: 22650, batch size: 52, loss[discriminator_loss=2.582, discriminator_real_loss=1.3, discriminator_fake_loss=1.283, generator_loss=30.01, generator_mel_loss=19.99, generator_kl_loss=1.996, generator_dur_loss=1.664, generator_adv_loss=2.268, generator_feat_match_loss=4.09, over 52.00 samples.], tot_loss[discriminator_loss=2.518, discriminator_real_loss=1.282, discriminator_fake_loss=1.237, generator_loss=30.9, generator_mel_loss=20.35, generator_kl_loss=2.02, generator_dur_loss=1.656, generator_adv_loss=2.403, generator_feat_match_loss=4.476, over 463.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:23:36,792 INFO [train.py:811] (3/4) Start epoch 614
2023-11-14 09:25:41,065 INFO [train.py:467] (3/4) Epoch 614, batch 19, global_batch_idx: 22700, batch size: 51, loss[discriminator_loss=2.574, discriminator_real_loss=1.392, discriminator_fake_loss=1.183, generator_loss=29.82, generator_mel_loss=19.94, generator_kl_loss=2.042, generator_dur_loss=1.65, generator_adv_loss=2.312, generator_feat_match_loss=3.873, over 51.00 samples.], tot_loss[discriminator_loss=2.512, discriminator_real_loss=1.275, discriminator_fake_loss=1.236, generator_loss=30.32, generator_mel_loss=20.03, generator_kl_loss=1.996, generator_dur_loss=1.654, generator_adv_loss=2.315, generator_feat_match_loss=4.325, over 1197.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:27:09,706 INFO [train.py:811] (3/4) Start epoch 615
2023-11-14 09:30:16,622 INFO [train.py:467] (3/4) Epoch 615, batch 32, global_batch_idx: 22750, batch size: 126, loss[discriminator_loss=2.75, discriminator_real_loss=1.243, discriminator_fake_loss=1.508, generator_loss=30.96, generator_mel_loss=20.76, generator_kl_loss=2.028, generator_dur_loss=1.651, generator_adv_loss=2.404, generator_feat_match_loss=4.121, over 126.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.311, discriminator_fake_loss=1.289, generator_loss=30.63, generator_mel_loss=20.58, generator_kl_loss=2.025, generator_dur_loss=1.653, generator_adv_loss=2.212, generator_feat_match_loss=4.161, over 2370.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:30:41,004 INFO [train.py:811] (3/4) Start epoch 616
2023-11-14 09:34:12,831 INFO [train.py:811] (3/4) Start epoch 617
2023-11-14 09:35:08,957 INFO [train.py:467] (3/4) Epoch 617, batch 8, global_batch_idx: 22800, batch size: 52, loss[discriminator_loss=2.645, discriminator_real_loss=1.322, discriminator_fake_loss=1.322, generator_loss=29.75, generator_mel_loss=20.32, generator_kl_loss=2.016, generator_dur_loss=1.642, generator_adv_loss=2.045, generator_feat_match_loss=3.734, over 52.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.33, discriminator_fake_loss=1.283, generator_loss=30.12, generator_mel_loss=20.37, generator_kl_loss=1.985, generator_dur_loss=1.645, generator_adv_loss=2.141, generator_feat_match_loss=3.981, over 630.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 09:35:08,958 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 09:35:20,814 INFO [train.py:517] (3/4) Epoch 617, validation: discriminator_loss=2.624, discriminator_real_loss=1.329, discriminator_fake_loss=1.295, generator_loss=31.14, generator_mel_loss=21.46, generator_kl_loss=2.126, generator_dur_loss=1.642, generator_adv_loss=2.023, generator_feat_match_loss=3.887, over 100.00 samples.
2023-11-14 09:35:20,815 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 09:38:01,845 INFO [train.py:811] (3/4) Start epoch 618
2023-11-14 09:40:09,688 INFO [train.py:467] (3/4) Epoch 618, batch 21, global_batch_idx: 22850, batch size: 52, loss[discriminator_loss=2.588, discriminator_real_loss=1.33, discriminator_fake_loss=1.258, generator_loss=29.59, generator_mel_loss=20, generator_kl_loss=1.943, generator_dur_loss=1.64, generator_adv_loss=2.273, generator_feat_match_loss=3.738, over 52.00 samples.], tot_loss[discriminator_loss=2.634, discriminator_real_loss=1.323, discriminator_fake_loss=1.311, generator_loss=30.5, generator_mel_loss=20.67, generator_kl_loss=2.032, generator_dur_loss=1.649, generator_adv_loss=2.171, generator_feat_match_loss=3.978, over 1717.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 09:41:29,568 INFO [train.py:811] (3/4) Start epoch 619
2023-11-14 09:44:53,758 INFO [train.py:467] (3/4) Epoch 619, batch 34, global_batch_idx: 22900, batch size: 63, loss[discriminator_loss=2.465, discriminator_real_loss=1.307, discriminator_fake_loss=1.157, generator_loss=31.21, generator_mel_loss=20.35, generator_kl_loss=1.948, generator_dur_loss=1.681, generator_adv_loss=2.281, generator_feat_match_loss=4.953, over 63.00 samples.], tot_loss[discriminator_loss=2.52, discriminator_real_loss=1.285, discriminator_fake_loss=1.236, generator_loss=30.86, generator_mel_loss=20.26, generator_kl_loss=1.981, generator_dur_loss=1.65, generator_adv_loss=2.375, generator_feat_match_loss=4.594, over 2431.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:45:03,845 INFO [train.py:811] (3/4) Start epoch 620
2023-11-14 09:48:40,626 INFO [train.py:811] (3/4) Start epoch 621
2023-11-14 09:49:58,331 INFO [train.py:467] (3/4) Epoch 621, batch 10, global_batch_idx: 22950, batch size: 73, loss[discriminator_loss=2.578, discriminator_real_loss=1.182, discriminator_fake_loss=1.396, generator_loss=30.48, generator_mel_loss=20.51, generator_kl_loss=2.049, generator_dur_loss=1.657, generator_adv_loss=2.215, generator_feat_match_loss=4.051, over 73.00 samples.], tot_loss[discriminator_loss=2.568, discriminator_real_loss=1.281, discriminator_fake_loss=1.287, generator_loss=30.24, generator_mel_loss=20.44, generator_kl_loss=2.002, generator_dur_loss=1.648, generator_adv_loss=2.135, generator_feat_match_loss=4.014, over 824.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:52:13,959 INFO [train.py:811] (3/4) Start epoch 622
2023-11-14 09:54:25,811 INFO [train.py:467] (3/4) Epoch 622, batch 23, global_batch_idx: 23000, batch size: 95, loss[discriminator_loss=2.672, discriminator_real_loss=1.375, discriminator_fake_loss=1.297, generator_loss=30.83, generator_mel_loss=20.84, generator_kl_loss=2.034, generator_dur_loss=1.654, generator_adv_loss=2.215, generator_feat_match_loss=4.086, over 95.00 samples.], tot_loss[discriminator_loss=2.643, discriminator_real_loss=1.348, discriminator_fake_loss=1.295, generator_loss=30.25, generator_mel_loss=20.56, generator_kl_loss=1.984, generator_dur_loss=1.653, generator_adv_loss=2.131, generator_feat_match_loss=3.926, over 1715.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:54:25,812 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 09:54:37,385 INFO [train.py:517] (3/4) Epoch 622, validation: discriminator_loss=2.656, discriminator_real_loss=1.36, discriminator_fake_loss=1.297, generator_loss=31.1, generator_mel_loss=21.42, generator_kl_loss=2.152, generator_dur_loss=1.639, generator_adv_loss=1.983, generator_feat_match_loss=3.901, over 100.00 samples.
2023-11-14 09:54:37,386 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 09:55:57,102 INFO [train.py:811] (3/4) Start epoch 623
2023-11-14 09:59:26,703 INFO [train.py:467] (3/4) Epoch 623, batch 36, global_batch_idx: 23050, batch size: 51, loss[discriminator_loss=2.527, discriminator_real_loss=1.228, discriminator_fake_loss=1.301, generator_loss=30.7, generator_mel_loss=20.71, generator_kl_loss=1.961, generator_dur_loss=1.651, generator_adv_loss=2.393, generator_feat_match_loss=3.982, over 51.00 samples.], tot_loss[discriminator_loss=2.657, discriminator_real_loss=1.347, discriminator_fake_loss=1.31, generator_loss=30.36, generator_mel_loss=20.64, generator_kl_loss=1.992, generator_dur_loss=1.649, generator_adv_loss=2.133, generator_feat_match_loss=3.943, over 2702.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 09:59:27,367 INFO [train.py:811] (3/4) Start epoch 624
2023-11-14 10:02:54,683 INFO [train.py:811] (3/4) Start epoch 625
2023-11-14 10:04:14,216 INFO [train.py:467] (3/4) Epoch 625, batch 12, global_batch_idx: 23100, batch size: 154, loss[discriminator_loss=2.65, discriminator_real_loss=1.259, discriminator_fake_loss=1.392, generator_loss=31.37, generator_mel_loss=20.96, generator_kl_loss=1.986, generator_dur_loss=1.635, generator_adv_loss=2.369, generator_feat_match_loss=4.414, over 154.00 samples.], tot_loss[discriminator_loss=2.599, discriminator_real_loss=1.303, discriminator_fake_loss=1.296, generator_loss=31.01, generator_mel_loss=20.7, generator_kl_loss=2.002, generator_dur_loss=1.647, generator_adv_loss=2.284, generator_feat_match_loss=4.369, over 1006.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:06:27,564 INFO [train.py:811] (3/4) Start epoch 626
2023-11-14 10:08:53,786 INFO [train.py:467] (3/4) Epoch 626, batch 25, global_batch_idx: 23150, batch size: 56, loss[discriminator_loss=2.578, discriminator_real_loss=1.308, discriminator_fake_loss=1.271, generator_loss=29.83, generator_mel_loss=20.23, generator_kl_loss=1.904, generator_dur_loss=1.642, generator_adv_loss=2.281, generator_feat_match_loss=3.777, over 56.00 samples.], tot_loss[discriminator_loss=2.56, discriminator_real_loss=1.295, discriminator_fake_loss=1.264, generator_loss=30.32, generator_mel_loss=20.3, generator_kl_loss=1.985, generator_dur_loss=1.65, generator_adv_loss=2.218, generator_feat_match_loss=4.163, over 1809.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:09:57,212 INFO [train.py:811] (3/4) Start epoch 627
2023-11-14 10:13:34,912 INFO [train.py:811] (3/4) Start epoch 628
2023-11-14 10:13:57,603 INFO [train.py:467] (3/4) Epoch 628, batch 1, global_batch_idx: 23200, batch size: 153, loss[discriminator_loss=2.852, discriminator_real_loss=1.403, discriminator_fake_loss=1.449, generator_loss=29.56, generator_mel_loss=20.18, generator_kl_loss=2.084, generator_dur_loss=1.649, generator_adv_loss=1.979, generator_feat_match_loss=3.662, over 153.00 samples.], tot_loss[discriminator_loss=2.809, discriminator_real_loss=1.363, discriminator_fake_loss=1.446, generator_loss=29.65, generator_mel_loss=20.23, generator_kl_loss=2.045, generator_dur_loss=1.656, generator_adv_loss=2.009, generator_feat_match_loss=3.707, over 213.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 10:13:57,605 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 10:14:09,967 INFO [train.py:517] (3/4) Epoch 628, validation: discriminator_loss=2.78, discriminator_real_loss=1.448, discriminator_fake_loss=1.331, generator_loss=30.2, generator_mel_loss=20.83, generator_kl_loss=2.13, generator_dur_loss=1.641, generator_adv_loss=1.933, generator_feat_match_loss=3.66, over 100.00 samples.
2023-11-14 10:14:09,968 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 10:17:15,495 INFO [train.py:811] (3/4) Start epoch 629
2023-11-14 10:18:46,044 INFO [train.py:467] (3/4) Epoch 629, batch 14, global_batch_idx: 23250, batch size: 73, loss[discriminator_loss=2.523, discriminator_real_loss=1.207, discriminator_fake_loss=1.317, generator_loss=31.32, generator_mel_loss=20.88, generator_kl_loss=2.019, generator_dur_loss=1.65, generator_adv_loss=2.387, generator_feat_match_loss=4.383, over 73.00 samples.], tot_loss[discriminator_loss=2.565, discriminator_real_loss=1.302, discriminator_fake_loss=1.263, generator_loss=30.42, generator_mel_loss=20.45, generator_kl_loss=2.001, generator_dur_loss=1.639, generator_adv_loss=2.19, generator_feat_match_loss=4.137, over 1247.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 10:20:46,964 INFO [train.py:811] (3/4) Start epoch 630
2023-11-14 10:23:35,024 INFO [train.py:467] (3/4) Epoch 630, batch 27, global_batch_idx: 23300, batch size: 126, loss[discriminator_loss=2.648, discriminator_real_loss=1.348, discriminator_fake_loss=1.302, generator_loss=30.68, generator_mel_loss=20.62, generator_kl_loss=1.981, generator_dur_loss=1.657, generator_adv_loss=2.299, generator_feat_match_loss=4.125, over 126.00 samples.], tot_loss[discriminator_loss=2.592, discriminator_real_loss=1.317, discriminator_fake_loss=1.275, generator_loss=30.14, generator_mel_loss=20.22, generator_kl_loss=1.977, generator_dur_loss=1.652, generator_adv_loss=2.203, generator_feat_match_loss=4.089, over 2158.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 10:24:19,964 INFO [train.py:811] (3/4) Start epoch 631
2023-11-14 10:27:52,468 INFO [train.py:811] (3/4) Start epoch 632
2023-11-14 10:28:21,123 INFO [train.py:467] (3/4) Epoch 632, batch 3, global_batch_idx: 23350, batch size: 50, loss[discriminator_loss=2.434, discriminator_real_loss=1.328, discriminator_fake_loss=1.106, generator_loss=30.92, generator_mel_loss=20.05, generator_kl_loss=1.921, generator_dur_loss=1.672, generator_adv_loss=2.445, generator_feat_match_loss=4.836, over 50.00 samples.], tot_loss[discriminator_loss=2.573, discriminator_real_loss=1.379, discriminator_fake_loss=1.194, generator_loss=30.26, generator_mel_loss=19.96, generator_kl_loss=2.021, generator_dur_loss=1.645, generator_adv_loss=2.432, generator_feat_match_loss=4.21, over 281.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:31:22,519 INFO [train.py:811] (3/4) Start epoch 633
2023-11-14 10:33:10,318 INFO [train.py:467] (3/4) Epoch 633, batch 16, global_batch_idx: 23400, batch size: 65, loss[discriminator_loss=2.598, discriminator_real_loss=1.318, discriminator_fake_loss=1.278, generator_loss=30, generator_mel_loss=20.35, generator_kl_loss=1.948, generator_dur_loss=1.659, generator_adv_loss=2.143, generator_feat_match_loss=3.902, over 65.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.309, discriminator_fake_loss=1.273, generator_loss=30.3, generator_mel_loss=20.4, generator_kl_loss=2.024, generator_dur_loss=1.648, generator_adv_loss=2.146, generator_feat_match_loss=4.084, over 1292.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:33:10,319 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 10:33:21,999 INFO [train.py:517] (3/4) Epoch 633, validation: discriminator_loss=2.527, discriminator_real_loss=1.207, discriminator_fake_loss=1.32, generator_loss=30.54, generator_mel_loss=20.69, generator_kl_loss=2.167, generator_dur_loss=1.634, generator_adv_loss=2.041, generator_feat_match_loss=4.006, over 100.00 samples.
2023-11-14 10:33:22,000 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 10:35:03,935 INFO [train.py:811] (3/4) Start epoch 634
2023-11-14 10:37:54,479 INFO [train.py:467] (3/4) Epoch 634, batch 29, global_batch_idx: 23450, batch size: 90, loss[discriminator_loss=2.527, discriminator_real_loss=1.304, discriminator_fake_loss=1.225, generator_loss=30.68, generator_mel_loss=20.37, generator_kl_loss=2.035, generator_dur_loss=1.646, generator_adv_loss=2.225, generator_feat_match_loss=4.41, over 90.00 samples.], tot_loss[discriminator_loss=2.539, discriminator_real_loss=1.266, discriminator_fake_loss=1.273, generator_loss=30.79, generator_mel_loss=20.41, generator_kl_loss=1.992, generator_dur_loss=1.649, generator_adv_loss=2.296, generator_feat_match_loss=4.438, over 2165.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:38:37,577 INFO [train.py:811] (3/4) Start epoch 635
2023-11-14 10:42:18,998 INFO [train.py:811] (3/4) Start epoch 636
2023-11-14 10:42:58,835 INFO [train.py:467] (3/4) Epoch 636, batch 5, global_batch_idx: 23500, batch size: 67, loss[discriminator_loss=2.373, discriminator_real_loss=1.16, discriminator_fake_loss=1.213, generator_loss=30.95, generator_mel_loss=19.57, generator_kl_loss=1.855, generator_dur_loss=1.655, generator_adv_loss=2.723, generator_feat_match_loss=5.148, over 67.00 samples.], tot_loss[discriminator_loss=2.367, discriminator_real_loss=1.222, discriminator_fake_loss=1.144, generator_loss=30.82, generator_mel_loss=19.92, generator_kl_loss=1.947, generator_dur_loss=1.649, generator_adv_loss=2.394, generator_feat_match_loss=4.912, over 427.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:45:54,909 INFO [train.py:811] (3/4) Start epoch 637
2023-11-14 10:47:54,949 INFO [train.py:467] (3/4) Epoch 637, batch 18, global_batch_idx: 23550, batch size: 81, loss[discriminator_loss=2.623, discriminator_real_loss=1.175, discriminator_fake_loss=1.448, generator_loss=29.6, generator_mel_loss=20.14, generator_kl_loss=2.012, generator_dur_loss=1.647, generator_adv_loss=2.049, generator_feat_match_loss=3.754, over 81.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.329, discriminator_fake_loss=1.294, generator_loss=30.36, generator_mel_loss=20.48, generator_kl_loss=2.025, generator_dur_loss=1.646, generator_adv_loss=2.136, generator_feat_match_loss=4.078, over 1533.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 10:49:25,719 INFO [train.py:811] (3/4) Start epoch 638
2023-11-14 10:52:32,053 INFO [train.py:467] (3/4) Epoch 638, batch 31, global_batch_idx: 23600, batch size: 67, loss[discriminator_loss=2.707, discriminator_real_loss=1.295, discriminator_fake_loss=1.413, generator_loss=30.57, generator_mel_loss=20.59, generator_kl_loss=2.06, generator_dur_loss=1.672, generator_adv_loss=2.09, generator_feat_match_loss=4.16, over 67.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.32, discriminator_fake_loss=1.308, generator_loss=30.4, generator_mel_loss=20.54, generator_kl_loss=2.052, generator_dur_loss=1.653, generator_adv_loss=2.128, generator_feat_match_loss=4.025, over 2261.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 10:52:32,055 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 10:52:43,775 INFO [train.py:517] (3/4) Epoch 638, validation: discriminator_loss=2.672, discriminator_real_loss=1.26, discriminator_fake_loss=1.412, generator_loss=30.22, generator_mel_loss=20.97, generator_kl_loss=2.202, generator_dur_loss=1.635, generator_adv_loss=1.772, generator_feat_match_loss=3.642, over 100.00 samples.
2023-11-14 10:52:43,776 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 10:53:10,205 INFO [train.py:811] (3/4) Start epoch 639
2023-11-14 10:56:44,352 INFO [train.py:811] (3/4) Start epoch 640
2023-11-14 10:57:40,747 INFO [train.py:467] (3/4) Epoch 640, batch 7, global_batch_idx: 23650, batch size: 64, loss[discriminator_loss=2.543, discriminator_real_loss=1.201, discriminator_fake_loss=1.342, generator_loss=30.58, generator_mel_loss=20.29, generator_kl_loss=1.981, generator_dur_loss=1.651, generator_adv_loss=2.182, generator_feat_match_loss=4.477, over 64.00 samples.], tot_loss[discriminator_loss=2.622, discriminator_real_loss=1.321, discriminator_fake_loss=1.3, generator_loss=30.32, generator_mel_loss=20.3, generator_kl_loss=1.971, generator_dur_loss=1.65, generator_adv_loss=2.194, generator_feat_match_loss=4.207, over 606.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 11:00:13,311 INFO [train.py:811] (3/4) Start epoch 641
2023-11-14 11:02:21,388 INFO [train.py:467] (3/4) Epoch 641, batch 20, global_batch_idx: 23700, batch size: 95, loss[discriminator_loss=2.73, discriminator_real_loss=1.48, discriminator_fake_loss=1.249, generator_loss=30.32, generator_mel_loss=20.66, generator_kl_loss=2.038, generator_dur_loss=1.642, generator_adv_loss=2.156, generator_feat_match_loss=3.824, over 95.00 samples.], tot_loss[discriminator_loss=2.596, discriminator_real_loss=1.32, discriminator_fake_loss=1.276, generator_loss=30.51, generator_mel_loss=20.48, generator_kl_loss=2.004, generator_dur_loss=1.649, generator_adv_loss=2.189, generator_feat_match_loss=4.18, over 1653.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 16.0
2023-11-14 11:03:49,020 INFO [train.py:811] (3/4) Start epoch 642
2023-11-14 11:07:06,480 INFO [train.py:467] (3/4) Epoch 642, batch 33, global_batch_idx: 23750, batch size: 126, loss[discriminator_loss=2.438, discriminator_real_loss=1.298, discriminator_fake_loss=1.139, generator_loss=30.68, generator_mel_loss=20.52, generator_kl_loss=2.044, generator_dur_loss=1.619, generator_adv_loss=2.18, generator_feat_match_loss=4.312, over 126.00 samples.], tot_loss[discriminator_loss=2.556, discriminator_real_loss=1.305, discriminator_fake_loss=1.251, generator_loss=30.66, generator_mel_loss=20.35, generator_kl_loss=2.018, generator_dur_loss=1.65, generator_adv_loss=2.289, generator_feat_match_loss=4.352, over 2570.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 11:07:21,885 INFO [train.py:811] (3/4) Start epoch 643
2023-11-14 11:10:59,617 INFO [train.py:811] (3/4) Start epoch 644
2023-11-14 11:12:06,806 INFO [train.py:467] (3/4) Epoch 644, batch 9, global_batch_idx: 23800, batch size: 126, loss[discriminator_loss=2.293, discriminator_real_loss=1.111, discriminator_fake_loss=1.181, generator_loss=31.86, generator_mel_loss=20.36, generator_kl_loss=1.988, generator_dur_loss=1.617, generator_adv_loss=2.523, generator_feat_match_loss=5.363, over 126.00 samples.], tot_loss[discriminator_loss=2.435, discriminator_real_loss=1.247, discriminator_fake_loss=1.189, generator_loss=31.12, generator_mel_loss=20.15, generator_kl_loss=2.017, generator_dur_loss=1.644, generator_adv_loss=2.466, generator_feat_match_loss=4.843, over 804.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 11:12:06,808 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 11:12:18,216 INFO [train.py:517] (3/4) Epoch 644, validation: discriminator_loss=2.394, discriminator_real_loss=1.111, discriminator_fake_loss=1.283, generator_loss=31.18, generator_mel_loss=20.58, generator_kl_loss=2.156, generator_dur_loss=1.638, generator_adv_loss=1.979, generator_feat_match_loss=4.832, over 100.00 samples.
2023-11-14 11:12:18,217 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 11:14:49,391 INFO [train.py:811] (3/4) Start epoch 645
2023-11-14 11:17:05,333 INFO [train.py:467] (3/4) Epoch 645, batch 22, global_batch_idx: 23850, batch size: 52, loss[discriminator_loss=2.545, discriminator_real_loss=1.3, discriminator_fake_loss=1.245, generator_loss=29.92, generator_mel_loss=19.86, generator_kl_loss=1.979, generator_dur_loss=1.662, generator_adv_loss=2.229, generator_feat_match_loss=4.188, over 52.00 samples.], tot_loss[discriminator_loss=2.595, discriminator_real_loss=1.316, discriminator_fake_loss=1.279, generator_loss=30.38, generator_mel_loss=20.17, generator_kl_loss=1.995, generator_dur_loss=1.65, generator_adv_loss=2.218, generator_feat_match_loss=4.344, over 1692.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 11:18:21,228 INFO [train.py:811] (3/4) Start epoch 646
2023-11-14 11:21:51,441 INFO [train.py:467] (3/4) Epoch 646, batch 35, global_batch_idx: 23900, batch size: 60, loss[discriminator_loss=2.648, discriminator_real_loss=1.357, discriminator_fake_loss=1.292, generator_loss=30.6, generator_mel_loss=20.61, generator_kl_loss=2, generator_dur_loss=1.664, generator_adv_loss=2.246, generator_feat_match_loss=4.086, over 60.00 samples.], tot_loss[discriminator_loss=2.612, discriminator_real_loss=1.327, discriminator_fake_loss=1.285, generator_loss=30.52, generator_mel_loss=20.52, generator_kl_loss=2.074, generator_dur_loss=1.648, generator_adv_loss=2.163, generator_feat_match_loss=4.116, over 2723.00 samples.], cur_lr_g: 1.85e-04, cur_lr_d: 1.85e-04, grad_scale: 8.0
2023-11-14 11:21:56,614 INFO [train.py:811] (3/4) Start epoch 647
2023-11-14 11:25:22,419 INFO [train.py:811] (3/4) Start epoch 648
2023-11-14 11:26:36,079 INFO [train.py:467] (3/4) Epoch 648, batch 11, global_batch_idx: 23950, batch size: 51, loss[discriminator_loss=2.527, discriminator_real_loss=1.278, discriminator_fake_loss=1.249, generator_loss=30.34, generator_mel_loss=20.53, generator_kl_loss=1.985, generator_dur_loss=1.645, generator_adv_loss=2.156, generator_feat_match_loss=4.02, over 51.00 samples.], tot_loss[discriminator_loss=2.623, discriminator_real_loss=1.321, discriminator_fake_loss=1.302, generator_loss=30.31, generator_mel_loss=20.51, generator_kl_loss=1.982, generator_dur_loss=1.647, generator_adv_loss=2.14, generator_feat_match_loss=4.025, over 877.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:28:54,814 INFO [train.py:811] (3/4) Start epoch 649
2023-11-14 11:31:26,729 INFO [train.py:467] (3/4) Epoch 649, batch 24, global_batch_idx: 24000, batch size: 85, loss[discriminator_loss=2.496, discriminator_real_loss=1.372, discriminator_fake_loss=1.125, generator_loss=30.65, generator_mel_loss=20.64, generator_kl_loss=2, generator_dur_loss=1.635, generator_adv_loss=2.1, generator_feat_match_loss=4.277, over 85.00 samples.], tot_loss[discriminator_loss=2.616, discriminator_real_loss=1.332, discriminator_fake_loss=1.285, generator_loss=30.47, generator_mel_loss=20.47, generator_kl_loss=2.005, generator_dur_loss=1.648, generator_adv_loss=2.219, generator_feat_match_loss=4.126, over 1847.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 11:31:26,731 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 11:31:38,303 INFO [train.py:517] (3/4) Epoch 649, validation: discriminator_loss=2.62, discriminator_real_loss=1.149, discriminator_fake_loss=1.472, generator_loss=31.22, generator_mel_loss=21.2, generator_kl_loss=2.164, generator_dur_loss=1.641, generator_adv_loss=1.789, generator_feat_match_loss=4.427, over 100.00 samples.
2023-11-14 11:31:38,304 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 11:32:40,861 INFO [train.py:811] (3/4) Start epoch 650
2023-11-14 11:36:15,294 INFO [train.py:811] (3/4) Start epoch 651
2023-11-14 11:36:31,711 INFO [train.py:467] (3/4) Epoch 651, batch 0, global_batch_idx: 24050, batch size: 90, loss[discriminator_loss=2.5, discriminator_real_loss=1.144, discriminator_fake_loss=1.357, generator_loss=30.35, generator_mel_loss=20.16, generator_kl_loss=2.124, generator_dur_loss=1.633, generator_adv_loss=2.309, generator_feat_match_loss=4.129, over 90.00 samples.], tot_loss[discriminator_loss=2.5, discriminator_real_loss=1.144, discriminator_fake_loss=1.357, generator_loss=30.35, generator_mel_loss=20.16, generator_kl_loss=2.124, generator_dur_loss=1.633, generator_adv_loss=2.309, generator_feat_match_loss=4.129, over 90.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:39:52,644 INFO [train.py:811] (3/4) Start epoch 652
2023-11-14 11:41:29,874 INFO [train.py:467] (3/4) Epoch 652, batch 13, global_batch_idx: 24100, batch size: 54, loss[discriminator_loss=2.738, discriminator_real_loss=1.519, discriminator_fake_loss=1.22, generator_loss=29.5, generator_mel_loss=19.86, generator_kl_loss=1.882, generator_dur_loss=1.68, generator_adv_loss=2.107, generator_feat_match_loss=3.971, over 54.00 samples.], tot_loss[discriminator_loss=2.599, discriminator_real_loss=1.333, discriminator_fake_loss=1.266, generator_loss=30.62, generator_mel_loss=20.33, generator_kl_loss=2.032, generator_dur_loss=1.651, generator_adv_loss=2.26, generator_feat_match_loss=4.35, over 1116.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:43:33,208 INFO [train.py:811] (3/4) Start epoch 653
2023-11-14 11:46:13,080 INFO [train.py:467] (3/4) Epoch 653, batch 26, global_batch_idx: 24150, batch size: 69, loss[discriminator_loss=2.559, discriminator_real_loss=1.392, discriminator_fake_loss=1.167, generator_loss=29.96, generator_mel_loss=19.8, generator_kl_loss=2.144, generator_dur_loss=1.671, generator_adv_loss=2.316, generator_feat_match_loss=4.031, over 69.00 samples.], tot_loss[discriminator_loss=2.565, discriminator_real_loss=1.303, discriminator_fake_loss=1.262, generator_loss=30.1, generator_mel_loss=20.2, generator_kl_loss=1.995, generator_dur_loss=1.646, generator_adv_loss=2.166, generator_feat_match_loss=4.097, over 1996.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:47:09,351 INFO [train.py:811] (3/4) Start epoch 654
2023-11-14 11:50:39,221 INFO [train.py:811] (3/4) Start epoch 655
2023-11-14 11:51:04,860 INFO [train.py:467] (3/4) Epoch 655, batch 2, global_batch_idx: 24200, batch size: 65, loss[discriminator_loss=2.48, discriminator_real_loss=1.256, discriminator_fake_loss=1.224, generator_loss=31.07, generator_mel_loss=20.28, generator_kl_loss=2.062, generator_dur_loss=1.651, generator_adv_loss=2.418, generator_feat_match_loss=4.66, over 65.00 samples.], tot_loss[discriminator_loss=2.526, discriminator_real_loss=1.259, discriminator_fake_loss=1.267, generator_loss=30.29, generator_mel_loss=19.9, generator_kl_loss=2, generator_dur_loss=1.645, generator_adv_loss=2.307, generator_feat_match_loss=4.434, over 181.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:51:04,860 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 11:51:17,868 INFO [train.py:517] (3/4) Epoch 655, validation: discriminator_loss=2.579, discriminator_real_loss=1.29, discriminator_fake_loss=1.289, generator_loss=31.08, generator_mel_loss=20.77, generator_kl_loss=2.158, generator_dur_loss=1.64, generator_adv_loss=2.05, generator_feat_match_loss=4.467, over 100.00 samples.
2023-11-14 11:51:17,870 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 11:54:29,943 INFO [train.py:811] (3/4) Start epoch 656
2023-11-14 11:56:12,151 INFO [train.py:467] (3/4) Epoch 656, batch 15, global_batch_idx: 24250, batch size: 79, loss[discriminator_loss=2.598, discriminator_real_loss=1.324, discriminator_fake_loss=1.272, generator_loss=30.26, generator_mel_loss=20.34, generator_kl_loss=1.953, generator_dur_loss=1.625, generator_adv_loss=2.064, generator_feat_match_loss=4.273, over 79.00 samples.], tot_loss[discriminator_loss=2.559, discriminator_real_loss=1.3, discriminator_fake_loss=1.259, generator_loss=30.5, generator_mel_loss=20.31, generator_kl_loss=2.034, generator_dur_loss=1.643, generator_adv_loss=2.215, generator_feat_match_loss=4.297, over 1195.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 11:58:04,115 INFO [train.py:811] (3/4) Start epoch 657
2023-11-14 12:00:59,470 INFO [train.py:467] (3/4) Epoch 657, batch 28, global_batch_idx: 24300, batch size: 53, loss[discriminator_loss=2.555, discriminator_real_loss=1.235, discriminator_fake_loss=1.319, generator_loss=30.16, generator_mel_loss=20.36, generator_kl_loss=2.024, generator_dur_loss=1.665, generator_adv_loss=2.23, generator_feat_match_loss=3.883, over 53.00 samples.], tot_loss[discriminator_loss=2.634, discriminator_real_loss=1.332, discriminator_fake_loss=1.302, generator_loss=30.37, generator_mel_loss=20.36, generator_kl_loss=2.01, generator_dur_loss=1.645, generator_adv_loss=2.198, generator_feat_match_loss=4.162, over 2150.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:01:44,751 INFO [train.py:811] (3/4) Start epoch 658
2023-11-14 12:05:15,313 INFO [train.py:811] (3/4) Start epoch 659
2023-11-14 12:06:02,188 INFO [train.py:467] (3/4) Epoch 659, batch 4, global_batch_idx: 24350, batch size: 76, loss[discriminator_loss=2.309, discriminator_real_loss=1.161, discriminator_fake_loss=1.146, generator_loss=31.67, generator_mel_loss=19.94, generator_kl_loss=1.987, generator_dur_loss=1.653, generator_adv_loss=2.732, generator_feat_match_loss=5.359, over 76.00 samples.], tot_loss[discriminator_loss=2.462, discriminator_real_loss=1.275, discriminator_fake_loss=1.187, generator_loss=31.14, generator_mel_loss=20.1, generator_kl_loss=2.03, generator_dur_loss=1.638, generator_adv_loss=2.478, generator_feat_match_loss=4.887, over 437.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:08:52,125 INFO [train.py:811] (3/4) Start epoch 660
2023-11-14 12:10:43,925 INFO [train.py:467] (3/4) Epoch 660, batch 17, global_batch_idx: 24400, batch size: 63, loss[discriminator_loss=2.598, discriminator_real_loss=1.429, discriminator_fake_loss=1.169, generator_loss=29.72, generator_mel_loss=19.96, generator_kl_loss=2.055, generator_dur_loss=1.624, generator_adv_loss=2.084, generator_feat_match_loss=4, over 63.00 samples.], tot_loss[discriminator_loss=2.555, discriminator_real_loss=1.287, discriminator_fake_loss=1.269, generator_loss=30.32, generator_mel_loss=20.19, generator_kl_loss=2.01, generator_dur_loss=1.641, generator_adv_loss=2.219, generator_feat_match_loss=4.257, over 1338.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 12:10:43,927 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 12:10:55,823 INFO [train.py:517] (3/4) Epoch 660, validation: discriminator_loss=2.644, discriminator_real_loss=1.146, discriminator_fake_loss=1.498, generator_loss=30.52, generator_mel_loss=20.88, generator_kl_loss=2.231, generator_dur_loss=1.644, generator_adv_loss=1.815, generator_feat_match_loss=3.949, over 100.00 samples.
2023-11-14 12:10:55,825 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 12:12:42,237 INFO [train.py:811] (3/4) Start epoch 661
2023-11-14 12:15:44,467 INFO [train.py:467] (3/4) Epoch 661, batch 30, global_batch_idx: 24450, batch size: 60, loss[discriminator_loss=2.418, discriminator_real_loss=1.236, discriminator_fake_loss=1.182, generator_loss=31.06, generator_mel_loss=20.01, generator_kl_loss=2.035, generator_dur_loss=1.636, generator_adv_loss=2.283, generator_feat_match_loss=5.094, over 60.00 samples.], tot_loss[discriminator_loss=2.573, discriminator_real_loss=1.315, discriminator_fake_loss=1.258, generator_loss=30.72, generator_mel_loss=20.36, generator_kl_loss=2.022, generator_dur_loss=1.652, generator_adv_loss=2.282, generator_feat_match_loss=4.401, over 2193.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:16:19,733 INFO [train.py:811] (3/4) Start epoch 662
2023-11-14 12:19:55,042 INFO [train.py:811] (3/4) Start epoch 663
2023-11-14 12:20:42,979 INFO [train.py:467] (3/4) Epoch 663, batch 6, global_batch_idx: 24500, batch size: 69, loss[discriminator_loss=2.516, discriminator_real_loss=1.256, discriminator_fake_loss=1.261, generator_loss=30.24, generator_mel_loss=20.16, generator_kl_loss=1.969, generator_dur_loss=1.639, generator_adv_loss=2.17, generator_feat_match_loss=4.297, over 69.00 samples.], tot_loss[discriminator_loss=2.587, discriminator_real_loss=1.321, discriminator_fake_loss=1.265, generator_loss=30.52, generator_mel_loss=20.44, generator_kl_loss=2.034, generator_dur_loss=1.652, generator_adv_loss=2.164, generator_feat_match_loss=4.235, over 530.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:23:27,244 INFO [train.py:811] (3/4) Start epoch 664
2023-11-14 12:25:28,989 INFO [train.py:467] (3/4) Epoch 664, batch 19, global_batch_idx: 24550, batch size: 110, loss[discriminator_loss=2.523, discriminator_real_loss=1.316, discriminator_fake_loss=1.206, generator_loss=31.1, generator_mel_loss=20.65, generator_kl_loss=2.023, generator_dur_loss=1.648, generator_adv_loss=2.273, generator_feat_match_loss=4.5, over 110.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.343, discriminator_fake_loss=1.272, generator_loss=30.37, generator_mel_loss=20.27, generator_kl_loss=2.013, generator_dur_loss=1.646, generator_adv_loss=2.241, generator_feat_match_loss=4.196, over 1499.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:27:02,637 INFO [train.py:811] (3/4) Start epoch 665
2023-11-14 12:30:08,315 INFO [train.py:467] (3/4) Epoch 665, batch 32, global_batch_idx: 24600, batch size: 85, loss[discriminator_loss=2.578, discriminator_real_loss=1.317, discriminator_fake_loss=1.26, generator_loss=30.75, generator_mel_loss=20.18, generator_kl_loss=2.007, generator_dur_loss=1.628, generator_adv_loss=2.596, generator_feat_match_loss=4.336, over 85.00 samples.], tot_loss[discriminator_loss=2.587, discriminator_real_loss=1.312, discriminator_fake_loss=1.274, generator_loss=30.59, generator_mel_loss=20.43, generator_kl_loss=2.008, generator_dur_loss=1.65, generator_adv_loss=2.231, generator_feat_match_loss=4.273, over 2321.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:30:08,317 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 12:30:19,982 INFO [train.py:517] (3/4) Epoch 665, validation: discriminator_loss=2.543, discriminator_real_loss=1.435, discriminator_fake_loss=1.108, generator_loss=31.54, generator_mel_loss=20.99, generator_kl_loss=2.102, generator_dur_loss=1.638, generator_adv_loss=2.564, generator_feat_match_loss=4.24, over 100.00 samples.
2023-11-14 12:30:19,983 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 12:30:43,570 INFO [train.py:811] (3/4) Start epoch 666
2023-11-14 12:34:21,509 INFO [train.py:811] (3/4) Start epoch 667
2023-11-14 12:35:22,826 INFO [train.py:467] (3/4) Epoch 667, batch 8, global_batch_idx: 24650, batch size: 61, loss[discriminator_loss=2.74, discriminator_real_loss=1.151, discriminator_fake_loss=1.589, generator_loss=29.49, generator_mel_loss=19.79, generator_kl_loss=1.951, generator_dur_loss=1.642, generator_adv_loss=2.377, generator_feat_match_loss=3.73, over 61.00 samples.], tot_loss[discriminator_loss=2.556, discriminator_real_loss=1.273, discriminator_fake_loss=1.283, generator_loss=30.91, generator_mel_loss=20.22, generator_kl_loss=2.005, generator_dur_loss=1.642, generator_adv_loss=2.39, generator_feat_match_loss=4.65, over 660.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:37:53,824 INFO [train.py:811] (3/4) Start epoch 668
2023-11-14 12:40:02,952 INFO [train.py:467] (3/4) Epoch 668, batch 21, global_batch_idx: 24700, batch size: 85, loss[discriminator_loss=2.52, discriminator_real_loss=1.255, discriminator_fake_loss=1.264, generator_loss=30.81, generator_mel_loss=20.71, generator_kl_loss=2.044, generator_dur_loss=1.633, generator_adv_loss=2.109, generator_feat_match_loss=4.309, over 85.00 samples.], tot_loss[discriminator_loss=2.588, discriminator_real_loss=1.305, discriminator_fake_loss=1.283, generator_loss=30.21, generator_mel_loss=20.22, generator_kl_loss=2.004, generator_dur_loss=1.642, generator_adv_loss=2.139, generator_feat_match_loss=4.202, over 1586.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:41:33,322 INFO [train.py:811] (3/4) Start epoch 669
2023-11-14 12:44:52,243 INFO [train.py:467] (3/4) Epoch 669, batch 34, global_batch_idx: 24750, batch size: 79, loss[discriminator_loss=2.543, discriminator_real_loss=1.124, discriminator_fake_loss=1.418, generator_loss=30.56, generator_mel_loss=20.12, generator_kl_loss=2.11, generator_dur_loss=1.656, generator_adv_loss=2.258, generator_feat_match_loss=4.41, over 79.00 samples.], tot_loss[discriminator_loss=2.594, discriminator_real_loss=1.312, discriminator_fake_loss=1.282, generator_loss=30.63, generator_mel_loss=20.4, generator_kl_loss=2.018, generator_dur_loss=1.648, generator_adv_loss=2.248, generator_feat_match_loss=4.312, over 2379.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:45:05,630 INFO [train.py:811] (3/4) Start epoch 670
2023-11-14 12:48:36,937 INFO [train.py:811] (3/4) Start epoch 671
2023-11-14 12:49:48,435 INFO [train.py:467] (3/4) Epoch 671, batch 10, global_batch_idx: 24800, batch size: 64, loss[discriminator_loss=2.512, discriminator_real_loss=1.322, discriminator_fake_loss=1.19, generator_loss=31.19, generator_mel_loss=20.38, generator_kl_loss=1.977, generator_dur_loss=1.652, generator_adv_loss=2.301, generator_feat_match_loss=4.883, over 64.00 samples.], tot_loss[discriminator_loss=2.521, discriminator_real_loss=1.261, discriminator_fake_loss=1.26, generator_loss=30.61, generator_mel_loss=20.26, generator_kl_loss=2.02, generator_dur_loss=1.649, generator_adv_loss=2.244, generator_feat_match_loss=4.435, over 757.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 12:49:48,437 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 12:50:00,621 INFO [train.py:517] (3/4) Epoch 671, validation: discriminator_loss=2.513, discriminator_real_loss=1.152, discriminator_fake_loss=1.361, generator_loss=30.93, generator_mel_loss=20.81, generator_kl_loss=2.15, generator_dur_loss=1.638, generator_adv_loss=1.923, generator_feat_match_loss=4.411, over 100.00 samples.
2023-11-14 12:50:00,622 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 12:52:23,760 INFO [train.py:811] (3/4) Start epoch 672
2023-11-14 12:54:45,557 INFO [train.py:467] (3/4) Epoch 672, batch 23, global_batch_idx: 24850, batch size: 59, loss[discriminator_loss=2.551, discriminator_real_loss=1.279, discriminator_fake_loss=1.272, generator_loss=30.94, generator_mel_loss=20.56, generator_kl_loss=2.128, generator_dur_loss=1.654, generator_adv_loss=2.301, generator_feat_match_loss=4.301, over 59.00 samples.], tot_loss[discriminator_loss=2.569, discriminator_real_loss=1.28, discriminator_fake_loss=1.288, generator_loss=30.24, generator_mel_loss=20.13, generator_kl_loss=2.021, generator_dur_loss=1.641, generator_adv_loss=2.195, generator_feat_match_loss=4.247, over 1830.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:55:56,052 INFO [train.py:811] (3/4) Start epoch 673
2023-11-14 12:59:29,728 INFO [train.py:467] (3/4) Epoch 673, batch 36, global_batch_idx: 24900, batch size: 79, loss[discriminator_loss=2.516, discriminator_real_loss=1.263, discriminator_fake_loss=1.254, generator_loss=30.63, generator_mel_loss=20.52, generator_kl_loss=2.035, generator_dur_loss=1.641, generator_adv_loss=2.109, generator_feat_match_loss=4.32, over 79.00 samples.], tot_loss[discriminator_loss=2.603, discriminator_real_loss=1.32, discriminator_fake_loss=1.283, generator_loss=30.39, generator_mel_loss=20.38, generator_kl_loss=2.043, generator_dur_loss=1.648, generator_adv_loss=2.155, generator_feat_match_loss=4.16, over 2586.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 12:59:30,413 INFO [train.py:811] (3/4) Start epoch 674
2023-11-14 13:03:01,800 INFO [train.py:811] (3/4) Start epoch 675
2023-11-14 13:04:20,052 INFO [train.py:467] (3/4) Epoch 675, batch 12, global_batch_idx: 24950, batch size: 95, loss[discriminator_loss=2.348, discriminator_real_loss=1.215, discriminator_fake_loss=1.132, generator_loss=32.39, generator_mel_loss=20.23, generator_kl_loss=2.078, generator_dur_loss=1.627, generator_adv_loss=2.559, generator_feat_match_loss=5.898, over 95.00 samples.], tot_loss[discriminator_loss=2.541, discriminator_real_loss=1.314, discriminator_fake_loss=1.227, generator_loss=31.05, generator_mel_loss=20.14, generator_kl_loss=1.971, generator_dur_loss=1.638, generator_adv_loss=2.423, generator_feat_match_loss=4.883, over 1007.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:06:42,568 INFO [train.py:811] (3/4) Start epoch 676
2023-11-14 13:09:11,621 INFO [train.py:467] (3/4) Epoch 676, batch 25, global_batch_idx: 25000, batch size: 153, loss[discriminator_loss=2.564, discriminator_real_loss=1.22, discriminator_fake_loss=1.345, generator_loss=30.99, generator_mel_loss=20.55, generator_kl_loss=2.096, generator_dur_loss=1.656, generator_adv_loss=2.205, generator_feat_match_loss=4.477, over 153.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.292, discriminator_fake_loss=1.275, generator_loss=30.36, generator_mel_loss=20.36, generator_kl_loss=2.033, generator_dur_loss=1.648, generator_adv_loss=2.158, generator_feat_match_loss=4.163, over 1803.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:09:11,623 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 13:09:23,430 INFO [train.py:517] (3/4) Epoch 676, validation: discriminator_loss=2.535, discriminator_real_loss=1.263, discriminator_fake_loss=1.272, generator_loss=31.18, generator_mel_loss=20.96, generator_kl_loss=2.146, generator_dur_loss=1.636, generator_adv_loss=2.061, generator_feat_match_loss=4.378, over 100.00 samples.
2023-11-14 13:09:23,431 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 13:10:27,103 INFO [train.py:811] (3/4) Start epoch 677
2023-11-14 13:13:59,602 INFO [train.py:811] (3/4) Start epoch 678
2023-11-14 13:14:22,104 INFO [train.py:467] (3/4) Epoch 678, batch 1, global_batch_idx: 25050, batch size: 60, loss[discriminator_loss=2.426, discriminator_real_loss=1.249, discriminator_fake_loss=1.178, generator_loss=31.03, generator_mel_loss=20.4, generator_kl_loss=2.004, generator_dur_loss=1.656, generator_adv_loss=2.309, generator_feat_match_loss=4.66, over 60.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.355, discriminator_fake_loss=1.184, generator_loss=31, generator_mel_loss=20.44, generator_kl_loss=1.992, generator_dur_loss=1.654, generator_adv_loss=2.349, generator_feat_match_loss=4.557, over 170.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:17:27,780 INFO [train.py:811] (3/4) Start epoch 679
2023-11-14 13:19:06,699 INFO [train.py:467] (3/4) Epoch 679, batch 14, global_batch_idx: 25100, batch size: 73, loss[discriminator_loss=2.566, discriminator_real_loss=1.417, discriminator_fake_loss=1.149, generator_loss=31.83, generator_mel_loss=20.84, generator_kl_loss=2.126, generator_dur_loss=1.645, generator_adv_loss=2.594, generator_feat_match_loss=4.633, over 73.00 samples.], tot_loss[discriminator_loss=2.625, discriminator_real_loss=1.318, discriminator_fake_loss=1.307, generator_loss=30.74, generator_mel_loss=20.38, generator_kl_loss=2.045, generator_dur_loss=1.638, generator_adv_loss=2.273, generator_feat_match_loss=4.399, over 1347.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:21:03,879 INFO [train.py:811] (3/4) Start epoch 680
2023-11-14 13:23:45,458 INFO [train.py:467] (3/4) Epoch 680, batch 27, global_batch_idx: 25150, batch size: 55, loss[discriminator_loss=2.539, discriminator_real_loss=1.407, discriminator_fake_loss=1.131, generator_loss=31.03, generator_mel_loss=20.39, generator_kl_loss=1.963, generator_dur_loss=1.673, generator_adv_loss=2.453, generator_feat_match_loss=4.551, over 55.00 samples.], tot_loss[discriminator_loss=2.538, discriminator_real_loss=1.277, discriminator_fake_loss=1.261, generator_loss=30.75, generator_mel_loss=20.36, generator_kl_loss=2.02, generator_dur_loss=1.64, generator_adv_loss=2.277, generator_feat_match_loss=4.455, over 1966.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:24:32,626 INFO [train.py:811] (3/4) Start epoch 681
2023-11-14 13:28:07,697 INFO [train.py:811] (3/4) Start epoch 682
2023-11-14 13:28:42,545 INFO [train.py:467] (3/4) Epoch 682, batch 3, global_batch_idx: 25200, batch size: 55, loss[discriminator_loss=2.309, discriminator_real_loss=1.176, discriminator_fake_loss=1.134, generator_loss=31.21, generator_mel_loss=19.86, generator_kl_loss=2.008, generator_dur_loss=1.669, generator_adv_loss=2.566, generator_feat_match_loss=5.105, over 55.00 samples.], tot_loss[discriminator_loss=2.378, discriminator_real_loss=1.212, discriminator_fake_loss=1.166, generator_loss=31.79, generator_mel_loss=20.35, generator_kl_loss=1.982, generator_dur_loss=1.642, generator_adv_loss=2.558, generator_feat_match_loss=5.256, over 378.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 13:28:42,547 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 13:28:56,168 INFO [train.py:517] (3/4) Epoch 682, validation: discriminator_loss=2.282, discriminator_real_loss=1.218, discriminator_fake_loss=1.064, generator_loss=32.04, generator_mel_loss=20.55, generator_kl_loss=2.159, generator_dur_loss=1.636, generator_adv_loss=2.418, generator_feat_match_loss=5.276, over 100.00 samples.
2023-11-14 13:28:56,169 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 13:31:51,151 INFO [train.py:811] (3/4) Start epoch 683
2023-11-14 13:33:37,746 INFO [train.py:467] (3/4) Epoch 683, batch 16, global_batch_idx: 25250, batch size: 54, loss[discriminator_loss=2.584, discriminator_real_loss=1.367, discriminator_fake_loss=1.217, generator_loss=29.42, generator_mel_loss=19.77, generator_kl_loss=2.015, generator_dur_loss=1.627, generator_adv_loss=2.049, generator_feat_match_loss=3.955, over 54.00 samples.], tot_loss[discriminator_loss=2.614, discriminator_real_loss=1.333, discriminator_fake_loss=1.281, generator_loss=30.24, generator_mel_loss=20.28, generator_kl_loss=2.012, generator_dur_loss=1.64, generator_adv_loss=2.151, generator_feat_match_loss=4.161, over 1083.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 13:35:24,466 INFO [train.py:811] (3/4) Start epoch 684
2023-11-14 13:38:20,601 INFO [train.py:467] (3/4) Epoch 684, batch 29, global_batch_idx: 25300, batch size: 63, loss[discriminator_loss=2.676, discriminator_real_loss=1.356, discriminator_fake_loss=1.32, generator_loss=30.46, generator_mel_loss=20.58, generator_kl_loss=2.029, generator_dur_loss=1.64, generator_adv_loss=2.359, generator_feat_match_loss=3.852, over 63.00 samples.], tot_loss[discriminator_loss=2.605, discriminator_real_loss=1.315, discriminator_fake_loss=1.29, generator_loss=30.52, generator_mel_loss=20.41, generator_kl_loss=1.996, generator_dur_loss=1.643, generator_adv_loss=2.226, generator_feat_match_loss=4.248, over 2084.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 13:38:59,304 INFO [train.py:811] (3/4) Start epoch 685
2023-11-14 13:42:38,373 INFO [train.py:811] (3/4) Start epoch 686
2023-11-14 13:43:29,177 INFO [train.py:467] (3/4) Epoch 686, batch 5, global_batch_idx: 25350, batch size: 65, loss[discriminator_loss=2.537, discriminator_real_loss=1.284, discriminator_fake_loss=1.253, generator_loss=30.53, generator_mel_loss=20.47, generator_kl_loss=1.886, generator_dur_loss=1.675, generator_adv_loss=2.338, generator_feat_match_loss=4.16, over 65.00 samples.], tot_loss[discriminator_loss=2.575, discriminator_real_loss=1.316, discriminator_fake_loss=1.259, generator_loss=30.88, generator_mel_loss=20.3, generator_kl_loss=2, generator_dur_loss=1.66, generator_adv_loss=2.313, generator_feat_match_loss=4.61, over 438.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 16.0
2023-11-14 13:46:12,265 INFO [train.py:811] (3/4) Start epoch 687
2023-11-14 13:48:08,381 INFO [train.py:467] (3/4) Epoch 687, batch 18, global_batch_idx: 25400, batch size: 73, loss[discriminator_loss=2.305, discriminator_real_loss=1.129, discriminator_fake_loss=1.177, generator_loss=30.8, generator_mel_loss=19.9, generator_kl_loss=1.893, generator_dur_loss=1.637, generator_adv_loss=2.369, generator_feat_match_loss=5.004, over 73.00 samples.], tot_loss[discriminator_loss=2.491, discriminator_real_loss=1.245, discriminator_fake_loss=1.247, generator_loss=30.79, generator_mel_loss=19.97, generator_kl_loss=2.016, generator_dur_loss=1.638, generator_adv_loss=2.395, generator_feat_match_loss=4.776, over 1378.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:48:08,383 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 13:48:20,531 INFO [train.py:517] (3/4) Epoch 687, validation: discriminator_loss=2.486, discriminator_real_loss=1.159, discriminator_fake_loss=1.328, generator_loss=30.94, generator_mel_loss=20.78, generator_kl_loss=2.145, generator_dur_loss=1.627, generator_adv_loss=1.893, generator_feat_match_loss=4.492, over 100.00 samples.
2023-11-14 13:48:20,532 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 13:50:01,125 INFO [train.py:811] (3/4) Start epoch 688
2023-11-14 13:53:06,013 INFO [train.py:467] (3/4) Epoch 688, batch 31, global_batch_idx: 25450, batch size: 76, loss[discriminator_loss=2.533, discriminator_real_loss=1.337, discriminator_fake_loss=1.196, generator_loss=30.71, generator_mel_loss=20.54, generator_kl_loss=2.038, generator_dur_loss=1.637, generator_adv_loss=2.109, generator_feat_match_loss=4.379, over 76.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.322, discriminator_fake_loss=1.259, generator_loss=30.31, generator_mel_loss=20.31, generator_kl_loss=2.022, generator_dur_loss=1.645, generator_adv_loss=2.137, generator_feat_match_loss=4.199, over 2260.00 samples.], cur_lr_g: 1.84e-04, cur_lr_d: 1.84e-04, grad_scale: 8.0
2023-11-14 13:53:36,030 INFO [train.py:811] (3/4) Start epoch 689
2023-11-14 13:57:05,694 INFO [train.py:811] (3/4) Start epoch 690
2023-11-14 13:58:01,047 INFO [train.py:467] (3/4) Epoch 690, batch 7, global_batch_idx: 25500, batch size: 65, loss[discriminator_loss=2.602, discriminator_real_loss=1.363, discriminator_fake_loss=1.237, generator_loss=30, generator_mel_loss=20.14, generator_kl_loss=2.028, generator_dur_loss=1.647, generator_adv_loss=1.986, generator_feat_match_loss=4.203, over 65.00 samples.], tot_loss[discriminator_loss=2.653, discriminator_real_loss=1.33, discriminator_fake_loss=1.323, generator_loss=30.23, generator_mel_loss=20.21, generator_kl_loss=2.025, generator_dur_loss=1.645, generator_adv_loss=2.216, generator_feat_match_loss=4.132, over 548.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:00:44,129 INFO [train.py:811] (3/4) Start epoch 691
2023-11-14 14:02:49,299 INFO [train.py:467] (3/4) Epoch 691, batch 20, global_batch_idx: 25550, batch size: 52, loss[discriminator_loss=2.488, discriminator_real_loss=1.247, discriminator_fake_loss=1.241, generator_loss=30.98, generator_mel_loss=20.24, generator_kl_loss=2.066, generator_dur_loss=1.644, generator_adv_loss=2.367, generator_feat_match_loss=4.664, over 52.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.288, discriminator_fake_loss=1.286, generator_loss=30.61, generator_mel_loss=20.1, generator_kl_loss=2.007, generator_dur_loss=1.648, generator_adv_loss=2.29, generator_feat_match_loss=4.57, over 1415.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:04:19,973 INFO [train.py:811] (3/4) Start epoch 692
2023-11-14 14:07:38,774 INFO [train.py:467] (3/4) Epoch 692, batch 33, global_batch_idx: 25600, batch size: 81, loss[discriminator_loss=2.531, discriminator_real_loss=1.304, discriminator_fake_loss=1.229, generator_loss=30.63, generator_mel_loss=20.04, generator_kl_loss=2.026, generator_dur_loss=1.627, generator_adv_loss=2.396, generator_feat_match_loss=4.539, over 81.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.298, discriminator_fake_loss=1.266, generator_loss=30.36, generator_mel_loss=20.24, generator_kl_loss=2.036, generator_dur_loss=1.642, generator_adv_loss=2.18, generator_feat_match_loss=4.265, over 2568.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 14:07:38,776 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 14:07:50,035 INFO [train.py:517] (3/4) Epoch 692, validation: discriminator_loss=2.648, discriminator_real_loss=1.176, discriminator_fake_loss=1.471, generator_loss=30.91, generator_mel_loss=21.14, generator_kl_loss=2.264, generator_dur_loss=1.628, generator_adv_loss=1.743, generator_feat_match_loss=4.13, over 100.00 samples.
2023-11-14 14:07:50,037 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 14:08:02,662 INFO [train.py:811] (3/4) Start epoch 693
2023-11-14 14:11:34,041 INFO [train.py:811] (3/4) Start epoch 694
2023-11-14 14:12:37,429 INFO [train.py:467] (3/4) Epoch 694, batch 9, global_batch_idx: 25650, batch size: 69, loss[discriminator_loss=2.641, discriminator_real_loss=1.347, discriminator_fake_loss=1.293, generator_loss=30.32, generator_mel_loss=20.11, generator_kl_loss=2.042, generator_dur_loss=1.639, generator_adv_loss=2.166, generator_feat_match_loss=4.363, over 69.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.322, discriminator_fake_loss=1.252, generator_loss=30.5, generator_mel_loss=20.24, generator_kl_loss=2.004, generator_dur_loss=1.656, generator_adv_loss=2.247, generator_feat_match_loss=4.348, over 607.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:15:06,001 INFO [train.py:811] (3/4) Start epoch 695
2023-11-14 14:17:23,170 INFO [train.py:467] (3/4) Epoch 695, batch 22, global_batch_idx: 25700, batch size: 53, loss[discriminator_loss=2.613, discriminator_real_loss=1.428, discriminator_fake_loss=1.185, generator_loss=29.81, generator_mel_loss=20, generator_kl_loss=1.99, generator_dur_loss=1.68, generator_adv_loss=2.137, generator_feat_match_loss=4.004, over 53.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.314, discriminator_fake_loss=1.271, generator_loss=30.35, generator_mel_loss=20.22, generator_kl_loss=2.031, generator_dur_loss=1.646, generator_adv_loss=2.218, generator_feat_match_loss=4.229, over 1715.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:18:36,432 INFO [train.py:811] (3/4) Start epoch 696
2023-11-14 14:21:53,187 INFO [train.py:467] (3/4) Epoch 696, batch 35, global_batch_idx: 25750, batch size: 51, loss[discriminator_loss=2.645, discriminator_real_loss=1.17, discriminator_fake_loss=1.475, generator_loss=29.78, generator_mel_loss=19.87, generator_kl_loss=1.995, generator_dur_loss=1.665, generator_adv_loss=2.004, generator_feat_match_loss=4.246, over 51.00 samples.], tot_loss[discriminator_loss=2.551, discriminator_real_loss=1.276, discriminator_fake_loss=1.275, generator_loss=30.73, generator_mel_loss=20.21, generator_kl_loss=2.023, generator_dur_loss=1.644, generator_adv_loss=2.317, generator_feat_match_loss=4.539, over 2710.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:22:01,654 INFO [train.py:811] (3/4) Start epoch 697
2023-11-14 14:25:36,571 INFO [train.py:811] (3/4) Start epoch 698
2023-11-14 14:26:58,928 INFO [train.py:467] (3/4) Epoch 698, batch 11, global_batch_idx: 25800, batch size: 61, loss[discriminator_loss=2.793, discriminator_real_loss=1.408, discriminator_fake_loss=1.385, generator_loss=29.46, generator_mel_loss=20.05, generator_kl_loss=1.962, generator_dur_loss=1.636, generator_adv_loss=1.989, generator_feat_match_loss=3.822, over 61.00 samples.], tot_loss[discriminator_loss=2.426, discriminator_real_loss=1.241, discriminator_fake_loss=1.185, generator_loss=31.06, generator_mel_loss=19.98, generator_kl_loss=1.998, generator_dur_loss=1.639, generator_adv_loss=2.412, generator_feat_match_loss=5.031, over 863.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:26:58,930 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 14:27:10,095 INFO [train.py:517] (3/4) Epoch 698, validation: discriminator_loss=2.54, discriminator_real_loss=1.189, discriminator_fake_loss=1.35, generator_loss=30.44, generator_mel_loss=20.34, generator_kl_loss=2.23, generator_dur_loss=1.631, generator_adv_loss=2.048, generator_feat_match_loss=4.19, over 100.00 samples.
2023-11-14 14:27:10,097 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 14:29:22,819 INFO [train.py:811] (3/4) Start epoch 699
2023-11-14 14:31:40,901 INFO [train.py:467] (3/4) Epoch 699, batch 24, global_batch_idx: 25850, batch size: 61, loss[discriminator_loss=2.617, discriminator_real_loss=1.306, discriminator_fake_loss=1.312, generator_loss=29.47, generator_mel_loss=19.57, generator_kl_loss=2.053, generator_dur_loss=1.65, generator_adv_loss=2.242, generator_feat_match_loss=3.953, over 61.00 samples.], tot_loss[discriminator_loss=2.554, discriminator_real_loss=1.285, discriminator_fake_loss=1.269, generator_loss=30.41, generator_mel_loss=20.21, generator_kl_loss=2.036, generator_dur_loss=1.639, generator_adv_loss=2.203, generator_feat_match_loss=4.324, over 1863.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:32:52,121 INFO [train.py:811] (3/4) Start epoch 700
2023-11-14 14:36:23,013 INFO [train.py:811] (3/4) Start epoch 701
2023-11-14 14:36:40,047 INFO [train.py:467] (3/4) Epoch 701, batch 0, global_batch_idx: 25900, batch size: 65, loss[discriminator_loss=2.762, discriminator_real_loss=1.162, discriminator_fake_loss=1.599, generator_loss=30.58, generator_mel_loss=20.38, generator_kl_loss=1.965, generator_dur_loss=1.67, generator_adv_loss=2.191, generator_feat_match_loss=4.379, over 65.00 samples.], tot_loss[discriminator_loss=2.762, discriminator_real_loss=1.162, discriminator_fake_loss=1.599, generator_loss=30.58, generator_mel_loss=20.38, generator_kl_loss=1.965, generator_dur_loss=1.67, generator_adv_loss=2.191, generator_feat_match_loss=4.379, over 65.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:39:57,566 INFO [train.py:811] (3/4) Start epoch 702
2023-11-14 14:41:25,229 INFO [train.py:467] (3/4) Epoch 702, batch 13, global_batch_idx: 25950, batch size: 56, loss[discriminator_loss=2.52, discriminator_real_loss=1.172, discriminator_fake_loss=1.348, generator_loss=31.16, generator_mel_loss=20.56, generator_kl_loss=2.06, generator_dur_loss=1.646, generator_adv_loss=2.23, generator_feat_match_loss=4.664, over 56.00 samples.], tot_loss[discriminator_loss=2.555, discriminator_real_loss=1.264, discriminator_fake_loss=1.291, generator_loss=30.51, generator_mel_loss=20.27, generator_kl_loss=2.071, generator_dur_loss=1.638, generator_adv_loss=2.156, generator_feat_match_loss=4.381, over 1075.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 14:43:31,251 INFO [train.py:811] (3/4) Start epoch 703
2023-11-14 14:46:10,176 INFO [train.py:467] (3/4) Epoch 703, batch 26, global_batch_idx: 26000, batch size: 56, loss[discriminator_loss=2.547, discriminator_real_loss=1.317, discriminator_fake_loss=1.229, generator_loss=31.1, generator_mel_loss=20.46, generator_kl_loss=2.07, generator_dur_loss=1.644, generator_adv_loss=2.324, generator_feat_match_loss=4.602, over 56.00 samples.], tot_loss[discriminator_loss=2.611, discriminator_real_loss=1.319, discriminator_fake_loss=1.292, generator_loss=30.39, generator_mel_loss=20.3, generator_kl_loss=2.026, generator_dur_loss=1.635, generator_adv_loss=2.174, generator_feat_match_loss=4.252, over 1825.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 14:46:10,178 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 14:46:21,215 INFO [train.py:517] (3/4) Epoch 703, validation: discriminator_loss=2.567, discriminator_real_loss=1.261, discriminator_fake_loss=1.305, generator_loss=31.06, generator_mel_loss=20.82, generator_kl_loss=2.197, generator_dur_loss=1.635, generator_adv_loss=2.174, generator_feat_match_loss=4.234, over 100.00 samples.
2023-11-14 14:46:21,216 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 14:47:22,530 INFO [train.py:811] (3/4) Start epoch 704
2023-11-14 14:50:54,916 INFO [train.py:811] (3/4) Start epoch 705
2023-11-14 14:51:19,639 INFO [train.py:467] (3/4) Epoch 705, batch 2, global_batch_idx: 26050, batch size: 90, loss[discriminator_loss=2.689, discriminator_real_loss=1.299, discriminator_fake_loss=1.391, generator_loss=31.09, generator_mel_loss=20.82, generator_kl_loss=2.049, generator_dur_loss=1.66, generator_adv_loss=2.23, generator_feat_match_loss=4.336, over 90.00 samples.], tot_loss[discriminator_loss=2.725, discriminator_real_loss=1.355, discriminator_fake_loss=1.37, generator_loss=30.44, generator_mel_loss=20.55, generator_kl_loss=1.997, generator_dur_loss=1.674, generator_adv_loss=2.14, generator_feat_match_loss=4.077, over 197.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 14:54:27,588 INFO [train.py:811] (3/4) Start epoch 706
2023-11-14 14:56:04,335 INFO [train.py:467] (3/4) Epoch 706, batch 15, global_batch_idx: 26100, batch size: 54, loss[discriminator_loss=2.617, discriminator_real_loss=1.348, discriminator_fake_loss=1.27, generator_loss=30.36, generator_mel_loss=20.62, generator_kl_loss=2.033, generator_dur_loss=1.683, generator_adv_loss=2.09, generator_feat_match_loss=3.938, over 54.00 samples.], tot_loss[discriminator_loss=2.631, discriminator_real_loss=1.348, discriminator_fake_loss=1.283, generator_loss=30.52, generator_mel_loss=20.42, generator_kl_loss=2.013, generator_dur_loss=1.654, generator_adv_loss=2.2, generator_feat_match_loss=4.228, over 1039.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 14:57:56,717 INFO [train.py:811] (3/4) Start epoch 707
2023-11-14 15:00:41,818 INFO [train.py:467] (3/4) Epoch 707, batch 28, global_batch_idx: 26150, batch size: 69, loss[discriminator_loss=2.598, discriminator_real_loss=1.184, discriminator_fake_loss=1.415, generator_loss=30.01, generator_mel_loss=20.22, generator_kl_loss=2.021, generator_dur_loss=1.64, generator_adv_loss=1.953, generator_feat_match_loss=4.176, over 69.00 samples.], tot_loss[discriminator_loss=2.526, discriminator_real_loss=1.265, discriminator_fake_loss=1.261, generator_loss=30.87, generator_mel_loss=20.19, generator_kl_loss=2.006, generator_dur_loss=1.639, generator_adv_loss=2.368, generator_feat_match_loss=4.667, over 2214.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:01:28,957 INFO [train.py:811] (3/4) Start epoch 708
2023-11-14 15:05:04,118 INFO [train.py:811] (3/4) Start epoch 709
2023-11-14 15:05:38,916 INFO [train.py:467] (3/4) Epoch 709, batch 4, global_batch_idx: 26200, batch size: 73, loss[discriminator_loss=2.492, discriminator_real_loss=1.329, discriminator_fake_loss=1.163, generator_loss=30.5, generator_mel_loss=20.19, generator_kl_loss=1.97, generator_dur_loss=1.641, generator_adv_loss=2.213, generator_feat_match_loss=4.48, over 73.00 samples.], tot_loss[discriminator_loss=2.548, discriminator_real_loss=1.332, discriminator_fake_loss=1.215, generator_loss=30.36, generator_mel_loss=20.12, generator_kl_loss=2.012, generator_dur_loss=1.64, generator_adv_loss=2.274, generator_feat_match_loss=4.312, over 320.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:05:38,918 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 15:05:50,929 INFO [train.py:517] (3/4) Epoch 709, validation: discriminator_loss=2.445, discriminator_real_loss=1.196, discriminator_fake_loss=1.249, generator_loss=31.68, generator_mel_loss=20.94, generator_kl_loss=2.097, generator_dur_loss=1.63, generator_adv_loss=2.2, generator_feat_match_loss=4.81, over 100.00 samples.
2023-11-14 15:05:50,930 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 15:08:42,241 INFO [train.py:811] (3/4) Start epoch 710
2023-11-14 15:10:39,636 INFO [train.py:467] (3/4) Epoch 710, batch 17, global_batch_idx: 26250, batch size: 61, loss[discriminator_loss=2.523, discriminator_real_loss=1.335, discriminator_fake_loss=1.189, generator_loss=30.04, generator_mel_loss=19.8, generator_kl_loss=2.103, generator_dur_loss=1.654, generator_adv_loss=2.426, generator_feat_match_loss=4.059, over 61.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.331, discriminator_fake_loss=1.27, generator_loss=30.3, generator_mel_loss=19.94, generator_kl_loss=2.018, generator_dur_loss=1.645, generator_adv_loss=2.242, generator_feat_match_loss=4.459, over 1366.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:12:18,855 INFO [train.py:811] (3/4) Start epoch 711
2023-11-14 15:15:28,180 INFO [train.py:467] (3/4) Epoch 711, batch 30, global_batch_idx: 26300, batch size: 85, loss[discriminator_loss=2.574, discriminator_real_loss=1.261, discriminator_fake_loss=1.312, generator_loss=30.7, generator_mel_loss=20.38, generator_kl_loss=1.902, generator_dur_loss=1.637, generator_adv_loss=2.258, generator_feat_match_loss=4.523, over 85.00 samples.], tot_loss[discriminator_loss=2.582, discriminator_real_loss=1.31, discriminator_fake_loss=1.272, generator_loss=30.66, generator_mel_loss=20.42, generator_kl_loss=2.027, generator_dur_loss=1.642, generator_adv_loss=2.198, generator_feat_match_loss=4.377, over 2492.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:15:56,958 INFO [train.py:811] (3/4) Start epoch 712
2023-11-14 15:19:27,518 INFO [train.py:811] (3/4) Start epoch 713
2023-11-14 15:20:19,645 INFO [train.py:467] (3/4) Epoch 713, batch 6, global_batch_idx: 26350, batch size: 69, loss[discriminator_loss=2.617, discriminator_real_loss=1.353, discriminator_fake_loss=1.265, generator_loss=30.86, generator_mel_loss=20.65, generator_kl_loss=2.138, generator_dur_loss=1.657, generator_adv_loss=2.16, generator_feat_match_loss=4.254, over 69.00 samples.], tot_loss[discriminator_loss=2.557, discriminator_real_loss=1.302, discriminator_fake_loss=1.255, generator_loss=30.55, generator_mel_loss=20.21, generator_kl_loss=2.052, generator_dur_loss=1.646, generator_adv_loss=2.202, generator_feat_match_loss=4.436, over 460.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:23:03,217 INFO [train.py:811] (3/4) Start epoch 714
2023-11-14 15:25:03,863 INFO [train.py:467] (3/4) Epoch 714, batch 19, global_batch_idx: 26400, batch size: 49, loss[discriminator_loss=2.562, discriminator_real_loss=1.324, discriminator_fake_loss=1.237, generator_loss=30.52, generator_mel_loss=19.94, generator_kl_loss=2.033, generator_dur_loss=1.638, generator_adv_loss=2.168, generator_feat_match_loss=4.738, over 49.00 samples.], tot_loss[discriminator_loss=2.587, discriminator_real_loss=1.304, discriminator_fake_loss=1.283, generator_loss=30.57, generator_mel_loss=20.39, generator_kl_loss=2.041, generator_dur_loss=1.645, generator_adv_loss=2.182, generator_feat_match_loss=4.311, over 1366.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 15:25:03,864 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 15:25:15,256 INFO [train.py:517] (3/4) Epoch 714, validation: discriminator_loss=2.513, discriminator_real_loss=1.094, discriminator_fake_loss=1.419, generator_loss=31.62, generator_mel_loss=21.33, generator_kl_loss=2.166, generator_dur_loss=1.632, generator_adv_loss=1.889, generator_feat_match_loss=4.603, over 100.00 samples.
2023-11-14 15:25:15,257 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 15:26:49,378 INFO [train.py:811] (3/4) Start epoch 715
2023-11-14 15:30:04,967 INFO [train.py:467] (3/4) Epoch 715, batch 32, global_batch_idx: 26450, batch size: 56, loss[discriminator_loss=2.551, discriminator_real_loss=1.204, discriminator_fake_loss=1.346, generator_loss=30.25, generator_mel_loss=19.67, generator_kl_loss=2.002, generator_dur_loss=1.633, generator_adv_loss=2.332, generator_feat_match_loss=4.609, over 56.00 samples.], tot_loss[discriminator_loss=2.556, discriminator_real_loss=1.299, discriminator_fake_loss=1.257, generator_loss=30.4, generator_mel_loss=20.06, generator_kl_loss=2.011, generator_dur_loss=1.642, generator_adv_loss=2.245, generator_feat_match_loss=4.442, over 2357.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:30:22,288 INFO [train.py:811] (3/4) Start epoch 716
2023-11-14 15:33:51,595 INFO [train.py:811] (3/4) Start epoch 717
2023-11-14 15:34:46,931 INFO [train.py:467] (3/4) Epoch 717, batch 8, global_batch_idx: 26500, batch size: 56, loss[discriminator_loss=2.574, discriminator_real_loss=1.362, discriminator_fake_loss=1.211, generator_loss=30.18, generator_mel_loss=19.99, generator_kl_loss=1.931, generator_dur_loss=1.623, generator_adv_loss=2.379, generator_feat_match_loss=4.258, over 56.00 samples.], tot_loss[discriminator_loss=2.525, discriminator_real_loss=1.297, discriminator_fake_loss=1.228, generator_loss=30.4, generator_mel_loss=20.1, generator_kl_loss=1.976, generator_dur_loss=1.632, generator_adv_loss=2.301, generator_feat_match_loss=4.389, over 609.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:37:21,461 INFO [train.py:811] (3/4) Start epoch 718
2023-11-14 15:39:34,792 INFO [train.py:467] (3/4) Epoch 718, batch 21, global_batch_idx: 26550, batch size: 101, loss[discriminator_loss=2.512, discriminator_real_loss=1.326, discriminator_fake_loss=1.187, generator_loss=30.13, generator_mel_loss=20.07, generator_kl_loss=2.089, generator_dur_loss=1.646, generator_adv_loss=2.105, generator_feat_match_loss=4.227, over 101.00 samples.], tot_loss[discriminator_loss=2.589, discriminator_real_loss=1.32, discriminator_fake_loss=1.268, generator_loss=30.46, generator_mel_loss=20.19, generator_kl_loss=2.035, generator_dur_loss=1.643, generator_adv_loss=2.228, generator_feat_match_loss=4.36, over 1564.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:40:53,933 INFO [train.py:811] (3/4) Start epoch 719
2023-11-14 15:44:15,700 INFO [train.py:467] (3/4) Epoch 719, batch 34, global_batch_idx: 26600, batch size: 81, loss[discriminator_loss=2.762, discriminator_real_loss=1.396, discriminator_fake_loss=1.367, generator_loss=29.13, generator_mel_loss=19.5, generator_kl_loss=2.04, generator_dur_loss=1.636, generator_adv_loss=2.059, generator_feat_match_loss=3.898, over 81.00 samples.], tot_loss[discriminator_loss=2.51, discriminator_real_loss=1.256, discriminator_fake_loss=1.255, generator_loss=30.79, generator_mel_loss=20.03, generator_kl_loss=2.012, generator_dur_loss=1.642, generator_adv_loss=2.367, generator_feat_match_loss=4.732, over 2675.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:44:15,702 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 15:44:26,715 INFO [train.py:517] (3/4) Epoch 719, validation: discriminator_loss=2.578, discriminator_real_loss=1.256, discriminator_fake_loss=1.322, generator_loss=30.62, generator_mel_loss=20.52, generator_kl_loss=2.137, generator_dur_loss=1.636, generator_adv_loss=2.032, generator_feat_match_loss=4.302, over 100.00 samples.
2023-11-14 15:44:26,716 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 15:44:37,025 INFO [train.py:811] (3/4) Start epoch 720
2023-11-14 15:48:07,763 INFO [train.py:811] (3/4) Start epoch 721
2023-11-14 15:49:16,261 INFO [train.py:467] (3/4) Epoch 721, batch 10, global_batch_idx: 26650, batch size: 64, loss[discriminator_loss=2.629, discriminator_real_loss=1.327, discriminator_fake_loss=1.303, generator_loss=29.74, generator_mel_loss=19.95, generator_kl_loss=2.049, generator_dur_loss=1.658, generator_adv_loss=2.104, generator_feat_match_loss=3.982, over 64.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.274, discriminator_fake_loss=1.266, generator_loss=30.12, generator_mel_loss=20.05, generator_kl_loss=2.006, generator_dur_loss=1.645, generator_adv_loss=2.148, generator_feat_match_loss=4.267, over 819.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:51:34,798 INFO [train.py:811] (3/4) Start epoch 722
2023-11-14 15:53:58,847 INFO [train.py:467] (3/4) Epoch 722, batch 23, global_batch_idx: 26700, batch size: 90, loss[discriminator_loss=2.5, discriminator_real_loss=1.336, discriminator_fake_loss=1.165, generator_loss=30.85, generator_mel_loss=20.19, generator_kl_loss=2.077, generator_dur_loss=1.634, generator_adv_loss=2.264, generator_feat_match_loss=4.68, over 90.00 samples.], tot_loss[discriminator_loss=2.594, discriminator_real_loss=1.315, discriminator_fake_loss=1.28, generator_loss=30.61, generator_mel_loss=20.28, generator_kl_loss=2.008, generator_dur_loss=1.639, generator_adv_loss=2.249, generator_feat_match_loss=4.434, over 1702.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:55:07,104 INFO [train.py:811] (3/4) Start epoch 723
2023-11-14 15:58:43,661 INFO [train.py:467] (3/4) Epoch 723, batch 36, global_batch_idx: 26750, batch size: 85, loss[discriminator_loss=2.609, discriminator_real_loss=1.354, discriminator_fake_loss=1.256, generator_loss=31.08, generator_mel_loss=20.17, generator_kl_loss=2.045, generator_dur_loss=1.647, generator_adv_loss=2.582, generator_feat_match_loss=4.633, over 85.00 samples.], tot_loss[discriminator_loss=2.571, discriminator_real_loss=1.306, discriminator_fake_loss=1.265, generator_loss=30.49, generator_mel_loss=20.22, generator_kl_loss=2.043, generator_dur_loss=1.645, generator_adv_loss=2.233, generator_feat_match_loss=4.352, over 2759.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 15:58:44,354 INFO [train.py:811] (3/4) Start epoch 724
2023-11-14 16:02:17,729 INFO [train.py:811] (3/4) Start epoch 725
2023-11-14 16:03:36,468 INFO [train.py:467] (3/4) Epoch 725, batch 12, global_batch_idx: 26800, batch size: 50, loss[discriminator_loss=2.547, discriminator_real_loss=1.325, discriminator_fake_loss=1.222, generator_loss=29.83, generator_mel_loss=19.89, generator_kl_loss=1.964, generator_dur_loss=1.597, generator_adv_loss=2.229, generator_feat_match_loss=4.148, over 50.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.311, discriminator_fake_loss=1.27, generator_loss=30.29, generator_mel_loss=20.14, generator_kl_loss=2.017, generator_dur_loss=1.635, generator_adv_loss=2.2, generator_feat_match_loss=4.301, over 957.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 16.0
2023-11-14 16:03:36,469 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 16:03:47,692 INFO [train.py:517] (3/4) Epoch 725, validation: discriminator_loss=2.553, discriminator_real_loss=1.265, discriminator_fake_loss=1.288, generator_loss=30.26, generator_mel_loss=20.49, generator_kl_loss=2.119, generator_dur_loss=1.638, generator_adv_loss=2.055, generator_feat_match_loss=3.957, over 100.00 samples.
2023-11-14 16:03:47,693 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 16:06:02,843 INFO [train.py:811] (3/4) Start epoch 726
2023-11-14 16:08:30,036 INFO [train.py:467] (3/4) Epoch 726, batch 25, global_batch_idx: 26850, batch size: 81, loss[discriminator_loss=2.531, discriminator_real_loss=1.302, discriminator_fake_loss=1.23, generator_loss=30.52, generator_mel_loss=19.85, generator_kl_loss=2.145, generator_dur_loss=1.655, generator_adv_loss=2.107, generator_feat_match_loss=4.766, over 81.00 samples.], tot_loss[discriminator_loss=2.58, discriminator_real_loss=1.306, discriminator_fake_loss=1.275, generator_loss=30.5, generator_mel_loss=20.18, generator_kl_loss=2.023, generator_dur_loss=1.647, generator_adv_loss=2.221, generator_feat_match_loss=4.424, over 1891.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:09:33,798 INFO [train.py:811] (3/4) Start epoch 727
2023-11-14 16:13:05,214 INFO [train.py:811] (3/4) Start epoch 728
2023-11-14 16:13:28,040 INFO [train.py:467] (3/4) Epoch 728, batch 1, global_batch_idx: 26900, batch size: 50, loss[discriminator_loss=2.578, discriminator_real_loss=1.279, discriminator_fake_loss=1.299, generator_loss=29.78, generator_mel_loss=19.57, generator_kl_loss=2.058, generator_dur_loss=1.66, generator_adv_loss=2.311, generator_feat_match_loss=4.172, over 50.00 samples.], tot_loss[discriminator_loss=2.528, discriminator_real_loss=1.239, discriminator_fake_loss=1.29, generator_loss=29.91, generator_mel_loss=19.68, generator_kl_loss=2.05, generator_dur_loss=1.647, generator_adv_loss=2.28, generator_feat_match_loss=4.255, over 113.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:16:38,002 INFO [train.py:811] (3/4) Start epoch 729
2023-11-14 16:18:11,095 INFO [train.py:467] (3/4) Epoch 729, batch 14, global_batch_idx: 26950, batch size: 95, loss[discriminator_loss=2.711, discriminator_real_loss=1.39, discriminator_fake_loss=1.32, generator_loss=30.03, generator_mel_loss=20.39, generator_kl_loss=2.026, generator_dur_loss=1.649, generator_adv_loss=2.072, generator_feat_match_loss=3.887, over 95.00 samples.], tot_loss[discriminator_loss=2.642, discriminator_real_loss=1.348, discriminator_fake_loss=1.293, generator_loss=30.22, generator_mel_loss=20.3, generator_kl_loss=2.021, generator_dur_loss=1.647, generator_adv_loss=2.152, generator_feat_match_loss=4.101, over 948.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:20:13,646 INFO [train.py:811] (3/4) Start epoch 730
2023-11-14 16:22:59,325 INFO [train.py:467] (3/4) Epoch 730, batch 27, global_batch_idx: 27000, batch size: 63, loss[discriminator_loss=2.602, discriminator_real_loss=1.385, discriminator_fake_loss=1.217, generator_loss=29.59, generator_mel_loss=19.81, generator_kl_loss=1.999, generator_dur_loss=1.665, generator_adv_loss=2.074, generator_feat_match_loss=4.043, over 63.00 samples.], tot_loss[discriminator_loss=2.642, discriminator_real_loss=1.336, discriminator_fake_loss=1.306, generator_loss=30.4, generator_mel_loss=20.32, generator_kl_loss=2.029, generator_dur_loss=1.642, generator_adv_loss=2.153, generator_feat_match_loss=4.253, over 2068.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:22:59,327 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 16:23:10,799 INFO [train.py:517] (3/4) Epoch 730, validation: discriminator_loss=2.568, discriminator_real_loss=1.117, discriminator_fake_loss=1.45, generator_loss=30.82, generator_mel_loss=20.99, generator_kl_loss=2.167, generator_dur_loss=1.636, generator_adv_loss=1.875, generator_feat_match_loss=4.148, over 100.00 samples.
2023-11-14 16:23:10,801 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 16:23:59,561 INFO [train.py:811] (3/4) Start epoch 731
2023-11-14 16:27:30,214 INFO [train.py:811] (3/4) Start epoch 732
2023-11-14 16:28:04,146 INFO [train.py:467] (3/4) Epoch 732, batch 3, global_batch_idx: 27050, batch size: 60, loss[discriminator_loss=2.65, discriminator_real_loss=1.319, discriminator_fake_loss=1.331, generator_loss=30.88, generator_mel_loss=20.78, generator_kl_loss=2.038, generator_dur_loss=1.654, generator_adv_loss=2.162, generator_feat_match_loss=4.242, over 60.00 samples.], tot_loss[discriminator_loss=2.652, discriminator_real_loss=1.367, discriminator_fake_loss=1.286, generator_loss=30.6, generator_mel_loss=20.44, generator_kl_loss=1.999, generator_dur_loss=1.642, generator_adv_loss=2.207, generator_feat_match_loss=4.312, over 330.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:30:57,335 INFO [train.py:811] (3/4) Start epoch 733
2023-11-14 16:32:39,192 INFO [train.py:467] (3/4) Epoch 733, batch 16, global_batch_idx: 27100, batch size: 52, loss[discriminator_loss=2.57, discriminator_real_loss=1.477, discriminator_fake_loss=1.094, generator_loss=30.63, generator_mel_loss=20.3, generator_kl_loss=1.962, generator_dur_loss=1.654, generator_adv_loss=2.42, generator_feat_match_loss=4.297, over 52.00 samples.], tot_loss[discriminator_loss=2.569, discriminator_real_loss=1.301, discriminator_fake_loss=1.268, generator_loss=30.59, generator_mel_loss=20.18, generator_kl_loss=2.04, generator_dur_loss=1.646, generator_adv_loss=2.288, generator_feat_match_loss=4.438, over 1170.00 samples.], cur_lr_g: 1.83e-04, cur_lr_d: 1.83e-04, grad_scale: 8.0
2023-11-14 16:34:26,623 INFO [train.py:811] (3/4) Start epoch 734
2023-11-14 16:37:13,378 INFO [train.py:467] (3/4) Epoch 734, batch 29, global_batch_idx: 27150, batch size: 85, loss[discriminator_loss=2.629, discriminator_real_loss=1.263, discriminator_fake_loss=1.365, generator_loss=30.22, generator_mel_loss=20.03, generator_kl_loss=1.982, generator_dur_loss=1.651, generator_adv_loss=1.98, generator_feat_match_loss=4.574, over 85.00 samples.], tot_loss[discriminator_loss=2.655, discriminator_real_loss=1.386, discriminator_fake_loss=1.268, generator_loss=30.55, generator_mel_loss=19.95, generator_kl_loss=1.999, generator_dur_loss=1.638, generator_adv_loss=2.297, generator_feat_match_loss=4.663, over 2181.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 16:37:54,728 INFO [train.py:811] (3/4) Start epoch 735
2023-11-14 16:41:26,424 INFO [train.py:811] (3/4) Start epoch 736
2023-11-14 16:42:14,517 INFO [train.py:467] (3/4) Epoch 736, batch 5, global_batch_idx: 27200, batch size: 73, loss[discriminator_loss=2.539, discriminator_real_loss=1.3, discriminator_fake_loss=1.239, generator_loss=29.99, generator_mel_loss=19.88, generator_kl_loss=2.047, generator_dur_loss=1.646, generator_adv_loss=2.105, generator_feat_match_loss=4.305, over 73.00 samples.], tot_loss[discriminator_loss=2.549, discriminator_real_loss=1.295, discriminator_fake_loss=1.254, generator_loss=30.43, generator_mel_loss=20.21, generator_kl_loss=2.061, generator_dur_loss=1.637, generator_adv_loss=2.189, generator_feat_match_loss=4.337, over 569.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 16:42:14,519 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 16:42:26,535 INFO [train.py:517] (3/4) Epoch 736, validation: discriminator_loss=2.571, discriminator_real_loss=1.193, discriminator_fake_loss=1.377, generator_loss=31.04, generator_mel_loss=20.92, generator_kl_loss=2.167, generator_dur_loss=1.631, generator_adv_loss=1.9, generator_feat_match_loss=4.418, over 100.00 samples.
2023-11-14 16:42:26,536 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 16:45:13,236 INFO [train.py:811] (3/4) Start epoch 737
2023-11-14 16:47:14,279 INFO [train.py:467] (3/4) Epoch 737, batch 18, global_batch_idx: 27250, batch size: 101, loss[discriminator_loss=2.559, discriminator_real_loss=1.227, discriminator_fake_loss=1.331, generator_loss=30.16, generator_mel_loss=20.16, generator_kl_loss=2.003, generator_dur_loss=1.646, generator_adv_loss=2.18, generator_feat_match_loss=4.172, over 101.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.318, discriminator_fake_loss=1.249, generator_loss=30.49, generator_mel_loss=19.93, generator_kl_loss=2.015, generator_dur_loss=1.633, generator_adv_loss=2.351, generator_feat_match_loss=4.56, over 1494.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 16:48:47,133 INFO [train.py:811] (3/4) Start epoch 738
2023-11-14 16:51:48,697 INFO [train.py:467] (3/4) Epoch 738, batch 31, global_batch_idx: 27300, batch size: 101, loss[discriminator_loss=2.602, discriminator_real_loss=1.23, discriminator_fake_loss=1.372, generator_loss=31.05, generator_mel_loss=20.5, generator_kl_loss=2.072, generator_dur_loss=1.629, generator_adv_loss=2.344, generator_feat_match_loss=4.504, over 101.00 samples.], tot_loss[discriminator_loss=2.579, discriminator_real_loss=1.304, discriminator_fake_loss=1.275, generator_loss=30.49, generator_mel_loss=20.26, generator_kl_loss=2.047, generator_dur_loss=1.64, generator_adv_loss=2.198, generator_feat_match_loss=4.346, over 2441.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 16:52:16,142 INFO [train.py:811] (3/4) Start epoch 739
2023-11-14 16:55:49,376 INFO [train.py:811] (3/4) Start epoch 740
2023-11-14 16:56:38,571 INFO [train.py:467] (3/4) Epoch 740, batch 7, global_batch_idx: 27350, batch size: 49, loss[discriminator_loss=2.566, discriminator_real_loss=1.327, discriminator_fake_loss=1.24, generator_loss=30.42, generator_mel_loss=20.21, generator_kl_loss=1.931, generator_dur_loss=1.65, generator_adv_loss=2.195, generator_feat_match_loss=4.438, over 49.00 samples.], tot_loss[discriminator_loss=2.617, discriminator_real_loss=1.296, discriminator_fake_loss=1.321, generator_loss=29.98, generator_mel_loss=20.1, generator_kl_loss=2.009, generator_dur_loss=1.644, generator_adv_loss=2.1, generator_feat_match_loss=4.13, over 465.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 16:59:21,982 INFO [train.py:811] (3/4) Start epoch 741
2023-11-14 17:01:30,368 INFO [train.py:467] (3/4) Epoch 741, batch 20, global_batch_idx: 27400, batch size: 50, loss[discriminator_loss=2.578, discriminator_real_loss=1.312, discriminator_fake_loss=1.266, generator_loss=30.55, generator_mel_loss=19.97, generator_kl_loss=2.008, generator_dur_loss=1.626, generator_adv_loss=2.385, generator_feat_match_loss=4.555, over 50.00 samples.], tot_loss[discriminator_loss=2.536, discriminator_real_loss=1.286, discriminator_fake_loss=1.251, generator_loss=30.53, generator_mel_loss=20.16, generator_kl_loss=2.031, generator_dur_loss=1.647, generator_adv_loss=2.236, generator_feat_match_loss=4.45, over 1377.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:01:30,369 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 17:01:41,521 INFO [train.py:517] (3/4) Epoch 741, validation: discriminator_loss=2.556, discriminator_real_loss=1.263, discriminator_fake_loss=1.293, generator_loss=31.45, generator_mel_loss=21.03, generator_kl_loss=2.177, generator_dur_loss=1.634, generator_adv_loss=2.05, generator_feat_match_loss=4.559, over 100.00 samples.
2023-11-14 17:01:41,522 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 17:03:10,654 INFO [train.py:811] (3/4) Start epoch 742
2023-11-14 17:06:26,351 INFO [train.py:467] (3/4) Epoch 742, batch 33, global_batch_idx: 27450, batch size: 58, loss[discriminator_loss=2.57, discriminator_real_loss=1.31, discriminator_fake_loss=1.261, generator_loss=30.14, generator_mel_loss=20.07, generator_kl_loss=1.947, generator_dur_loss=1.657, generator_adv_loss=2.07, generator_feat_match_loss=4.391, over 58.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.316, discriminator_fake_loss=1.249, generator_loss=30.4, generator_mel_loss=19.95, generator_kl_loss=2.009, generator_dur_loss=1.639, generator_adv_loss=2.3, generator_feat_match_loss=4.501, over 2276.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:06:44,847 INFO [train.py:811] (3/4) Start epoch 743
2023-11-14 17:10:20,522 INFO [train.py:811] (3/4) Start epoch 744
2023-11-14 17:11:24,085 INFO [train.py:467] (3/4) Epoch 744, batch 9, global_batch_idx: 27500, batch size: 110, loss[discriminator_loss=2.699, discriminator_real_loss=1.489, discriminator_fake_loss=1.211, generator_loss=30.33, generator_mel_loss=20.49, generator_kl_loss=2.025, generator_dur_loss=1.63, generator_adv_loss=2.078, generator_feat_match_loss=4.105, over 110.00 samples.], tot_loss[discriminator_loss=2.621, discriminator_real_loss=1.358, discriminator_fake_loss=1.263, generator_loss=30.22, generator_mel_loss=20.28, generator_kl_loss=2.016, generator_dur_loss=1.644, generator_adv_loss=2.118, generator_feat_match_loss=4.159, over 660.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:13:54,205 INFO [train.py:811] (3/4) Start epoch 745
2023-11-14 17:16:09,419 INFO [train.py:467] (3/4) Epoch 745, batch 22, global_batch_idx: 27550, batch size: 95, loss[discriminator_loss=2.785, discriminator_real_loss=1.339, discriminator_fake_loss=1.447, generator_loss=29.41, generator_mel_loss=20.18, generator_kl_loss=2.011, generator_dur_loss=1.615, generator_adv_loss=1.86, generator_feat_match_loss=3.744, over 95.00 samples.], tot_loss[discriminator_loss=2.644, discriminator_real_loss=1.359, discriminator_fake_loss=1.285, generator_loss=30.69, generator_mel_loss=20.26, generator_kl_loss=2.003, generator_dur_loss=1.643, generator_adv_loss=2.29, generator_feat_match_loss=4.497, over 1606.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:17:23,398 INFO [train.py:811] (3/4) Start epoch 746
2023-11-14 17:20:45,048 INFO [train.py:467] (3/4) Epoch 746, batch 35, global_batch_idx: 27600, batch size: 52, loss[discriminator_loss=2.512, discriminator_real_loss=1.301, discriminator_fake_loss=1.212, generator_loss=30.52, generator_mel_loss=20.15, generator_kl_loss=2.035, generator_dur_loss=1.646, generator_adv_loss=2.277, generator_feat_match_loss=4.414, over 52.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.305, discriminator_fake_loss=1.273, generator_loss=30.51, generator_mel_loss=20.21, generator_kl_loss=2.071, generator_dur_loss=1.634, generator_adv_loss=2.193, generator_feat_match_loss=4.399, over 2932.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:20:45,049 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 17:20:56,459 INFO [train.py:517] (3/4) Epoch 746, validation: discriminator_loss=2.5, discriminator_real_loss=1.217, discriminator_fake_loss=1.283, generator_loss=30.84, generator_mel_loss=20.69, generator_kl_loss=2.212, generator_dur_loss=1.631, generator_adv_loss=2.049, generator_feat_match_loss=4.261, over 100.00 samples.
2023-11-14 17:20:56,459 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 17:21:03,737 INFO [train.py:811] (3/4) Start epoch 747
2023-11-14 17:24:28,950 INFO [train.py:811] (3/4) Start epoch 748
2023-11-14 17:25:48,707 INFO [train.py:467] (3/4) Epoch 748, batch 11, global_batch_idx: 27650, batch size: 63, loss[discriminator_loss=2.613, discriminator_real_loss=1.186, discriminator_fake_loss=1.428, generator_loss=29.91, generator_mel_loss=19.8, generator_kl_loss=1.967, generator_dur_loss=1.643, generator_adv_loss=2.188, generator_feat_match_loss=4.309, over 63.00 samples.], tot_loss[discriminator_loss=2.577, discriminator_real_loss=1.304, discriminator_fake_loss=1.273, generator_loss=30.36, generator_mel_loss=20.03, generator_kl_loss=1.999, generator_dur_loss=1.639, generator_adv_loss=2.218, generator_feat_match_loss=4.478, over 896.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:28:05,998 INFO [train.py:811] (3/4) Start epoch 749
2023-11-14 17:30:37,283 INFO [train.py:467] (3/4) Epoch 749, batch 24, global_batch_idx: 27700, batch size: 64, loss[discriminator_loss=2.508, discriminator_real_loss=1.403, discriminator_fake_loss=1.104, generator_loss=30.12, generator_mel_loss=19.7, generator_kl_loss=1.952, generator_dur_loss=1.622, generator_adv_loss=2.336, generator_feat_match_loss=4.512, over 64.00 samples.], tot_loss[discriminator_loss=2.559, discriminator_real_loss=1.3, discriminator_fake_loss=1.259, generator_loss=30.39, generator_mel_loss=20.01, generator_kl_loss=2.011, generator_dur_loss=1.643, generator_adv_loss=2.27, generator_feat_match_loss=4.456, over 1638.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:31:40,091 INFO [train.py:811] (3/4) Start epoch 750
2023-11-14 17:35:12,258 INFO [train.py:811] (3/4) Start epoch 751
2023-11-14 17:35:29,024 INFO [train.py:467] (3/4) Epoch 751, batch 0, global_batch_idx: 27750, batch size: 81, loss[discriminator_loss=2.469, discriminator_real_loss=1.341, discriminator_fake_loss=1.129, generator_loss=30.83, generator_mel_loss=20.21, generator_kl_loss=1.919, generator_dur_loss=1.628, generator_adv_loss=2.305, generator_feat_match_loss=4.77, over 81.00 samples.], tot_loss[discriminator_loss=2.469, discriminator_real_loss=1.341, discriminator_fake_loss=1.129, generator_loss=30.83, generator_mel_loss=20.21, generator_kl_loss=1.919, generator_dur_loss=1.628, generator_adv_loss=2.305, generator_feat_match_loss=4.77, over 81.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:38:41,331 INFO [train.py:811] (3/4) Start epoch 752
2023-11-14 17:40:07,590 INFO [train.py:467] (3/4) Epoch 752, batch 13, global_batch_idx: 27800, batch size: 73, loss[discriminator_loss=2.676, discriminator_real_loss=1.35, discriminator_fake_loss=1.325, generator_loss=30.13, generator_mel_loss=20.22, generator_kl_loss=2.059, generator_dur_loss=1.641, generator_adv_loss=2.135, generator_feat_match_loss=4.082, over 73.00 samples.], tot_loss[discriminator_loss=2.552, discriminator_real_loss=1.278, discriminator_fake_loss=1.274, generator_loss=30.8, generator_mel_loss=20.35, generator_kl_loss=2.028, generator_dur_loss=1.637, generator_adv_loss=2.253, generator_feat_match_loss=4.532, over 1074.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:40:07,592 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 17:40:18,589 INFO [train.py:517] (3/4) Epoch 752, validation: discriminator_loss=2.55, discriminator_real_loss=1.234, discriminator_fake_loss=1.316, generator_loss=31.17, generator_mel_loss=20.9, generator_kl_loss=2.111, generator_dur_loss=1.63, generator_adv_loss=2.089, generator_feat_match_loss=4.446, over 100.00 samples.
2023-11-14 17:40:18,590 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 17:42:25,986 INFO [train.py:811] (3/4) Start epoch 753
2023-11-14 17:45:01,976 INFO [train.py:467] (3/4) Epoch 753, batch 26, global_batch_idx: 27850, batch size: 52, loss[discriminator_loss=2.588, discriminator_real_loss=1.428, discriminator_fake_loss=1.16, generator_loss=29.56, generator_mel_loss=19.84, generator_kl_loss=1.943, generator_dur_loss=1.632, generator_adv_loss=2.152, generator_feat_match_loss=3.99, over 52.00 samples.], tot_loss[discriminator_loss=2.554, discriminator_real_loss=1.298, discriminator_fake_loss=1.255, generator_loss=30.54, generator_mel_loss=20.06, generator_kl_loss=2.027, generator_dur_loss=1.634, generator_adv_loss=2.254, generator_feat_match_loss=4.565, over 2052.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:45:57,994 INFO [train.py:811] (3/4) Start epoch 754
2023-11-14 17:49:36,483 INFO [train.py:811] (3/4) Start epoch 755
2023-11-14 17:50:05,007 INFO [train.py:467] (3/4) Epoch 755, batch 2, global_batch_idx: 27900, batch size: 153, loss[discriminator_loss=2.639, discriminator_real_loss=1.461, discriminator_fake_loss=1.178, generator_loss=30.7, generator_mel_loss=20.28, generator_kl_loss=2.019, generator_dur_loss=1.619, generator_adv_loss=2.219, generator_feat_match_loss=4.57, over 153.00 samples.], tot_loss[discriminator_loss=2.624, discriminator_real_loss=1.431, discriminator_fake_loss=1.193, generator_loss=30.64, generator_mel_loss=20.24, generator_kl_loss=2.03, generator_dur_loss=1.631, generator_adv_loss=2.303, generator_feat_match_loss=4.432, over 293.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:53:10,085 INFO [train.py:811] (3/4) Start epoch 756
2023-11-14 17:54:47,929 INFO [train.py:467] (3/4) Epoch 756, batch 15, global_batch_idx: 27950, batch size: 52, loss[discriminator_loss=2.705, discriminator_real_loss=1.316, discriminator_fake_loss=1.389, generator_loss=29.85, generator_mel_loss=20.32, generator_kl_loss=1.949, generator_dur_loss=1.657, generator_adv_loss=2.062, generator_feat_match_loss=3.869, over 52.00 samples.], tot_loss[discriminator_loss=2.515, discriminator_real_loss=1.255, discriminator_fake_loss=1.26, generator_loss=30.72, generator_mel_loss=20.08, generator_kl_loss=2.004, generator_dur_loss=1.634, generator_adv_loss=2.33, generator_feat_match_loss=4.679, over 1148.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 17:56:40,955 INFO [train.py:811] (3/4) Start epoch 757
2023-11-14 17:59:26,265 INFO [train.py:467] (3/4) Epoch 757, batch 28, global_batch_idx: 28000, batch size: 85, loss[discriminator_loss=2.613, discriminator_real_loss=1.312, discriminator_fake_loss=1.301, generator_loss=30.18, generator_mel_loss=20.08, generator_kl_loss=2.007, generator_dur_loss=1.651, generator_adv_loss=2.236, generator_feat_match_loss=4.207, over 85.00 samples.], tot_loss[discriminator_loss=2.548, discriminator_real_loss=1.29, discriminator_fake_loss=1.258, generator_loss=30.42, generator_mel_loss=20.08, generator_kl_loss=2.03, generator_dur_loss=1.639, generator_adv_loss=2.202, generator_feat_match_loss=4.467, over 2215.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 17:59:26,267 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 17:59:37,436 INFO [train.py:517] (3/4) Epoch 757, validation: discriminator_loss=2.594, discriminator_real_loss=1.328, discriminator_fake_loss=1.265, generator_loss=30.85, generator_mel_loss=20.74, generator_kl_loss=2.225, generator_dur_loss=1.632, generator_adv_loss=2.084, generator_feat_match_loss=4.173, over 100.00 samples.
2023-11-14 17:59:37,437 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 18:00:27,653 INFO [train.py:811] (3/4) Start epoch 758
2023-11-14 18:03:56,886 INFO [train.py:811] (3/4) Start epoch 759
2023-11-14 18:04:33,810 INFO [train.py:467] (3/4) Epoch 759, batch 4, global_batch_idx: 28050, batch size: 67, loss[discriminator_loss=2.818, discriminator_real_loss=1.388, discriminator_fake_loss=1.431, generator_loss=29.98, generator_mel_loss=20.26, generator_kl_loss=2.06, generator_dur_loss=1.652, generator_adv_loss=1.996, generator_feat_match_loss=4.02, over 67.00 samples.], tot_loss[discriminator_loss=2.728, discriminator_real_loss=1.376, discriminator_fake_loss=1.352, generator_loss=30.38, generator_mel_loss=20.42, generator_kl_loss=1.988, generator_dur_loss=1.638, generator_adv_loss=2.145, generator_feat_match_loss=4.184, over 306.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 18:07:31,562 INFO [train.py:811] (3/4) Start epoch 760
2023-11-14 18:09:31,067 INFO [train.py:467] (3/4) Epoch 760, batch 17, global_batch_idx: 28100, batch size: 85, loss[discriminator_loss=2.604, discriminator_real_loss=1.363, discriminator_fake_loss=1.24, generator_loss=30.54, generator_mel_loss=19.91, generator_kl_loss=2.048, generator_dur_loss=1.613, generator_adv_loss=2.285, generator_feat_match_loss=4.688, over 85.00 samples.], tot_loss[discriminator_loss=2.588, discriminator_real_loss=1.313, discriminator_fake_loss=1.275, generator_loss=30.37, generator_mel_loss=20.08, generator_kl_loss=2.005, generator_dur_loss=1.631, generator_adv_loss=2.226, generator_feat_match_loss=4.427, over 1456.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:11:07,402 INFO [train.py:811] (3/4) Start epoch 761
2023-11-14 18:14:03,896 INFO [train.py:467] (3/4) Epoch 761, batch 30, global_batch_idx: 28150, batch size: 110, loss[discriminator_loss=2.473, discriminator_real_loss=1.259, discriminator_fake_loss=1.214, generator_loss=30.66, generator_mel_loss=20, generator_kl_loss=2.018, generator_dur_loss=1.597, generator_adv_loss=2.336, generator_feat_match_loss=4.715, over 110.00 samples.], tot_loss[discriminator_loss=2.518, discriminator_real_loss=1.283, discriminator_fake_loss=1.235, generator_loss=30.54, generator_mel_loss=20, generator_kl_loss=2.012, generator_dur_loss=1.637, generator_adv_loss=2.286, generator_feat_match_loss=4.602, over 2447.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:14:41,823 INFO [train.py:811] (3/4) Start epoch 762
2023-11-14 18:18:15,760 INFO [train.py:811] (3/4) Start epoch 763
2023-11-14 18:19:10,712 INFO [train.py:467] (3/4) Epoch 763, batch 6, global_batch_idx: 28200, batch size: 71, loss[discriminator_loss=2.578, discriminator_real_loss=1.392, discriminator_fake_loss=1.186, generator_loss=30.02, generator_mel_loss=19.49, generator_kl_loss=2.05, generator_dur_loss=1.633, generator_adv_loss=2.324, generator_feat_match_loss=4.523, over 71.00 samples.], tot_loss[discriminator_loss=2.51, discriminator_real_loss=1.244, discriminator_fake_loss=1.266, generator_loss=30.41, generator_mel_loss=19.8, generator_kl_loss=2.051, generator_dur_loss=1.63, generator_adv_loss=2.252, generator_feat_match_loss=4.678, over 580.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:19:10,714 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 18:19:22,746 INFO [train.py:517] (3/4) Epoch 763, validation: discriminator_loss=2.437, discriminator_real_loss=1.185, discriminator_fake_loss=1.252, generator_loss=31.03, generator_mel_loss=20.34, generator_kl_loss=2.192, generator_dur_loss=1.639, generator_adv_loss=2.205, generator_feat_match_loss=4.655, over 100.00 samples.
2023-11-14 18:19:22,748 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 18:22:02,790 INFO [train.py:811] (3/4) Start epoch 764
2023-11-14 18:23:56,331 INFO [train.py:467] (3/4) Epoch 764, batch 19, global_batch_idx: 28250, batch size: 95, loss[discriminator_loss=2.646, discriminator_real_loss=1.402, discriminator_fake_loss=1.244, generator_loss=30.82, generator_mel_loss=20.45, generator_kl_loss=2.077, generator_dur_loss=1.607, generator_adv_loss=2.203, generator_feat_match_loss=4.48, over 95.00 samples.], tot_loss[discriminator_loss=2.582, discriminator_real_loss=1.308, discriminator_fake_loss=1.273, generator_loss=30.58, generator_mel_loss=20.25, generator_kl_loss=2.038, generator_dur_loss=1.638, generator_adv_loss=2.217, generator_feat_match_loss=4.433, over 1471.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:25:32,431 INFO [train.py:811] (3/4) Start epoch 765
2023-11-14 18:28:42,634 INFO [train.py:467] (3/4) Epoch 765, batch 32, global_batch_idx: 28300, batch size: 60, loss[discriminator_loss=2.561, discriminator_real_loss=1.322, discriminator_fake_loss=1.238, generator_loss=30.71, generator_mel_loss=20.21, generator_kl_loss=2.091, generator_dur_loss=1.659, generator_adv_loss=2.367, generator_feat_match_loss=4.379, over 60.00 samples.], tot_loss[discriminator_loss=2.599, discriminator_real_loss=1.325, discriminator_fake_loss=1.273, generator_loss=30.62, generator_mel_loss=20.25, generator_kl_loss=2.037, generator_dur_loss=1.639, generator_adv_loss=2.247, generator_feat_match_loss=4.446, over 2376.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:29:01,077 INFO [train.py:811] (3/4) Start epoch 766
2023-11-14 18:32:25,214 INFO [train.py:811] (3/4) Start epoch 767
2023-11-14 18:33:20,979 INFO [train.py:467] (3/4) Epoch 767, batch 8, global_batch_idx: 28350, batch size: 126, loss[discriminator_loss=2.559, discriminator_real_loss=1.229, discriminator_fake_loss=1.329, generator_loss=31.34, generator_mel_loss=20.6, generator_kl_loss=1.949, generator_dur_loss=1.617, generator_adv_loss=2.398, generator_feat_match_loss=4.777, over 126.00 samples.], tot_loss[discriminator_loss=2.561, discriminator_real_loss=1.283, discriminator_fake_loss=1.278, generator_loss=30.63, generator_mel_loss=20.22, generator_kl_loss=2.023, generator_dur_loss=1.635, generator_adv_loss=2.224, generator_feat_match_loss=4.535, over 651.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:35:56,200 INFO [train.py:811] (3/4) Start epoch 768
2023-11-14 18:38:02,875 INFO [train.py:467] (3/4) Epoch 768, batch 21, global_batch_idx: 28400, batch size: 79, loss[discriminator_loss=2.633, discriminator_real_loss=1.321, discriminator_fake_loss=1.312, generator_loss=30.63, generator_mel_loss=20.42, generator_kl_loss=2.074, generator_dur_loss=1.652, generator_adv_loss=2.199, generator_feat_match_loss=4.289, over 79.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.3, discriminator_fake_loss=1.281, generator_loss=30.7, generator_mel_loss=20.25, generator_kl_loss=2.024, generator_dur_loss=1.639, generator_adv_loss=2.26, generator_feat_match_loss=4.528, over 1455.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 18:38:02,877 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 18:38:14,259 INFO [train.py:517] (3/4) Epoch 768, validation: discriminator_loss=2.528, discriminator_real_loss=1.15, discriminator_fake_loss=1.378, generator_loss=31.68, generator_mel_loss=20.95, generator_kl_loss=2.279, generator_dur_loss=1.63, generator_adv_loss=2.086, generator_feat_match_loss=4.736, over 100.00 samples.
2023-11-14 18:38:14,260 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 18:39:42,336 INFO [train.py:811] (3/4) Start epoch 769
2023-11-14 18:43:10,887 INFO [train.py:467] (3/4) Epoch 769, batch 34, global_batch_idx: 28450, batch size: 61, loss[discriminator_loss=2.561, discriminator_real_loss=1.244, discriminator_fake_loss=1.316, generator_loss=30.71, generator_mel_loss=20.18, generator_kl_loss=2.081, generator_dur_loss=1.625, generator_adv_loss=2.256, generator_feat_match_loss=4.566, over 61.00 samples.], tot_loss[discriminator_loss=2.534, discriminator_real_loss=1.284, discriminator_fake_loss=1.25, generator_loss=30.54, generator_mel_loss=19.95, generator_kl_loss=2.005, generator_dur_loss=1.637, generator_adv_loss=2.307, generator_feat_match_loss=4.641, over 2444.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 16.0
2023-11-14 18:43:19,961 INFO [train.py:811] (3/4) Start epoch 770
2023-11-14 18:46:55,287 INFO [train.py:811] (3/4) Start epoch 771
2023-11-14 18:48:01,823 INFO [train.py:467] (3/4) Epoch 771, batch 10, global_batch_idx: 28500, batch size: 154, loss[discriminator_loss=2.426, discriminator_real_loss=1.154, discriminator_fake_loss=1.271, generator_loss=31.55, generator_mel_loss=20.54, generator_kl_loss=2.026, generator_dur_loss=1.632, generator_adv_loss=2.273, generator_feat_match_loss=5.078, over 154.00 samples.], tot_loss[discriminator_loss=2.508, discriminator_real_loss=1.258, discriminator_fake_loss=1.25, generator_loss=30.69, generator_mel_loss=20.12, generator_kl_loss=2.057, generator_dur_loss=1.647, generator_adv_loss=2.261, generator_feat_match_loss=4.613, over 713.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:50:31,247 INFO [train.py:811] (3/4) Start epoch 772
2023-11-14 18:52:56,382 INFO [train.py:467] (3/4) Epoch 772, batch 23, global_batch_idx: 28550, batch size: 54, loss[discriminator_loss=2.666, discriminator_real_loss=1.549, discriminator_fake_loss=1.117, generator_loss=29.92, generator_mel_loss=19.63, generator_kl_loss=2.012, generator_dur_loss=1.606, generator_adv_loss=2.318, generator_feat_match_loss=4.352, over 54.00 samples.], tot_loss[discriminator_loss=2.551, discriminator_real_loss=1.294, discriminator_fake_loss=1.258, generator_loss=30.64, generator_mel_loss=19.88, generator_kl_loss=1.999, generator_dur_loss=1.632, generator_adv_loss=2.365, generator_feat_match_loss=4.761, over 1954.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:54:04,227 INFO [train.py:811] (3/4) Start epoch 773
2023-11-14 18:57:42,112 INFO [train.py:467] (3/4) Epoch 773, batch 36, global_batch_idx: 28600, batch size: 67, loss[discriminator_loss=2.617, discriminator_real_loss=1.293, discriminator_fake_loss=1.324, generator_loss=30.52, generator_mel_loss=20.07, generator_kl_loss=2.032, generator_dur_loss=1.65, generator_adv_loss=2.33, generator_feat_match_loss=4.445, over 67.00 samples.], tot_loss[discriminator_loss=2.596, discriminator_real_loss=1.304, discriminator_fake_loss=1.292, generator_loss=30.57, generator_mel_loss=20.32, generator_kl_loss=2.022, generator_dur_loss=1.636, generator_adv_loss=2.182, generator_feat_match_loss=4.41, over 2754.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 18:57:42,114 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 18:57:53,394 INFO [train.py:517] (3/4) Epoch 773, validation: discriminator_loss=2.675, discriminator_real_loss=1.355, discriminator_fake_loss=1.32, generator_loss=30.56, generator_mel_loss=20.63, generator_kl_loss=2.129, generator_dur_loss=1.633, generator_adv_loss=2.041, generator_feat_match_loss=4.135, over 100.00 samples.
2023-11-14 18:57:53,395 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 18:57:54,077 INFO [train.py:811] (3/4) Start epoch 774
2023-11-14 19:01:28,311 INFO [train.py:811] (3/4) Start epoch 775
2023-11-14 19:02:44,117 INFO [train.py:467] (3/4) Epoch 775, batch 12, global_batch_idx: 28650, batch size: 90, loss[discriminator_loss=2.436, discriminator_real_loss=1.257, discriminator_fake_loss=1.179, generator_loss=31.09, generator_mel_loss=20.08, generator_kl_loss=2.068, generator_dur_loss=1.638, generator_adv_loss=2.373, generator_feat_match_loss=4.93, over 90.00 samples.], tot_loss[discriminator_loss=2.566, discriminator_real_loss=1.322, discriminator_fake_loss=1.243, generator_loss=30.47, generator_mel_loss=19.93, generator_kl_loss=1.998, generator_dur_loss=1.649, generator_adv_loss=2.349, generator_feat_match_loss=4.54, over 840.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 19:05:01,598 INFO [train.py:811] (3/4) Start epoch 776
2023-11-14 19:07:36,707 INFO [train.py:467] (3/4) Epoch 776, batch 25, global_batch_idx: 28700, batch size: 55, loss[discriminator_loss=2.609, discriminator_real_loss=1.174, discriminator_fake_loss=1.436, generator_loss=31.3, generator_mel_loss=20.43, generator_kl_loss=2.012, generator_dur_loss=1.645, generator_adv_loss=2.252, generator_feat_match_loss=4.965, over 55.00 samples.], tot_loss[discriminator_loss=2.556, discriminator_real_loss=1.306, discriminator_fake_loss=1.25, generator_loss=30.54, generator_mel_loss=20.04, generator_kl_loss=2.047, generator_dur_loss=1.634, generator_adv_loss=2.271, generator_feat_match_loss=4.547, over 1799.00 samples.], cur_lr_g: 1.82e-04, cur_lr_d: 1.82e-04, grad_scale: 8.0
2023-11-14 19:08:35,855 INFO [train.py:811] (3/4) Start epoch 777
2023-11-14 19:12:11,891 INFO [train.py:811] (3/4) Start epoch 778
2023-11-14 19:12:31,629 INFO [train.py:467] (3/4) Epoch 778, batch 1, global_batch_idx: 28750, batch size: 73, loss[discriminator_loss=2.449, discriminator_real_loss=1.223, discriminator_fake_loss=1.226, generator_loss=31.18, generator_mel_loss=20, generator_kl_loss=2.054, generator_dur_loss=1.646, generator_adv_loss=2.477, generator_feat_match_loss=5.004, over 73.00 samples.], tot_loss[discriminator_loss=2.479, discriminator_real_loss=1.287, discriminator_fake_loss=1.192, generator_loss=30.96, generator_mel_loss=19.99, generator_kl_loss=2.042, generator_dur_loss=1.652, generator_adv_loss=2.423, generator_feat_match_loss=4.858, over 144.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:15:47,952 INFO [train.py:811] (3/4) Start epoch 779
2023-11-14 19:17:21,547 INFO [train.py:467] (3/4) Epoch 779, batch 14, global_batch_idx: 28800, batch size: 56, loss[discriminator_loss=2.387, discriminator_real_loss=1.194, discriminator_fake_loss=1.191, generator_loss=30.78, generator_mel_loss=19.69, generator_kl_loss=1.976, generator_dur_loss=1.663, generator_adv_loss=2.461, generator_feat_match_loss=4.992, over 56.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.277, discriminator_fake_loss=1.263, generator_loss=30.6, generator_mel_loss=19.94, generator_kl_loss=2.01, generator_dur_loss=1.64, generator_adv_loss=2.321, generator_feat_match_loss=4.683, over 1068.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 19:17:21,549 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 19:17:32,666 INFO [train.py:517] (3/4) Epoch 779, validation: discriminator_loss=2.42, discriminator_real_loss=1.15, discriminator_fake_loss=1.27, generator_loss=30.91, generator_mel_loss=20.48, generator_kl_loss=2.196, generator_dur_loss=1.632, generator_adv_loss=2.052, generator_feat_match_loss=4.545, over 100.00 samples.
2023-11-14 19:17:32,667 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 19:19:25,372 INFO [train.py:811] (3/4) Start epoch 780
2023-11-14 19:21:58,796 INFO [train.py:467] (3/4) Epoch 780, batch 27, global_batch_idx: 28850, batch size: 52, loss[discriminator_loss=2.516, discriminator_real_loss=1.28, discriminator_fake_loss=1.236, generator_loss=30.23, generator_mel_loss=20.04, generator_kl_loss=1.98, generator_dur_loss=1.652, generator_adv_loss=2.096, generator_feat_match_loss=4.461, over 52.00 samples.], tot_loss[discriminator_loss=2.497, discriminator_real_loss=1.265, discriminator_fake_loss=1.232, generator_loss=30.78, generator_mel_loss=19.95, generator_kl_loss=2.022, generator_dur_loss=1.636, generator_adv_loss=2.325, generator_feat_match_loss=4.845, over 2039.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:22:53,612 INFO [train.py:811] (3/4) Start epoch 781
2023-11-14 19:26:27,322 INFO [train.py:811] (3/4) Start epoch 782
2023-11-14 19:26:58,514 INFO [train.py:467] (3/4) Epoch 782, batch 3, global_batch_idx: 28900, batch size: 50, loss[discriminator_loss=2.309, discriminator_real_loss=1.193, discriminator_fake_loss=1.116, generator_loss=31.25, generator_mel_loss=19.73, generator_kl_loss=1.959, generator_dur_loss=1.626, generator_adv_loss=2.371, generator_feat_match_loss=5.562, over 50.00 samples.], tot_loss[discriminator_loss=2.456, discriminator_real_loss=1.286, discriminator_fake_loss=1.17, generator_loss=31.55, generator_mel_loss=20.02, generator_kl_loss=2.005, generator_dur_loss=1.638, generator_adv_loss=2.486, generator_feat_match_loss=5.404, over 310.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:29:55,885 INFO [train.py:811] (3/4) Start epoch 783
2023-11-14 19:31:32,902 INFO [train.py:467] (3/4) Epoch 783, batch 16, global_batch_idx: 28950, batch size: 52, loss[discriminator_loss=2.492, discriminator_real_loss=1.247, discriminator_fake_loss=1.245, generator_loss=30.47, generator_mel_loss=20.11, generator_kl_loss=2.019, generator_dur_loss=1.643, generator_adv_loss=2.277, generator_feat_match_loss=4.418, over 52.00 samples.], tot_loss[discriminator_loss=2.572, discriminator_real_loss=1.297, discriminator_fake_loss=1.275, generator_loss=30.26, generator_mel_loss=20.08, generator_kl_loss=2.035, generator_dur_loss=1.635, generator_adv_loss=2.157, generator_feat_match_loss=4.355, over 1219.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:33:27,471 INFO [train.py:811] (3/4) Start epoch 784
2023-11-14 19:36:12,987 INFO [train.py:467] (3/4) Epoch 784, batch 29, global_batch_idx: 29000, batch size: 101, loss[discriminator_loss=2.6, discriminator_real_loss=1.385, discriminator_fake_loss=1.215, generator_loss=30.47, generator_mel_loss=20.13, generator_kl_loss=2.028, generator_dur_loss=1.643, generator_adv_loss=2.154, generator_feat_match_loss=4.512, over 101.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.305, discriminator_fake_loss=1.276, generator_loss=30.52, generator_mel_loss=20.17, generator_kl_loss=2.018, generator_dur_loss=1.634, generator_adv_loss=2.222, generator_feat_match_loss=4.481, over 2440.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:36:12,988 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 19:36:23,999 INFO [train.py:517] (3/4) Epoch 784, validation: discriminator_loss=2.483, discriminator_real_loss=1.163, discriminator_fake_loss=1.32, generator_loss=30.95, generator_mel_loss=20.64, generator_kl_loss=2.146, generator_dur_loss=1.631, generator_adv_loss=1.886, generator_feat_match_loss=4.646, over 100.00 samples.
2023-11-14 19:36:24,001 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 19:37:06,633 INFO [train.py:811] (3/4) Start epoch 785
2023-11-14 19:40:37,305 INFO [train.py:811] (3/4) Start epoch 786
2023-11-14 19:41:19,863 INFO [train.py:467] (3/4) Epoch 786, batch 5, global_batch_idx: 29050, batch size: 55, loss[discriminator_loss=2.643, discriminator_real_loss=1.334, discriminator_fake_loss=1.309, generator_loss=30.78, generator_mel_loss=20.11, generator_kl_loss=2.046, generator_dur_loss=1.653, generator_adv_loss=2.41, generator_feat_match_loss=4.559, over 55.00 samples.], tot_loss[discriminator_loss=2.546, discriminator_real_loss=1.26, discriminator_fake_loss=1.286, generator_loss=30.8, generator_mel_loss=20.16, generator_kl_loss=2.041, generator_dur_loss=1.637, generator_adv_loss=2.261, generator_feat_match_loss=4.702, over 524.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:44:08,643 INFO [train.py:811] (3/4) Start epoch 787
2023-11-14 19:46:00,353 INFO [train.py:467] (3/4) Epoch 787, batch 18, global_batch_idx: 29100, batch size: 52, loss[discriminator_loss=2.467, discriminator_real_loss=1.178, discriminator_fake_loss=1.289, generator_loss=30.93, generator_mel_loss=19.69, generator_kl_loss=2.005, generator_dur_loss=1.625, generator_adv_loss=2.514, generator_feat_match_loss=5.102, over 52.00 samples.], tot_loss[discriminator_loss=2.504, discriminator_real_loss=1.263, discriminator_fake_loss=1.241, generator_loss=30.71, generator_mel_loss=19.91, generator_kl_loss=2.009, generator_dur_loss=1.642, generator_adv_loss=2.365, generator_feat_match_loss=4.776, over 1238.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:47:40,833 INFO [train.py:811] (3/4) Start epoch 788
2023-11-14 19:50:37,918 INFO [train.py:467] (3/4) Epoch 788, batch 31, global_batch_idx: 29150, batch size: 71, loss[discriminator_loss=2.516, discriminator_real_loss=1.255, discriminator_fake_loss=1.262, generator_loss=29.82, generator_mel_loss=19.68, generator_kl_loss=1.948, generator_dur_loss=1.62, generator_adv_loss=2.137, generator_feat_match_loss=4.434, over 71.00 samples.], tot_loss[discriminator_loss=2.503, discriminator_real_loss=1.273, discriminator_fake_loss=1.23, generator_loss=30.74, generator_mel_loss=19.76, generator_kl_loss=2.018, generator_dur_loss=1.637, generator_adv_loss=2.387, generator_feat_match_loss=4.944, over 2158.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 19:51:10,066 INFO [train.py:811] (3/4) Start epoch 789
2023-11-14 19:54:42,965 INFO [train.py:811] (3/4) Start epoch 790
2023-11-14 19:55:32,828 INFO [train.py:467] (3/4) Epoch 790, batch 7, global_batch_idx: 29200, batch size: 85, loss[discriminator_loss=2.609, discriminator_real_loss=1.29, discriminator_fake_loss=1.318, generator_loss=30.41, generator_mel_loss=20.24, generator_kl_loss=2.005, generator_dur_loss=1.618, generator_adv_loss=2.184, generator_feat_match_loss=4.363, over 85.00 samples.], tot_loss[discriminator_loss=2.605, discriminator_real_loss=1.309, discriminator_fake_loss=1.296, generator_loss=30.59, generator_mel_loss=20.33, generator_kl_loss=2.018, generator_dur_loss=1.633, generator_adv_loss=2.178, generator_feat_match_loss=4.436, over 647.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 19:55:32,830 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 19:55:44,845 INFO [train.py:517] (3/4) Epoch 790, validation: discriminator_loss=2.615, discriminator_real_loss=1.313, discriminator_fake_loss=1.301, generator_loss=31.74, generator_mel_loss=21.48, generator_kl_loss=2.282, generator_dur_loss=1.629, generator_adv_loss=1.995, generator_feat_match_loss=4.356, over 100.00 samples.
2023-11-14 19:55:44,846 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 19:58:31,984 INFO [train.py:811] (3/4) Start epoch 791
2023-11-14 20:00:37,026 INFO [train.py:467] (3/4) Epoch 791, batch 20, global_batch_idx: 29250, batch size: 69, loss[discriminator_loss=2.551, discriminator_real_loss=1.167, discriminator_fake_loss=1.383, generator_loss=30.93, generator_mel_loss=20.08, generator_kl_loss=1.976, generator_dur_loss=1.604, generator_adv_loss=2.297, generator_feat_match_loss=4.969, over 69.00 samples.], tot_loss[discriminator_loss=2.613, discriminator_real_loss=1.31, discriminator_fake_loss=1.303, generator_loss=30.58, generator_mel_loss=20.19, generator_kl_loss=2.019, generator_dur_loss=1.635, generator_adv_loss=2.228, generator_feat_match_loss=4.511, over 1570.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 20:02:08,350 INFO [train.py:811] (3/4) Start epoch 792
2023-11-14 20:05:28,728 INFO [train.py:467] (3/4) Epoch 792, batch 33, global_batch_idx: 29300, batch size: 55, loss[discriminator_loss=2.301, discriminator_real_loss=1.123, discriminator_fake_loss=1.178, generator_loss=31.5, generator_mel_loss=19.89, generator_kl_loss=2.002, generator_dur_loss=1.669, generator_adv_loss=2.492, generator_feat_match_loss=5.449, over 55.00 samples.], tot_loss[discriminator_loss=2.488, discriminator_real_loss=1.256, discriminator_fake_loss=1.232, generator_loss=30.67, generator_mel_loss=19.88, generator_kl_loss=2.014, generator_dur_loss=1.637, generator_adv_loss=2.339, generator_feat_match_loss=4.801, over 2316.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:05:46,643 INFO [train.py:811] (3/4) Start epoch 793
2023-11-14 20:09:12,185 INFO [train.py:811] (3/4) Start epoch 794
2023-11-14 20:10:12,377 INFO [train.py:467] (3/4) Epoch 794, batch 9, global_batch_idx: 29350, batch size: 50, loss[discriminator_loss=2.641, discriminator_real_loss=1.192, discriminator_fake_loss=1.449, generator_loss=30.25, generator_mel_loss=20, generator_kl_loss=2.033, generator_dur_loss=1.621, generator_adv_loss=2.07, generator_feat_match_loss=4.531, over 50.00 samples.], tot_loss[discriminator_loss=2.575, discriminator_real_loss=1.29, discriminator_fake_loss=1.285, generator_loss=30.46, generator_mel_loss=20.06, generator_kl_loss=2.018, generator_dur_loss=1.646, generator_adv_loss=2.151, generator_feat_match_loss=4.581, over 617.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:12:45,761 INFO [train.py:811] (3/4) Start epoch 795
2023-11-14 20:15:04,198 INFO [train.py:467] (3/4) Epoch 795, batch 22, global_batch_idx: 29400, batch size: 58, loss[discriminator_loss=2.668, discriminator_real_loss=1.479, discriminator_fake_loss=1.188, generator_loss=30.56, generator_mel_loss=20.26, generator_kl_loss=2.052, generator_dur_loss=1.623, generator_adv_loss=2.18, generator_feat_match_loss=4.453, over 58.00 samples.], tot_loss[discriminator_loss=2.647, discriminator_real_loss=1.348, discriminator_fake_loss=1.298, generator_loss=30.58, generator_mel_loss=20.34, generator_kl_loss=2.076, generator_dur_loss=1.642, generator_adv_loss=2.156, generator_feat_match_loss=4.368, over 1770.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:15:04,200 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 20:15:15,187 INFO [train.py:517] (3/4) Epoch 795, validation: discriminator_loss=2.594, discriminator_real_loss=1.251, discriminator_fake_loss=1.343, generator_loss=30.8, generator_mel_loss=20.47, generator_kl_loss=2.226, generator_dur_loss=1.628, generator_adv_loss=1.974, generator_feat_match_loss=4.503, over 100.00 samples.
2023-11-14 20:15:15,188 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 20:16:30,364 INFO [train.py:811] (3/4) Start epoch 796
2023-11-14 20:20:00,202 INFO [train.py:467] (3/4) Epoch 796, batch 35, global_batch_idx: 29450, batch size: 110, loss[discriminator_loss=2.553, discriminator_real_loss=1.287, discriminator_fake_loss=1.266, generator_loss=30.88, generator_mel_loss=20.44, generator_kl_loss=2.036, generator_dur_loss=1.634, generator_adv_loss=2.211, generator_feat_match_loss=4.562, over 110.00 samples.], tot_loss[discriminator_loss=2.562, discriminator_real_loss=1.307, discriminator_fake_loss=1.255, generator_loss=30.61, generator_mel_loss=20.06, generator_kl_loss=2.012, generator_dur_loss=1.636, generator_adv_loss=2.261, generator_feat_match_loss=4.634, over 2801.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:20:06,780 INFO [train.py:811] (3/4) Start epoch 797
2023-11-14 20:23:40,369 INFO [train.py:811] (3/4) Start epoch 798
2023-11-14 20:24:56,865 INFO [train.py:467] (3/4) Epoch 798, batch 11, global_batch_idx: 29500, batch size: 90, loss[discriminator_loss=2.436, discriminator_real_loss=1.156, discriminator_fake_loss=1.279, generator_loss=31.06, generator_mel_loss=20.07, generator_kl_loss=2.127, generator_dur_loss=1.618, generator_adv_loss=2.127, generator_feat_match_loss=5.117, over 90.00 samples.], tot_loss[discriminator_loss=2.536, discriminator_real_loss=1.274, discriminator_fake_loss=1.262, generator_loss=30.63, generator_mel_loss=19.96, generator_kl_loss=2.027, generator_dur_loss=1.632, generator_adv_loss=2.313, generator_feat_match_loss=4.698, over 871.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:27:12,399 INFO [train.py:811] (3/4) Start epoch 799
2023-11-14 20:29:34,273 INFO [train.py:467] (3/4) Epoch 799, batch 24, global_batch_idx: 29550, batch size: 51, loss[discriminator_loss=2.398, discriminator_real_loss=1.248, discriminator_fake_loss=1.15, generator_loss=30.97, generator_mel_loss=20.13, generator_kl_loss=2.062, generator_dur_loss=1.645, generator_adv_loss=2.277, generator_feat_match_loss=4.859, over 51.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.291, discriminator_fake_loss=1.273, generator_loss=30.36, generator_mel_loss=19.95, generator_kl_loss=2.027, generator_dur_loss=1.636, generator_adv_loss=2.234, generator_feat_match_loss=4.51, over 1810.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:30:43,378 INFO [train.py:811] (3/4) Start epoch 800
2023-11-14 20:34:11,726 INFO [train.py:811] (3/4) Start epoch 801
2023-11-14 20:34:31,223 INFO [train.py:467] (3/4) Epoch 801, batch 0, global_batch_idx: 29600, batch size: 126, loss[discriminator_loss=2.371, discriminator_real_loss=1.229, discriminator_fake_loss=1.143, generator_loss=31.74, generator_mel_loss=19.99, generator_kl_loss=2.078, generator_dur_loss=1.644, generator_adv_loss=2.535, generator_feat_match_loss=5.484, over 126.00 samples.], tot_loss[discriminator_loss=2.371, discriminator_real_loss=1.229, discriminator_fake_loss=1.143, generator_loss=31.74, generator_mel_loss=19.99, generator_kl_loss=2.078, generator_dur_loss=1.644, generator_adv_loss=2.535, generator_feat_match_loss=5.484, over 126.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 20:34:31,225 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 20:34:43,063 INFO [train.py:517] (3/4) Epoch 801, validation: discriminator_loss=2.309, discriminator_real_loss=1.035, discriminator_fake_loss=1.273, generator_loss=31.85, generator_mel_loss=20.11, generator_kl_loss=2.18, generator_dur_loss=1.636, generator_adv_loss=2.45, generator_feat_match_loss=5.474, over 100.00 samples.
2023-11-14 20:34:43,064 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 20:38:02,166 INFO [train.py:811] (3/4) Start epoch 802
2023-11-14 20:39:22,647 INFO [train.py:467] (3/4) Epoch 802, batch 13, global_batch_idx: 29650, batch size: 126, loss[discriminator_loss=2.482, discriminator_real_loss=1.3, discriminator_fake_loss=1.183, generator_loss=31.58, generator_mel_loss=20.36, generator_kl_loss=2.142, generator_dur_loss=1.626, generator_adv_loss=2.367, generator_feat_match_loss=5.086, over 126.00 samples.], tot_loss[discriminator_loss=2.482, discriminator_real_loss=1.246, discriminator_fake_loss=1.236, generator_loss=30.76, generator_mel_loss=19.9, generator_kl_loss=2.034, generator_dur_loss=1.637, generator_adv_loss=2.349, generator_feat_match_loss=4.847, over 1062.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:41:39,812 INFO [train.py:811] (3/4) Start epoch 803
2023-11-14 20:44:13,179 INFO [train.py:467] (3/4) Epoch 803, batch 26, global_batch_idx: 29700, batch size: 52, loss[discriminator_loss=2.66, discriminator_real_loss=1.246, discriminator_fake_loss=1.415, generator_loss=30.59, generator_mel_loss=20.18, generator_kl_loss=2.111, generator_dur_loss=1.635, generator_adv_loss=2.244, generator_feat_match_loss=4.414, over 52.00 samples.], tot_loss[discriminator_loss=2.547, discriminator_real_loss=1.283, discriminator_fake_loss=1.263, generator_loss=30.77, generator_mel_loss=20.21, generator_kl_loss=2.033, generator_dur_loss=1.638, generator_adv_loss=2.273, generator_feat_match_loss=4.621, over 1960.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:45:14,670 INFO [train.py:811] (3/4) Start epoch 804
2023-11-14 20:48:49,228 INFO [train.py:811] (3/4) Start epoch 805
2023-11-14 20:49:18,568 INFO [train.py:467] (3/4) Epoch 805, batch 2, global_batch_idx: 29750, batch size: 53, loss[discriminator_loss=2.547, discriminator_real_loss=1.288, discriminator_fake_loss=1.26, generator_loss=29.16, generator_mel_loss=19.26, generator_kl_loss=2.029, generator_dur_loss=1.651, generator_adv_loss=2.205, generator_feat_match_loss=4.02, over 53.00 samples.], tot_loss[discriminator_loss=2.536, discriminator_real_loss=1.279, discriminator_fake_loss=1.256, generator_loss=30.48, generator_mel_loss=20.23, generator_kl_loss=2.046, generator_dur_loss=1.64, generator_adv_loss=2.121, generator_feat_match_loss=4.442, over 274.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:52:24,751 INFO [train.py:811] (3/4) Start epoch 806
2023-11-14 20:53:57,939 INFO [train.py:467] (3/4) Epoch 806, batch 15, global_batch_idx: 29800, batch size: 64, loss[discriminator_loss=2.426, discriminator_real_loss=1.144, discriminator_fake_loss=1.282, generator_loss=30.24, generator_mel_loss=19.39, generator_kl_loss=2.082, generator_dur_loss=1.627, generator_adv_loss=2.271, generator_feat_match_loss=4.871, over 64.00 samples.], tot_loss[discriminator_loss=2.507, discriminator_real_loss=1.246, discriminator_fake_loss=1.261, generator_loss=30.75, generator_mel_loss=19.96, generator_kl_loss=2.019, generator_dur_loss=1.639, generator_adv_loss=2.329, generator_feat_match_loss=4.802, over 1060.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:53:57,940 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 20:54:09,638 INFO [train.py:517] (3/4) Epoch 806, validation: discriminator_loss=2.527, discriminator_real_loss=1.069, discriminator_fake_loss=1.458, generator_loss=31.23, generator_mel_loss=20.8, generator_kl_loss=2.199, generator_dur_loss=1.638, generator_adv_loss=1.857, generator_feat_match_loss=4.737, over 100.00 samples.
2023-11-14 20:54:09,639 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 20:56:04,745 INFO [train.py:811] (3/4) Start epoch 807
2023-11-14 20:58:50,803 INFO [train.py:467] (3/4) Epoch 807, batch 28, global_batch_idx: 29850, batch size: 59, loss[discriminator_loss=2.477, discriminator_real_loss=1.341, discriminator_fake_loss=1.137, generator_loss=31.16, generator_mel_loss=20.13, generator_kl_loss=2.026, generator_dur_loss=1.657, generator_adv_loss=2.424, generator_feat_match_loss=4.93, over 59.00 samples.], tot_loss[discriminator_loss=2.553, discriminator_real_loss=1.296, discriminator_fake_loss=1.257, generator_loss=30.66, generator_mel_loss=20.11, generator_kl_loss=2.037, generator_dur_loss=1.636, generator_adv_loss=2.276, generator_feat_match_loss=4.597, over 2041.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 20:59:41,856 INFO [train.py:811] (3/4) Start epoch 808
2023-11-14 21:03:20,284 INFO [train.py:811] (3/4) Start epoch 809
2023-11-14 21:03:53,898 INFO [train.py:467] (3/4) Epoch 809, batch 4, global_batch_idx: 29900, batch size: 53, loss[discriminator_loss=2.578, discriminator_real_loss=1.287, discriminator_fake_loss=1.291, generator_loss=30.12, generator_mel_loss=20.01, generator_kl_loss=2.081, generator_dur_loss=1.648, generator_adv_loss=2.076, generator_feat_match_loss=4.312, over 53.00 samples.], tot_loss[discriminator_loss=2.56, discriminator_real_loss=1.299, discriminator_fake_loss=1.26, generator_loss=30.35, generator_mel_loss=19.79, generator_kl_loss=2.068, generator_dur_loss=1.64, generator_adv_loss=2.283, generator_feat_match_loss=4.568, over 334.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:06:52,453 INFO [train.py:811] (3/4) Start epoch 810
2023-11-14 21:08:44,193 INFO [train.py:467] (3/4) Epoch 810, batch 17, global_batch_idx: 29950, batch size: 69, loss[discriminator_loss=2.537, discriminator_real_loss=1.342, discriminator_fake_loss=1.195, generator_loss=30.8, generator_mel_loss=20.37, generator_kl_loss=2.001, generator_dur_loss=1.635, generator_adv_loss=2.359, generator_feat_match_loss=4.441, over 69.00 samples.], tot_loss[discriminator_loss=2.543, discriminator_real_loss=1.283, discriminator_fake_loss=1.26, generator_loss=30.49, generator_mel_loss=19.89, generator_kl_loss=2.027, generator_dur_loss=1.634, generator_adv_loss=2.281, generator_feat_match_loss=4.652, over 1363.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:10:25,759 INFO [train.py:811] (3/4) Start epoch 811
2023-11-14 21:13:17,655 INFO [train.py:467] (3/4) Epoch 811, batch 30, global_batch_idx: 30000, batch size: 69, loss[discriminator_loss=2.32, discriminator_real_loss=1.184, discriminator_fake_loss=1.137, generator_loss=31.76, generator_mel_loss=20.17, generator_kl_loss=1.999, generator_dur_loss=1.64, generator_adv_loss=2.473, generator_feat_match_loss=5.477, over 69.00 samples.], tot_loss[discriminator_loss=2.559, discriminator_real_loss=1.296, discriminator_fake_loss=1.263, generator_loss=30.8, generator_mel_loss=20.15, generator_kl_loss=2.055, generator_dur_loss=1.634, generator_adv_loss=2.287, generator_feat_match_loss=4.681, over 2380.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 21:13:17,657 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 21:13:28,584 INFO [train.py:517] (3/4) Epoch 811, validation: discriminator_loss=2.392, discriminator_real_loss=1.16, discriminator_fake_loss=1.232, generator_loss=32.08, generator_mel_loss=20.83, generator_kl_loss=2.196, generator_dur_loss=1.648, generator_adv_loss=2.263, generator_feat_match_loss=5.142, over 100.00 samples.
2023-11-14 21:13:28,585 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 21:14:05,688 INFO [train.py:811] (3/4) Start epoch 812
2023-11-14 21:17:42,480 INFO [train.py:811] (3/4) Start epoch 813
2023-11-14 21:18:27,684 INFO [train.py:467] (3/4) Epoch 813, batch 6, global_batch_idx: 30050, batch size: 71, loss[discriminator_loss=2.531, discriminator_real_loss=1.285, discriminator_fake_loss=1.246, generator_loss=30.7, generator_mel_loss=19.88, generator_kl_loss=1.981, generator_dur_loss=1.642, generator_adv_loss=2.369, generator_feat_match_loss=4.828, over 71.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.305, discriminator_fake_loss=1.235, generator_loss=30.69, generator_mel_loss=20.03, generator_kl_loss=2.03, generator_dur_loss=1.642, generator_adv_loss=2.311, generator_feat_match_loss=4.677, over 477.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 21:21:13,581 INFO [train.py:811] (3/4) Start epoch 814
2023-11-14 21:23:19,717 INFO [train.py:467] (3/4) Epoch 814, batch 19, global_batch_idx: 30100, batch size: 69, loss[discriminator_loss=2.5, discriminator_real_loss=1.198, discriminator_fake_loss=1.301, generator_loss=30.54, generator_mel_loss=20.32, generator_kl_loss=2.025, generator_dur_loss=1.631, generator_adv_loss=2.098, generator_feat_match_loss=4.461, over 69.00 samples.], tot_loss[discriminator_loss=2.546, discriminator_real_loss=1.284, discriminator_fake_loss=1.262, generator_loss=30.51, generator_mel_loss=20.02, generator_kl_loss=2.05, generator_dur_loss=1.636, generator_adv_loss=2.227, generator_feat_match_loss=4.583, over 1356.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 16.0
2023-11-14 21:24:42,751 INFO [train.py:811] (3/4) Start epoch 815
2023-11-14 21:27:58,748 INFO [train.py:467] (3/4) Epoch 815, batch 32, global_batch_idx: 30150, batch size: 58, loss[discriminator_loss=2.551, discriminator_real_loss=1.17, discriminator_fake_loss=1.38, generator_loss=29.63, generator_mel_loss=19.22, generator_kl_loss=1.991, generator_dur_loss=1.644, generator_adv_loss=2.416, generator_feat_match_loss=4.363, over 58.00 samples.], tot_loss[discriminator_loss=2.486, discriminator_real_loss=1.263, discriminator_fake_loss=1.224, generator_loss=30.93, generator_mel_loss=19.79, generator_kl_loss=2.021, generator_dur_loss=1.631, generator_adv_loss=2.414, generator_feat_match_loss=5.082, over 2660.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:28:16,209 INFO [train.py:811] (3/4) Start epoch 816
2023-11-14 21:31:51,857 INFO [train.py:811] (3/4) Start epoch 817
2023-11-14 21:32:52,176 INFO [train.py:467] (3/4) Epoch 817, batch 8, global_batch_idx: 30200, batch size: 126, loss[discriminator_loss=2.664, discriminator_real_loss=1.447, discriminator_fake_loss=1.216, generator_loss=30.52, generator_mel_loss=20.25, generator_kl_loss=2.051, generator_dur_loss=1.62, generator_adv_loss=2.129, generator_feat_match_loss=4.473, over 126.00 samples.], tot_loss[discriminator_loss=2.645, discriminator_real_loss=1.355, discriminator_fake_loss=1.29, generator_loss=30.5, generator_mel_loss=20.37, generator_kl_loss=2.044, generator_dur_loss=1.629, generator_adv_loss=2.122, generator_feat_match_loss=4.339, over 731.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:32:52,178 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 21:33:03,424 INFO [train.py:517] (3/4) Epoch 817, validation: discriminator_loss=2.652, discriminator_real_loss=1.239, discriminator_fake_loss=1.412, generator_loss=30.8, generator_mel_loss=20.69, generator_kl_loss=2.168, generator_dur_loss=1.631, generator_adv_loss=1.917, generator_feat_match_loss=4.389, over 100.00 samples.
2023-11-14 21:33:03,425 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 21:35:39,813 INFO [train.py:811] (3/4) Start epoch 818
2023-11-14 21:38:03,008 INFO [train.py:467] (3/4) Epoch 818, batch 21, global_batch_idx: 30250, batch size: 85, loss[discriminator_loss=2.516, discriminator_real_loss=1.435, discriminator_fake_loss=1.082, generator_loss=30.79, generator_mel_loss=19.99, generator_kl_loss=2.002, generator_dur_loss=1.66, generator_adv_loss=2.338, generator_feat_match_loss=4.801, over 85.00 samples.], tot_loss[discriminator_loss=2.568, discriminator_real_loss=1.302, discriminator_fake_loss=1.265, generator_loss=30.7, generator_mel_loss=20.18, generator_kl_loss=2.004, generator_dur_loss=1.638, generator_adv_loss=2.262, generator_feat_match_loss=4.618, over 1664.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:39:15,891 INFO [train.py:811] (3/4) Start epoch 819
2023-11-14 21:42:35,698 INFO [train.py:467] (3/4) Epoch 819, batch 34, global_batch_idx: 30300, batch size: 85, loss[discriminator_loss=2.617, discriminator_real_loss=1.413, discriminator_fake_loss=1.205, generator_loss=29.97, generator_mel_loss=19.89, generator_kl_loss=2.063, generator_dur_loss=1.642, generator_adv_loss=2.113, generator_feat_match_loss=4.266, over 85.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.368, discriminator_fake_loss=1.247, generator_loss=30.2, generator_mel_loss=19.67, generator_kl_loss=2.004, generator_dur_loss=1.635, generator_adv_loss=2.273, generator_feat_match_loss=4.62, over 2456.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:42:51,037 INFO [train.py:811] (3/4) Start epoch 820
2023-11-14 21:46:18,462 INFO [train.py:811] (3/4) Start epoch 821
2023-11-14 21:47:31,044 INFO [train.py:467] (3/4) Epoch 821, batch 10, global_batch_idx: 30350, batch size: 79, loss[discriminator_loss=2.625, discriminator_real_loss=1.347, discriminator_fake_loss=1.279, generator_loss=30.01, generator_mel_loss=19.96, generator_kl_loss=1.957, generator_dur_loss=1.628, generator_adv_loss=2.123, generator_feat_match_loss=4.344, over 79.00 samples.], tot_loss[discriminator_loss=2.577, discriminator_real_loss=1.321, discriminator_fake_loss=1.257, generator_loss=30.51, generator_mel_loss=20.14, generator_kl_loss=2.011, generator_dur_loss=1.632, generator_adv_loss=2.208, generator_feat_match_loss=4.52, over 830.00 samples.], cur_lr_g: 1.81e-04, cur_lr_d: 1.81e-04, grad_scale: 8.0
2023-11-14 21:49:52,253 INFO [train.py:811] (3/4) Start epoch 822
2023-11-14 21:52:15,893 INFO [train.py:467] (3/4) Epoch 822, batch 23, global_batch_idx: 30400, batch size: 110, loss[discriminator_loss=2.555, discriminator_real_loss=1.489, discriminator_fake_loss=1.064, generator_loss=31.09, generator_mel_loss=20.27, generator_kl_loss=2.089, generator_dur_loss=1.63, generator_adv_loss=2.262, generator_feat_match_loss=4.844, over 110.00 samples.], tot_loss[discriminator_loss=2.581, discriminator_real_loss=1.304, discriminator_fake_loss=1.278, generator_loss=30.61, generator_mel_loss=20.12, generator_kl_loss=2.014, generator_dur_loss=1.64, generator_adv_loss=2.25, generator_feat_match_loss=4.583, over 1771.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-14 21:52:15,895 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 21:52:26,906 INFO [train.py:517] (3/4) Epoch 822, validation: discriminator_loss=2.61, discriminator_real_loss=1.172, discriminator_fake_loss=1.438, generator_loss=30.48, generator_mel_loss=20.46, generator_kl_loss=2.197, generator_dur_loss=1.64, generator_adv_loss=1.842, generator_feat_match_loss=4.345, over 100.00 samples.
2023-11-14 21:52:26,907 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 21:53:36,131 INFO [train.py:811] (3/4) Start epoch 823
2023-11-14 21:57:10,106 INFO [train.py:467] (3/4) Epoch 823, batch 36, global_batch_idx: 30450, batch size: 60, loss[discriminator_loss=2.641, discriminator_real_loss=1.488, discriminator_fake_loss=1.153, generator_loss=31.34, generator_mel_loss=20.43, generator_kl_loss=2.022, generator_dur_loss=1.645, generator_adv_loss=2.477, generator_feat_match_loss=4.766, over 60.00 samples.], tot_loss[discriminator_loss=2.512, discriminator_real_loss=1.271, discriminator_fake_loss=1.24, generator_loss=30.87, generator_mel_loss=20.02, generator_kl_loss=2.037, generator_dur_loss=1.635, generator_adv_loss=2.332, generator_feat_match_loss=4.842, over 2828.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 21:57:10,867 INFO [train.py:811] (3/4) Start epoch 824
2023-11-14 22:00:36,968 INFO [train.py:811] (3/4) Start epoch 825
2023-11-14 22:02:01,462 INFO [train.py:467] (3/4) Epoch 825, batch 12, global_batch_idx: 30500, batch size: 85, loss[discriminator_loss=2.555, discriminator_real_loss=1.217, discriminator_fake_loss=1.339, generator_loss=30.1, generator_mel_loss=19.87, generator_kl_loss=2.013, generator_dur_loss=1.635, generator_adv_loss=2.152, generator_feat_match_loss=4.43, over 85.00 samples.], tot_loss[discriminator_loss=2.562, discriminator_real_loss=1.3, discriminator_fake_loss=1.262, generator_loss=30.91, generator_mel_loss=20.1, generator_kl_loss=2.037, generator_dur_loss=1.639, generator_adv_loss=2.35, generator_feat_match_loss=4.786, over 1057.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:04:09,039 INFO [train.py:811] (3/4) Start epoch 826
2023-11-14 22:06:42,025 INFO [train.py:467] (3/4) Epoch 826, batch 25, global_batch_idx: 30550, batch size: 61, loss[discriminator_loss=2.508, discriminator_real_loss=1.31, discriminator_fake_loss=1.199, generator_loss=30.44, generator_mel_loss=19.77, generator_kl_loss=1.988, generator_dur_loss=1.626, generator_adv_loss=2.344, generator_feat_match_loss=4.711, over 61.00 samples.], tot_loss[discriminator_loss=2.526, discriminator_real_loss=1.273, discriminator_fake_loss=1.254, generator_loss=30.61, generator_mel_loss=19.9, generator_kl_loss=2.015, generator_dur_loss=1.635, generator_adv_loss=2.297, generator_feat_match_loss=4.76, over 1870.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:07:38,064 INFO [train.py:811] (3/4) Start epoch 827
2023-11-14 22:11:16,378 INFO [train.py:811] (3/4) Start epoch 828
2023-11-14 22:11:39,087 INFO [train.py:467] (3/4) Epoch 828, batch 1, global_batch_idx: 30600, batch size: 64, loss[discriminator_loss=2.562, discriminator_real_loss=1.322, discriminator_fake_loss=1.239, generator_loss=30.47, generator_mel_loss=19.79, generator_kl_loss=1.997, generator_dur_loss=1.634, generator_adv_loss=2.377, generator_feat_match_loss=4.672, over 64.00 samples.], tot_loss[discriminator_loss=2.536, discriminator_real_loss=1.28, discriminator_fake_loss=1.255, generator_loss=30.65, generator_mel_loss=19.98, generator_kl_loss=1.963, generator_dur_loss=1.638, generator_adv_loss=2.396, generator_feat_match_loss=4.674, over 128.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:11:39,088 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 22:11:50,897 INFO [train.py:517] (3/4) Epoch 828, validation: discriminator_loss=2.486, discriminator_real_loss=1.237, discriminator_fake_loss=1.249, generator_loss=31.89, generator_mel_loss=20.74, generator_kl_loss=2.258, generator_dur_loss=1.635, generator_adv_loss=2.294, generator_feat_match_loss=4.963, over 100.00 samples.
2023-11-14 22:11:50,898 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 22:15:00,407 INFO [train.py:811] (3/4) Start epoch 829
2023-11-14 22:16:34,448 INFO [train.py:467] (3/4) Epoch 829, batch 14, global_batch_idx: 30650, batch size: 69, loss[discriminator_loss=2.508, discriminator_real_loss=1.238, discriminator_fake_loss=1.271, generator_loss=30.11, generator_mel_loss=19.82, generator_kl_loss=2.073, generator_dur_loss=1.645, generator_adv_loss=2.113, generator_feat_match_loss=4.465, over 69.00 samples.], tot_loss[discriminator_loss=2.492, discriminator_real_loss=1.259, discriminator_fake_loss=1.233, generator_loss=30.54, generator_mel_loss=19.95, generator_kl_loss=2.051, generator_dur_loss=1.638, generator_adv_loss=2.244, generator_feat_match_loss=4.66, over 1150.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:18:34,086 INFO [train.py:811] (3/4) Start epoch 830
2023-11-14 22:21:14,973 INFO [train.py:467] (3/4) Epoch 830, batch 27, global_batch_idx: 30700, batch size: 101, loss[discriminator_loss=2.734, discriminator_real_loss=1.329, discriminator_fake_loss=1.405, generator_loss=29.77, generator_mel_loss=19.74, generator_kl_loss=2.05, generator_dur_loss=1.623, generator_adv_loss=2.164, generator_feat_match_loss=4.188, over 101.00 samples.], tot_loss[discriminator_loss=2.566, discriminator_real_loss=1.297, discriminator_fake_loss=1.269, generator_loss=30.61, generator_mel_loss=19.84, generator_kl_loss=2.022, generator_dur_loss=1.639, generator_adv_loss=2.343, generator_feat_match_loss=4.768, over 1990.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:22:08,631 INFO [train.py:811] (3/4) Start epoch 831
2023-11-14 22:25:46,250 INFO [train.py:811] (3/4) Start epoch 832
2023-11-14 22:26:19,938 INFO [train.py:467] (3/4) Epoch 832, batch 3, global_batch_idx: 30750, batch size: 52, loss[discriminator_loss=2.473, discriminator_real_loss=1.252, discriminator_fake_loss=1.222, generator_loss=31.11, generator_mel_loss=20.22, generator_kl_loss=1.952, generator_dur_loss=1.653, generator_adv_loss=2.166, generator_feat_match_loss=5.117, over 52.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.374, discriminator_fake_loss=1.217, generator_loss=30.45, generator_mel_loss=19.94, generator_kl_loss=1.975, generator_dur_loss=1.64, generator_adv_loss=2.282, generator_feat_match_loss=4.613, over 224.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:29:15,815 INFO [train.py:811] (3/4) Start epoch 833
2023-11-14 22:30:54,174 INFO [train.py:467] (3/4) Epoch 833, batch 16, global_batch_idx: 30800, batch size: 52, loss[discriminator_loss=2.617, discriminator_real_loss=1.263, discriminator_fake_loss=1.355, generator_loss=29.62, generator_mel_loss=19.62, generator_kl_loss=2.074, generator_dur_loss=1.635, generator_adv_loss=2.299, generator_feat_match_loss=4, over 52.00 samples.], tot_loss[discriminator_loss=2.545, discriminator_real_loss=1.299, discriminator_fake_loss=1.247, generator_loss=30.42, generator_mel_loss=19.94, generator_kl_loss=2.004, generator_dur_loss=1.632, generator_adv_loss=2.253, generator_feat_match_loss=4.588, over 1301.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-14 22:30:54,176 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 22:31:05,524 INFO [train.py:517] (3/4) Epoch 833, validation: discriminator_loss=2.566, discriminator_real_loss=1.36, discriminator_fake_loss=1.206, generator_loss=30.99, generator_mel_loss=20.35, generator_kl_loss=2.14, generator_dur_loss=1.632, generator_adv_loss=2.373, generator_feat_match_loss=4.486, over 100.00 samples.
2023-11-14 22:31:05,526 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 22:32:55,743 INFO [train.py:811] (3/4) Start epoch 834
2023-11-14 22:35:40,743 INFO [train.py:467] (3/4) Epoch 834, batch 29, global_batch_idx: 30850, batch size: 73, loss[discriminator_loss=2.568, discriminator_real_loss=1.43, discriminator_fake_loss=1.139, generator_loss=30.2, generator_mel_loss=19.71, generator_kl_loss=2.039, generator_dur_loss=1.648, generator_adv_loss=2.361, generator_feat_match_loss=4.434, over 73.00 samples.], tot_loss[discriminator_loss=2.494, discriminator_real_loss=1.259, discriminator_fake_loss=1.235, generator_loss=30.69, generator_mel_loss=19.68, generator_kl_loss=2.021, generator_dur_loss=1.635, generator_adv_loss=2.375, generator_feat_match_loss=4.971, over 2091.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:36:20,493 INFO [train.py:811] (3/4) Start epoch 835
2023-11-14 22:39:51,928 INFO [train.py:811] (3/4) Start epoch 836
2023-11-14 22:40:29,152 INFO [train.py:467] (3/4) Epoch 836, batch 5, global_batch_idx: 30900, batch size: 53, loss[discriminator_loss=2.66, discriminator_real_loss=1.391, discriminator_fake_loss=1.271, generator_loss=29.31, generator_mel_loss=19.61, generator_kl_loss=1.958, generator_dur_loss=1.637, generator_adv_loss=2.053, generator_feat_match_loss=4.059, over 53.00 samples.], tot_loss[discriminator_loss=2.594, discriminator_real_loss=1.314, discriminator_fake_loss=1.281, generator_loss=30.23, generator_mel_loss=20.01, generator_kl_loss=1.994, generator_dur_loss=1.644, generator_adv_loss=2.167, generator_feat_match_loss=4.418, over 370.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:43:18,506 INFO [train.py:811] (3/4) Start epoch 837
2023-11-14 22:45:18,532 INFO [train.py:467] (3/4) Epoch 837, batch 18, global_batch_idx: 30950, batch size: 67, loss[discriminator_loss=2.602, discriminator_real_loss=1.31, discriminator_fake_loss=1.292, generator_loss=30.65, generator_mel_loss=20.18, generator_kl_loss=2.078, generator_dur_loss=1.652, generator_adv_loss=2.293, generator_feat_match_loss=4.453, over 67.00 samples.], tot_loss[discriminator_loss=2.634, discriminator_real_loss=1.336, discriminator_fake_loss=1.297, generator_loss=30.53, generator_mel_loss=20.31, generator_kl_loss=2.03, generator_dur_loss=1.631, generator_adv_loss=2.166, generator_feat_match_loss=4.395, over 1329.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:46:54,141 INFO [train.py:811] (3/4) Start epoch 838
2023-11-14 22:50:00,836 INFO [train.py:467] (3/4) Epoch 838, batch 31, global_batch_idx: 31000, batch size: 49, loss[discriminator_loss=2.48, discriminator_real_loss=1.229, discriminator_fake_loss=1.251, generator_loss=30.33, generator_mel_loss=19.5, generator_kl_loss=2.046, generator_dur_loss=1.643, generator_adv_loss=2.279, generator_feat_match_loss=4.859, over 49.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.319, discriminator_fake_loss=1.282, generator_loss=30.51, generator_mel_loss=20.13, generator_kl_loss=2.017, generator_dur_loss=1.639, generator_adv_loss=2.235, generator_feat_match_loss=4.489, over 2242.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:50:00,838 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 22:50:12,834 INFO [train.py:517] (3/4) Epoch 838, validation: discriminator_loss=2.427, discriminator_real_loss=1.072, discriminator_fake_loss=1.355, generator_loss=31, generator_mel_loss=20.78, generator_kl_loss=2.183, generator_dur_loss=1.634, generator_adv_loss=1.901, generator_feat_match_loss=4.502, over 100.00 samples.
2023-11-14 22:50:12,835 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 22:50:40,521 INFO [train.py:811] (3/4) Start epoch 839
2023-11-14 22:54:13,596 INFO [train.py:811] (3/4) Start epoch 840
2023-11-14 22:55:15,144 INFO [train.py:467] (3/4) Epoch 840, batch 7, global_batch_idx: 31050, batch size: 110, loss[discriminator_loss=2.473, discriminator_real_loss=1.368, discriminator_fake_loss=1.105, generator_loss=31.37, generator_mel_loss=20.23, generator_kl_loss=1.991, generator_dur_loss=1.637, generator_adv_loss=2.408, generator_feat_match_loss=5.105, over 110.00 samples.], tot_loss[discriminator_loss=2.484, discriminator_real_loss=1.281, discriminator_fake_loss=1.202, generator_loss=30.8, generator_mel_loss=19.91, generator_kl_loss=2.052, generator_dur_loss=1.636, generator_adv_loss=2.352, generator_feat_match_loss=4.853, over 730.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 22:57:49,313 INFO [train.py:811] (3/4) Start epoch 841
2023-11-14 22:59:43,229 INFO [train.py:467] (3/4) Epoch 841, batch 20, global_batch_idx: 31100, batch size: 76, loss[discriminator_loss=2.549, discriminator_real_loss=1.254, discriminator_fake_loss=1.295, generator_loss=30.28, generator_mel_loss=19.73, generator_kl_loss=1.995, generator_dur_loss=1.637, generator_adv_loss=2.221, generator_feat_match_loss=4.703, over 76.00 samples.], tot_loss[discriminator_loss=2.505, discriminator_real_loss=1.274, discriminator_fake_loss=1.232, generator_loss=30.54, generator_mel_loss=19.77, generator_kl_loss=2.026, generator_dur_loss=1.635, generator_adv_loss=2.351, generator_feat_match_loss=4.755, over 1492.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:01:14,139 INFO [train.py:811] (3/4) Start epoch 842
2023-11-14 23:04:19,354 INFO [train.py:467] (3/4) Epoch 842, batch 33, global_batch_idx: 31150, batch size: 69, loss[discriminator_loss=2.348, discriminator_real_loss=1.206, discriminator_fake_loss=1.143, generator_loss=31.83, generator_mel_loss=20.14, generator_kl_loss=1.972, generator_dur_loss=1.631, generator_adv_loss=2.465, generator_feat_match_loss=5.625, over 69.00 samples.], tot_loss[discriminator_loss=2.522, discriminator_real_loss=1.286, discriminator_fake_loss=1.236, generator_loss=30.83, generator_mel_loss=19.97, generator_kl_loss=2.016, generator_dur_loss=1.634, generator_adv_loss=2.338, generator_feat_match_loss=4.868, over 2308.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:04:39,211 INFO [train.py:811] (3/4) Start epoch 843
2023-11-14 23:08:10,261 INFO [train.py:811] (3/4) Start epoch 844
2023-11-14 23:09:09,520 INFO [train.py:467] (3/4) Epoch 844, batch 9, global_batch_idx: 31200, batch size: 65, loss[discriminator_loss=2.619, discriminator_real_loss=1.336, discriminator_fake_loss=1.283, generator_loss=30.08, generator_mel_loss=20.02, generator_kl_loss=1.983, generator_dur_loss=1.621, generator_adv_loss=2.088, generator_feat_match_loss=4.359, over 65.00 samples.], tot_loss[discriminator_loss=2.625, discriminator_real_loss=1.351, discriminator_fake_loss=1.274, generator_loss=30.46, generator_mel_loss=20.2, generator_kl_loss=2.007, generator_dur_loss=1.633, generator_adv_loss=2.153, generator_feat_match_loss=4.464, over 749.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-14 23:09:09,522 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 23:09:20,749 INFO [train.py:517] (3/4) Epoch 844, validation: discriminator_loss=2.619, discriminator_real_loss=1.188, discriminator_fake_loss=1.431, generator_loss=31.26, generator_mel_loss=20.99, generator_kl_loss=2.229, generator_dur_loss=1.629, generator_adv_loss=1.894, generator_feat_match_loss=4.517, over 100.00 samples.
2023-11-14 23:09:20,750 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 23:11:51,702 INFO [train.py:811] (3/4) Start epoch 845
2023-11-14 23:14:06,000 INFO [train.py:467] (3/4) Epoch 845, batch 22, global_batch_idx: 31250, batch size: 53, loss[discriminator_loss=2.639, discriminator_real_loss=1.466, discriminator_fake_loss=1.173, generator_loss=30.2, generator_mel_loss=19.93, generator_kl_loss=1.978, generator_dur_loss=1.652, generator_adv_loss=2.277, generator_feat_match_loss=4.367, over 53.00 samples.], tot_loss[discriminator_loss=2.628, discriminator_real_loss=1.346, discriminator_fake_loss=1.282, generator_loss=30.4, generator_mel_loss=20.16, generator_kl_loss=2.037, generator_dur_loss=1.635, generator_adv_loss=2.185, generator_feat_match_loss=4.384, over 1654.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-14 23:15:25,443 INFO [train.py:811] (3/4) Start epoch 846
2023-11-14 23:18:53,012 INFO [train.py:467] (3/4) Epoch 846, batch 35, global_batch_idx: 31300, batch size: 59, loss[discriminator_loss=2.637, discriminator_real_loss=1.338, discriminator_fake_loss=1.299, generator_loss=30.24, generator_mel_loss=19.97, generator_kl_loss=2.029, generator_dur_loss=1.644, generator_adv_loss=2.312, generator_feat_match_loss=4.281, over 59.00 samples.], tot_loss[discriminator_loss=2.603, discriminator_real_loss=1.32, discriminator_fake_loss=1.283, generator_loss=30.51, generator_mel_loss=20.18, generator_kl_loss=2.023, generator_dur_loss=1.637, generator_adv_loss=2.221, generator_feat_match_loss=4.452, over 2546.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:18:58,214 INFO [train.py:811] (3/4) Start epoch 847
2023-11-14 23:22:37,268 INFO [train.py:811] (3/4) Start epoch 848
2023-11-14 23:23:47,717 INFO [train.py:467] (3/4) Epoch 848, batch 11, global_batch_idx: 31350, batch size: 64, loss[discriminator_loss=2.48, discriminator_real_loss=1.259, discriminator_fake_loss=1.222, generator_loss=30.97, generator_mel_loss=19.94, generator_kl_loss=2.049, generator_dur_loss=1.631, generator_adv_loss=2.432, generator_feat_match_loss=4.914, over 64.00 samples.], tot_loss[discriminator_loss=2.532, discriminator_real_loss=1.268, discriminator_fake_loss=1.264, generator_loss=30.72, generator_mel_loss=20.07, generator_kl_loss=1.996, generator_dur_loss=1.635, generator_adv_loss=2.311, generator_feat_match_loss=4.707, over 836.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:26:09,603 INFO [train.py:811] (3/4) Start epoch 849
2023-11-14 23:28:37,719 INFO [train.py:467] (3/4) Epoch 849, batch 24, global_batch_idx: 31400, batch size: 126, loss[discriminator_loss=2.504, discriminator_real_loss=1.175, discriminator_fake_loss=1.33, generator_loss=30.98, generator_mel_loss=20.21, generator_kl_loss=2.002, generator_dur_loss=1.633, generator_adv_loss=2.215, generator_feat_match_loss=4.922, over 126.00 samples.], tot_loss[discriminator_loss=2.496, discriminator_real_loss=1.255, discriminator_fake_loss=1.241, generator_loss=30.86, generator_mel_loss=19.96, generator_kl_loss=2.049, generator_dur_loss=1.633, generator_adv_loss=2.343, generator_feat_match_loss=4.867, over 2108.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:28:37,721 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 23:28:48,942 INFO [train.py:517] (3/4) Epoch 849, validation: discriminator_loss=2.489, discriminator_real_loss=1.2, discriminator_fake_loss=1.289, generator_loss=31.3, generator_mel_loss=20.63, generator_kl_loss=2.258, generator_dur_loss=1.633, generator_adv_loss=1.962, generator_feat_match_loss=4.819, over 100.00 samples.
2023-11-14 23:28:48,943 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 23:29:55,951 INFO [train.py:811] (3/4) Start epoch 850
2023-11-14 23:33:25,590 INFO [train.py:811] (3/4) Start epoch 851
2023-11-14 23:33:40,108 INFO [train.py:467] (3/4) Epoch 851, batch 0, global_batch_idx: 31450, batch size: 65, loss[discriminator_loss=2.535, discriminator_real_loss=1.31, discriminator_fake_loss=1.227, generator_loss=30.7, generator_mel_loss=19.94, generator_kl_loss=2.09, generator_dur_loss=1.654, generator_adv_loss=2.379, generator_feat_match_loss=4.633, over 65.00 samples.], tot_loss[discriminator_loss=2.535, discriminator_real_loss=1.31, discriminator_fake_loss=1.227, generator_loss=30.7, generator_mel_loss=19.94, generator_kl_loss=2.09, generator_dur_loss=1.654, generator_adv_loss=2.379, generator_feat_match_loss=4.633, over 65.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:37:01,443 INFO [train.py:811] (3/4) Start epoch 852
2023-11-14 23:38:32,189 INFO [train.py:467] (3/4) Epoch 852, batch 13, global_batch_idx: 31500, batch size: 95, loss[discriminator_loss=2.459, discriminator_real_loss=1.181, discriminator_fake_loss=1.278, generator_loss=30.57, generator_mel_loss=19.64, generator_kl_loss=1.93, generator_dur_loss=1.636, generator_adv_loss=2.328, generator_feat_match_loss=5.039, over 95.00 samples.], tot_loss[discriminator_loss=2.53, discriminator_real_loss=1.267, discriminator_fake_loss=1.264, generator_loss=30.93, generator_mel_loss=20.03, generator_kl_loss=1.989, generator_dur_loss=1.634, generator_adv_loss=2.313, generator_feat_match_loss=4.966, over 1146.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:40:31,603 INFO [train.py:811] (3/4) Start epoch 853
2023-11-14 23:43:12,767 INFO [train.py:467] (3/4) Epoch 853, batch 26, global_batch_idx: 31550, batch size: 110, loss[discriminator_loss=2.529, discriminator_real_loss=1.287, discriminator_fake_loss=1.242, generator_loss=30.82, generator_mel_loss=20.1, generator_kl_loss=1.983, generator_dur_loss=1.624, generator_adv_loss=2.299, generator_feat_match_loss=4.812, over 110.00 samples.], tot_loss[discriminator_loss=2.556, discriminator_real_loss=1.303, discriminator_fake_loss=1.252, generator_loss=30.41, generator_mel_loss=19.95, generator_kl_loss=2.016, generator_dur_loss=1.634, generator_adv_loss=2.225, generator_feat_match_loss=4.583, over 2058.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:44:08,250 INFO [train.py:811] (3/4) Start epoch 854
2023-11-14 23:47:43,563 INFO [train.py:811] (3/4) Start epoch 855
2023-11-14 23:48:06,061 INFO [train.py:467] (3/4) Epoch 855, batch 2, global_batch_idx: 31600, batch size: 61, loss[discriminator_loss=2.41, discriminator_real_loss=1.185, discriminator_fake_loss=1.226, generator_loss=30.53, generator_mel_loss=19.43, generator_kl_loss=2.023, generator_dur_loss=1.634, generator_adv_loss=2.35, generator_feat_match_loss=5.094, over 61.00 samples.], tot_loss[discriminator_loss=2.488, discriminator_real_loss=1.263, discriminator_fake_loss=1.225, generator_loss=30.12, generator_mel_loss=19.44, generator_kl_loss=2.035, generator_dur_loss=1.636, generator_adv_loss=2.202, generator_feat_match_loss=4.801, over 195.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-14 23:48:06,062 INFO [train.py:508] (3/4) Computing validation loss
2023-11-14 23:48:18,034 INFO [train.py:517] (3/4) Epoch 855, validation: discriminator_loss=2.382, discriminator_real_loss=1.122, discriminator_fake_loss=1.26, generator_loss=31.99, generator_mel_loss=20.71, generator_kl_loss=2.234, generator_dur_loss=1.63, generator_adv_loss=2.125, generator_feat_match_loss=5.289, over 100.00 samples.
2023-11-14 23:48:18,035 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-14 23:51:27,828 INFO [train.py:811] (3/4) Start epoch 856
2023-11-14 23:53:03,529 INFO [train.py:467] (3/4) Epoch 856, batch 15, global_batch_idx: 31650, batch size: 101, loss[discriminator_loss=2.408, discriminator_real_loss=1.258, discriminator_fake_loss=1.15, generator_loss=31.47, generator_mel_loss=20.21, generator_kl_loss=2.039, generator_dur_loss=1.67, generator_adv_loss=2.443, generator_feat_match_loss=5.105, over 101.00 samples.], tot_loss[discriminator_loss=2.484, discriminator_real_loss=1.254, discriminator_fake_loss=1.23, generator_loss=30.93, generator_mel_loss=20.04, generator_kl_loss=2.028, generator_dur_loss=1.634, generator_adv_loss=2.331, generator_feat_match_loss=4.898, over 1198.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:55:02,775 INFO [train.py:811] (3/4) Start epoch 857
2023-11-14 23:57:43,721 INFO [train.py:467] (3/4) Epoch 857, batch 28, global_batch_idx: 31700, batch size: 76, loss[discriminator_loss=2.428, discriminator_real_loss=1.172, discriminator_fake_loss=1.256, generator_loss=31, generator_mel_loss=19.72, generator_kl_loss=2.017, generator_dur_loss=1.647, generator_adv_loss=2.496, generator_feat_match_loss=5.113, over 76.00 samples.], tot_loss[discriminator_loss=2.474, discriminator_real_loss=1.248, discriminator_fake_loss=1.226, generator_loss=30.75, generator_mel_loss=19.74, generator_kl_loss=2.01, generator_dur_loss=1.636, generator_adv_loss=2.38, generator_feat_match_loss=4.987, over 1978.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-14 23:58:30,424 INFO [train.py:811] (3/4) Start epoch 858
2023-11-15 00:02:05,211 INFO [train.py:811] (3/4) Start epoch 859
2023-11-15 00:02:41,926 INFO [train.py:467] (3/4) Epoch 859, batch 4, global_batch_idx: 31750, batch size: 69, loss[discriminator_loss=2.309, discriminator_real_loss=1.147, discriminator_fake_loss=1.16, generator_loss=31.11, generator_mel_loss=19.4, generator_kl_loss=2.089, generator_dur_loss=1.64, generator_adv_loss=2.523, generator_feat_match_loss=5.453, over 69.00 samples.], tot_loss[discriminator_loss=2.364, discriminator_real_loss=1.209, discriminator_fake_loss=1.155, generator_loss=31.31, generator_mel_loss=19.83, generator_kl_loss=2.028, generator_dur_loss=1.634, generator_adv_loss=2.392, generator_feat_match_loss=5.427, over 367.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-15 00:05:43,082 INFO [train.py:811] (3/4) Start epoch 860
2023-11-15 00:07:31,631 INFO [train.py:467] (3/4) Epoch 860, batch 17, global_batch_idx: 31800, batch size: 52, loss[discriminator_loss=2.525, discriminator_real_loss=1.234, discriminator_fake_loss=1.291, generator_loss=31.26, generator_mel_loss=20.33, generator_kl_loss=2.042, generator_dur_loss=1.66, generator_adv_loss=2.414, generator_feat_match_loss=4.809, over 52.00 samples.], tot_loss[discriminator_loss=2.555, discriminator_real_loss=1.293, discriminator_fake_loss=1.262, generator_loss=30.51, generator_mel_loss=20.01, generator_kl_loss=2.041, generator_dur_loss=1.635, generator_adv_loss=2.23, generator_feat_match_loss=4.597, over 1299.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-15 00:07:31,632 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 00:07:42,688 INFO [train.py:517] (3/4) Epoch 860, validation: discriminator_loss=2.631, discriminator_real_loss=1.324, discriminator_fake_loss=1.307, generator_loss=31.27, generator_mel_loss=20.89, generator_kl_loss=2.236, generator_dur_loss=1.624, generator_adv_loss=2.088, generator_feat_match_loss=4.43, over 100.00 samples.
2023-11-15 00:07:42,689 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 00:09:31,121 INFO [train.py:811] (3/4) Start epoch 861
2023-11-15 00:12:28,551 INFO [train.py:467] (3/4) Epoch 861, batch 30, global_batch_idx: 31850, batch size: 51, loss[discriminator_loss=2.547, discriminator_real_loss=1.18, discriminator_fake_loss=1.367, generator_loss=31.28, generator_mel_loss=20.52, generator_kl_loss=1.931, generator_dur_loss=1.661, generator_adv_loss=2.322, generator_feat_match_loss=4.844, over 51.00 samples.], tot_loss[discriminator_loss=2.583, discriminator_real_loss=1.304, discriminator_fake_loss=1.279, generator_loss=30.75, generator_mel_loss=20.22, generator_kl_loss=2.036, generator_dur_loss=1.631, generator_adv_loss=2.234, generator_feat_match_loss=4.634, over 2317.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-15 00:13:08,472 INFO [train.py:811] (3/4) Start epoch 862
2023-11-15 00:16:40,544 INFO [train.py:811] (3/4) Start epoch 863
2023-11-15 00:17:27,071 INFO [train.py:467] (3/4) Epoch 863, batch 6, global_batch_idx: 31900, batch size: 50, loss[discriminator_loss=2.605, discriminator_real_loss=1.417, discriminator_fake_loss=1.188, generator_loss=30.01, generator_mel_loss=19.32, generator_kl_loss=2.039, generator_dur_loss=1.641, generator_adv_loss=2.232, generator_feat_match_loss=4.777, over 50.00 samples.], tot_loss[discriminator_loss=2.491, discriminator_real_loss=1.268, discriminator_fake_loss=1.223, generator_loss=30.61, generator_mel_loss=19.74, generator_kl_loss=2.017, generator_dur_loss=1.642, generator_adv_loss=2.307, generator_feat_match_loss=4.905, over 478.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-15 00:20:14,410 INFO [train.py:811] (3/4) Start epoch 864
2023-11-15 00:22:19,689 INFO [train.py:467] (3/4) Epoch 864, batch 19, global_batch_idx: 31950, batch size: 71, loss[discriminator_loss=2.539, discriminator_real_loss=1.289, discriminator_fake_loss=1.251, generator_loss=30.69, generator_mel_loss=19.98, generator_kl_loss=2.011, generator_dur_loss=1.614, generator_adv_loss=2.426, generator_feat_match_loss=4.656, over 71.00 samples.], tot_loss[discriminator_loss=2.544, discriminator_real_loss=1.288, discriminator_fake_loss=1.256, generator_loss=30.41, generator_mel_loss=19.98, generator_kl_loss=2.022, generator_dur_loss=1.626, generator_adv_loss=2.19, generator_feat_match_loss=4.594, over 1716.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 8.0
2023-11-15 00:23:47,350 INFO [train.py:811] (3/4) Start epoch 865
2023-11-15 00:26:58,646 INFO [train.py:467] (3/4) Epoch 865, batch 32, global_batch_idx: 32000, batch size: 126, loss[discriminator_loss=2.676, discriminator_real_loss=1.309, discriminator_fake_loss=1.368, generator_loss=30.17, generator_mel_loss=19.88, generator_kl_loss=2.019, generator_dur_loss=1.611, generator_adv_loss=2.258, generator_feat_match_loss=4.398, over 126.00 samples.], tot_loss[discriminator_loss=2.543, discriminator_real_loss=1.275, discriminator_fake_loss=1.269, generator_loss=30.78, generator_mel_loss=20.04, generator_kl_loss=2.009, generator_dur_loss=1.632, generator_adv_loss=2.292, generator_feat_match_loss=4.807, over 2647.00 samples.], cur_lr_g: 1.80e-04, cur_lr_d: 1.80e-04, grad_scale: 16.0
2023-11-15 00:26:58,648 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 00:27:09,992 INFO [train.py:517] (3/4) Epoch 865, validation: discriminator_loss=2.595, discriminator_real_loss=1.359, discriminator_fake_loss=1.236, generator_loss=31.08, generator_mel_loss=20.52, generator_kl_loss=2.151, generator_dur_loss=1.63, generator_adv_loss=2.244, generator_feat_match_loss=4.534, over 100.00 samples.
2023-11-15 00:27:09,993 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 00:27:30,610 INFO [train.py:811] (3/4) Start epoch 866
2023-11-15 00:31:05,951 INFO [train.py:811] (3/4) Start epoch 867
2023-11-15 00:32:09,224 INFO [train.py:467] (3/4) Epoch 867, batch 8, global_batch_idx: 32050, batch size: 58, loss[discriminator_loss=2.59, discriminator_real_loss=1.129, discriminator_fake_loss=1.462, generator_loss=30.39, generator_mel_loss=19.82, generator_kl_loss=2.005, generator_dur_loss=1.628, generator_adv_loss=2.201, generator_feat_match_loss=4.734, over 58.00 samples.], tot_loss[discriminator_loss=2.529, discriminator_real_loss=1.326, discriminator_fake_loss=1.204, generator_loss=30.82, generator_mel_loss=19.75, generator_kl_loss=2.004, generator_dur_loss=1.623, generator_adv_loss=2.372, generator_feat_match_loss=5.065, over 771.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:34:43,469 INFO [train.py:811] (3/4) Start epoch 868
2023-11-15 00:36:46,747 INFO [train.py:467] (3/4) Epoch 868, batch 21, global_batch_idx: 32100, batch size: 56, loss[discriminator_loss=2.619, discriminator_real_loss=1.294, discriminator_fake_loss=1.325, generator_loss=31.27, generator_mel_loss=20.54, generator_kl_loss=2.034, generator_dur_loss=1.647, generator_adv_loss=2.279, generator_feat_match_loss=4.766, over 56.00 samples.], tot_loss[discriminator_loss=2.557, discriminator_real_loss=1.305, discriminator_fake_loss=1.252, generator_loss=30.45, generator_mel_loss=19.97, generator_kl_loss=2.005, generator_dur_loss=1.634, generator_adv_loss=2.21, generator_feat_match_loss=4.631, over 1510.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:38:13,935 INFO [train.py:811] (3/4) Start epoch 869
2023-11-15 00:41:37,528 INFO [train.py:467] (3/4) Epoch 869, batch 34, global_batch_idx: 32150, batch size: 53, loss[discriminator_loss=2.496, discriminator_real_loss=1.316, discriminator_fake_loss=1.18, generator_loss=30.86, generator_mel_loss=19.93, generator_kl_loss=2.056, generator_dur_loss=1.643, generator_adv_loss=2.305, generator_feat_match_loss=4.926, over 53.00 samples.], tot_loss[discriminator_loss=2.558, discriminator_real_loss=1.293, discriminator_fake_loss=1.265, generator_loss=30.44, generator_mel_loss=19.89, generator_kl_loss=2.021, generator_dur_loss=1.631, generator_adv_loss=2.253, generator_feat_match_loss=4.652, over 2450.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:41:47,372 INFO [train.py:811] (3/4) Start epoch 870
2023-11-15 00:45:22,050 INFO [train.py:811] (3/4) Start epoch 871
2023-11-15 00:46:29,963 INFO [train.py:467] (3/4) Epoch 871, batch 10, global_batch_idx: 32200, batch size: 53, loss[discriminator_loss=2.48, discriminator_real_loss=1.237, discriminator_fake_loss=1.244, generator_loss=30.64, generator_mel_loss=19.95, generator_kl_loss=1.961, generator_dur_loss=1.658, generator_adv_loss=2.189, generator_feat_match_loss=4.887, over 53.00 samples.], tot_loss[discriminator_loss=2.531, discriminator_real_loss=1.292, discriminator_fake_loss=1.239, generator_loss=30.98, generator_mel_loss=20.26, generator_kl_loss=2.037, generator_dur_loss=1.632, generator_adv_loss=2.279, generator_feat_match_loss=4.769, over 760.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:46:29,964 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 00:46:40,912 INFO [train.py:517] (3/4) Epoch 871, validation: discriminator_loss=2.558, discriminator_real_loss=1.112, discriminator_fake_loss=1.446, generator_loss=31.29, generator_mel_loss=20.74, generator_kl_loss=2.289, generator_dur_loss=1.629, generator_adv_loss=1.83, generator_feat_match_loss=4.802, over 100.00 samples.
2023-11-15 00:46:40,913 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 00:49:04,280 INFO [train.py:811] (3/4) Start epoch 872
2023-11-15 00:51:22,598 INFO [train.py:467] (3/4) Epoch 872, batch 23, global_batch_idx: 32250, batch size: 73, loss[discriminator_loss=2.748, discriminator_real_loss=1.164, discriminator_fake_loss=1.584, generator_loss=30.22, generator_mel_loss=19.68, generator_kl_loss=1.973, generator_dur_loss=1.637, generator_adv_loss=2.168, generator_feat_match_loss=4.766, over 73.00 samples.], tot_loss[discriminator_loss=2.447, discriminator_real_loss=1.248, discriminator_fake_loss=1.199, generator_loss=31.06, generator_mel_loss=19.73, generator_kl_loss=1.994, generator_dur_loss=1.628, generator_adv_loss=2.459, generator_feat_match_loss=5.252, over 1884.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:52:32,063 INFO [train.py:811] (3/4) Start epoch 873
2023-11-15 00:56:02,960 INFO [train.py:467] (3/4) Epoch 873, batch 36, global_batch_idx: 32300, batch size: 76, loss[discriminator_loss=2.5, discriminator_real_loss=1.292, discriminator_fake_loss=1.208, generator_loss=30.58, generator_mel_loss=19.96, generator_kl_loss=2.102, generator_dur_loss=1.643, generator_adv_loss=2.188, generator_feat_match_loss=4.684, over 76.00 samples.], tot_loss[discriminator_loss=2.508, discriminator_real_loss=1.257, discriminator_fake_loss=1.251, generator_loss=30.43, generator_mel_loss=19.68, generator_kl_loss=2.022, generator_dur_loss=1.63, generator_adv_loss=2.281, generator_feat_match_loss=4.824, over 2776.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 00:56:03,709 INFO [train.py:811] (3/4) Start epoch 874
2023-11-15 00:59:41,414 INFO [train.py:811] (3/4) Start epoch 875
2023-11-15 01:01:11,185 INFO [train.py:467] (3/4) Epoch 875, batch 12, global_batch_idx: 32350, batch size: 69, loss[discriminator_loss=2.469, discriminator_real_loss=1.262, discriminator_fake_loss=1.207, generator_loss=29.97, generator_mel_loss=19.32, generator_kl_loss=2.06, generator_dur_loss=1.637, generator_adv_loss=2.172, generator_feat_match_loss=4.781, over 69.00 samples.], tot_loss[discriminator_loss=2.609, discriminator_real_loss=1.339, discriminator_fake_loss=1.27, generator_loss=30.7, generator_mel_loss=19.94, generator_kl_loss=2.02, generator_dur_loss=1.627, generator_adv_loss=2.336, generator_feat_match_loss=4.775, over 1002.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:03:10,040 INFO [train.py:811] (3/4) Start epoch 876
2023-11-15 01:05:38,800 INFO [train.py:467] (3/4) Epoch 876, batch 25, global_batch_idx: 32400, batch size: 153, loss[discriminator_loss=2.559, discriminator_real_loss=1.252, discriminator_fake_loss=1.308, generator_loss=30.81, generator_mel_loss=19.98, generator_kl_loss=2.06, generator_dur_loss=1.624, generator_adv_loss=2.246, generator_feat_match_loss=4.906, over 153.00 samples.], tot_loss[discriminator_loss=2.547, discriminator_real_loss=1.277, discriminator_fake_loss=1.271, generator_loss=30.74, generator_mel_loss=19.95, generator_kl_loss=2.044, generator_dur_loss=1.626, generator_adv_loss=2.268, generator_feat_match_loss=4.85, over 2123.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 01:05:38,802 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 01:05:49,911 INFO [train.py:517] (3/4) Epoch 876, validation: discriminator_loss=2.488, discriminator_real_loss=1.209, discriminator_fake_loss=1.279, generator_loss=31.4, generator_mel_loss=20.55, generator_kl_loss=2.287, generator_dur_loss=1.635, generator_adv_loss=2.072, generator_feat_match_loss=4.859, over 100.00 samples.
2023-11-15 01:05:49,912 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 01:06:55,281 INFO [train.py:811] (3/4) Start epoch 877
2023-11-15 01:10:28,235 INFO [train.py:811] (3/4) Start epoch 878
2023-11-15 01:10:48,332 INFO [train.py:467] (3/4) Epoch 878, batch 1, global_batch_idx: 32450, batch size: 64, loss[discriminator_loss=2.363, discriminator_real_loss=1.279, discriminator_fake_loss=1.085, generator_loss=31.71, generator_mel_loss=19.95, generator_kl_loss=2.049, generator_dur_loss=1.672, generator_adv_loss=2.518, generator_feat_match_loss=5.516, over 64.00 samples.], tot_loss[discriminator_loss=2.411, discriminator_real_loss=1.276, discriminator_fake_loss=1.136, generator_loss=31.17, generator_mel_loss=19.59, generator_kl_loss=2.022, generator_dur_loss=1.663, generator_adv_loss=2.541, generator_feat_match_loss=5.352, over 131.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:14:04,008 INFO [train.py:811] (3/4) Start epoch 879
2023-11-15 01:15:30,522 INFO [train.py:467] (3/4) Epoch 879, batch 14, global_batch_idx: 32500, batch size: 64, loss[discriminator_loss=2.477, discriminator_real_loss=1.306, discriminator_fake_loss=1.172, generator_loss=31.34, generator_mel_loss=20.28, generator_kl_loss=2.054, generator_dur_loss=1.668, generator_adv_loss=2.303, generator_feat_match_loss=5.027, over 64.00 samples.], tot_loss[discriminator_loss=2.548, discriminator_real_loss=1.295, discriminator_fake_loss=1.254, generator_loss=30.67, generator_mel_loss=20.01, generator_kl_loss=2.053, generator_dur_loss=1.635, generator_adv_loss=2.247, generator_feat_match_loss=4.725, over 1015.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:17:34,689 INFO [train.py:811] (3/4) Start epoch 880
2023-11-15 01:20:20,354 INFO [train.py:467] (3/4) Epoch 880, batch 27, global_batch_idx: 32550, batch size: 52, loss[discriminator_loss=2.52, discriminator_real_loss=1.333, discriminator_fake_loss=1.188, generator_loss=30.25, generator_mel_loss=19.92, generator_kl_loss=1.96, generator_dur_loss=1.654, generator_adv_loss=2.135, generator_feat_match_loss=4.578, over 52.00 samples.], tot_loss[discriminator_loss=2.451, discriminator_real_loss=1.233, discriminator_fake_loss=1.218, generator_loss=30.78, generator_mel_loss=19.66, generator_kl_loss=2.001, generator_dur_loss=1.632, generator_adv_loss=2.376, generator_feat_match_loss=5.111, over 2032.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:21:07,243 INFO [train.py:811] (3/4) Start epoch 881
2023-11-15 01:24:31,794 INFO [train.py:811] (3/4) Start epoch 882
2023-11-15 01:25:01,340 INFO [train.py:467] (3/4) Epoch 882, batch 3, global_batch_idx: 32600, batch size: 76, loss[discriminator_loss=2.432, discriminator_real_loss=1.258, discriminator_fake_loss=1.174, generator_loss=31.14, generator_mel_loss=19.65, generator_kl_loss=2.117, generator_dur_loss=1.658, generator_adv_loss=2.494, generator_feat_match_loss=5.215, over 76.00 samples.], tot_loss[discriminator_loss=2.528, discriminator_real_loss=1.315, discriminator_fake_loss=1.213, generator_loss=30.45, generator_mel_loss=19.56, generator_kl_loss=2.043, generator_dur_loss=1.647, generator_adv_loss=2.395, generator_feat_match_loss=4.806, over 242.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:25:01,341 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 01:25:13,537 INFO [train.py:517] (3/4) Epoch 882, validation: discriminator_loss=2.375, discriminator_real_loss=1.179, discriminator_fake_loss=1.196, generator_loss=32, generator_mel_loss=20.37, generator_kl_loss=2.176, generator_dur_loss=1.642, generator_adv_loss=2.394, generator_feat_match_loss=5.41, over 100.00 samples.
2023-11-15 01:25:13,538 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 01:28:19,329 INFO [train.py:811] (3/4) Start epoch 883
2023-11-15 01:29:54,451 INFO [train.py:467] (3/4) Epoch 883, batch 16, global_batch_idx: 32650, batch size: 65, loss[discriminator_loss=2.461, discriminator_real_loss=1.32, discriminator_fake_loss=1.141, generator_loss=30.52, generator_mel_loss=19.62, generator_kl_loss=1.994, generator_dur_loss=1.663, generator_adv_loss=2.291, generator_feat_match_loss=4.953, over 65.00 samples.], tot_loss[discriminator_loss=2.566, discriminator_real_loss=1.304, discriminator_fake_loss=1.262, generator_loss=30.55, generator_mel_loss=20.03, generator_kl_loss=2.045, generator_dur_loss=1.632, generator_adv_loss=2.235, generator_feat_match_loss=4.615, over 1087.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:31:43,295 INFO [train.py:811] (3/4) Start epoch 884
2023-11-15 01:34:31,866 INFO [train.py:467] (3/4) Epoch 884, batch 29, global_batch_idx: 32700, batch size: 67, loss[discriminator_loss=2.578, discriminator_real_loss=1.274, discriminator_fake_loss=1.305, generator_loss=30.01, generator_mel_loss=19.86, generator_kl_loss=2.003, generator_dur_loss=1.642, generator_adv_loss=2.191, generator_feat_match_loss=4.32, over 67.00 samples.], tot_loss[discriminator_loss=2.538, discriminator_real_loss=1.277, discriminator_fake_loss=1.261, generator_loss=30.64, generator_mel_loss=19.9, generator_kl_loss=2.005, generator_dur_loss=1.631, generator_adv_loss=2.305, generator_feat_match_loss=4.802, over 2275.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:35:17,577 INFO [train.py:811] (3/4) Start epoch 885
2023-11-15 01:38:53,597 INFO [train.py:811] (3/4) Start epoch 886
2023-11-15 01:39:37,940 INFO [train.py:467] (3/4) Epoch 886, batch 5, global_batch_idx: 32750, batch size: 61, loss[discriminator_loss=2.572, discriminator_real_loss=1.284, discriminator_fake_loss=1.288, generator_loss=30.41, generator_mel_loss=20.02, generator_kl_loss=1.959, generator_dur_loss=1.638, generator_adv_loss=2.156, generator_feat_match_loss=4.633, over 61.00 samples.], tot_loss[discriminator_loss=2.58, discriminator_real_loss=1.298, discriminator_fake_loss=1.281, generator_loss=30.49, generator_mel_loss=20.18, generator_kl_loss=2.013, generator_dur_loss=1.624, generator_adv_loss=2.198, generator_feat_match_loss=4.481, over 433.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:42:23,390 INFO [train.py:811] (3/4) Start epoch 887
2023-11-15 01:44:16,984 INFO [train.py:467] (3/4) Epoch 887, batch 18, global_batch_idx: 32800, batch size: 63, loss[discriminator_loss=2.48, discriminator_real_loss=1.327, discriminator_fake_loss=1.153, generator_loss=31.09, generator_mel_loss=20.09, generator_kl_loss=1.902, generator_dur_loss=1.6, generator_adv_loss=2.262, generator_feat_match_loss=5.23, over 63.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.308, discriminator_fake_loss=1.266, generator_loss=30.63, generator_mel_loss=20.08, generator_kl_loss=2.012, generator_dur_loss=1.626, generator_adv_loss=2.255, generator_feat_match_loss=4.652, over 1370.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 01:44:16,985 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 01:44:28,139 INFO [train.py:517] (3/4) Epoch 887, validation: discriminator_loss=2.489, discriminator_real_loss=1.192, discriminator_fake_loss=1.297, generator_loss=31.07, generator_mel_loss=20.3, generator_kl_loss=2.217, generator_dur_loss=1.644, generator_adv_loss=2.108, generator_feat_match_loss=4.801, over 100.00 samples.
2023-11-15 01:44:28,141 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 01:46:05,381 INFO [train.py:811] (3/4) Start epoch 888
2023-11-15 01:48:53,896 INFO [train.py:467] (3/4) Epoch 888, batch 31, global_batch_idx: 32850, batch size: 101, loss[discriminator_loss=2.48, discriminator_real_loss=1.337, discriminator_fake_loss=1.145, generator_loss=31.65, generator_mel_loss=20.2, generator_kl_loss=2.071, generator_dur_loss=1.624, generator_adv_loss=2.449, generator_feat_match_loss=5.305, over 101.00 samples.], tot_loss[discriminator_loss=2.524, discriminator_real_loss=1.288, discriminator_fake_loss=1.236, generator_loss=30.51, generator_mel_loss=19.88, generator_kl_loss=2.021, generator_dur_loss=1.629, generator_adv_loss=2.27, generator_feat_match_loss=4.708, over 2327.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:49:29,422 INFO [train.py:811] (3/4) Start epoch 889
2023-11-15 01:53:01,055 INFO [train.py:811] (3/4) Start epoch 890
2023-11-15 01:53:55,617 INFO [train.py:467] (3/4) Epoch 890, batch 7, global_batch_idx: 32900, batch size: 153, loss[discriminator_loss=2.531, discriminator_real_loss=1.157, discriminator_fake_loss=1.375, generator_loss=30.9, generator_mel_loss=19.9, generator_kl_loss=2.111, generator_dur_loss=1.599, generator_adv_loss=2.285, generator_feat_match_loss=5.008, over 153.00 samples.], tot_loss[discriminator_loss=2.527, discriminator_real_loss=1.254, discriminator_fake_loss=1.273, generator_loss=30.48, generator_mel_loss=19.76, generator_kl_loss=2.012, generator_dur_loss=1.624, generator_adv_loss=2.301, generator_feat_match_loss=4.781, over 651.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 01:56:34,001 INFO [train.py:811] (3/4) Start epoch 891
2023-11-15 01:58:40,563 INFO [train.py:467] (3/4) Epoch 891, batch 20, global_batch_idx: 32950, batch size: 67, loss[discriminator_loss=2.799, discriminator_real_loss=1.645, discriminator_fake_loss=1.154, generator_loss=30.36, generator_mel_loss=19.77, generator_kl_loss=2.089, generator_dur_loss=1.631, generator_adv_loss=2.455, generator_feat_match_loss=4.422, over 67.00 samples.], tot_loss[discriminator_loss=2.524, discriminator_real_loss=1.274, discriminator_fake_loss=1.25, generator_loss=30.78, generator_mel_loss=19.92, generator_kl_loss=2.012, generator_dur_loss=1.633, generator_adv_loss=2.342, generator_feat_match_loss=4.872, over 1532.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:00:13,505 INFO [train.py:811] (3/4) Start epoch 892
2023-11-15 02:03:30,996 INFO [train.py:467] (3/4) Epoch 892, batch 33, global_batch_idx: 33000, batch size: 50, loss[discriminator_loss=2.604, discriminator_real_loss=1.141, discriminator_fake_loss=1.463, generator_loss=31.82, generator_mel_loss=20.71, generator_kl_loss=1.924, generator_dur_loss=1.639, generator_adv_loss=2.436, generator_feat_match_loss=5.113, over 50.00 samples.], tot_loss[discriminator_loss=2.503, discriminator_real_loss=1.263, discriminator_fake_loss=1.239, generator_loss=30.77, generator_mel_loss=19.99, generator_kl_loss=2.027, generator_dur_loss=1.631, generator_adv_loss=2.293, generator_feat_match_loss=4.832, over 2454.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:03:30,997 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 02:03:42,568 INFO [train.py:517] (3/4) Epoch 892, validation: discriminator_loss=2.723, discriminator_real_loss=1.326, discriminator_fake_loss=1.397, generator_loss=30.8, generator_mel_loss=20.53, generator_kl_loss=2.251, generator_dur_loss=1.632, generator_adv_loss=1.933, generator_feat_match_loss=4.461, over 100.00 samples.
2023-11-15 02:03:42,570 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 02:03:58,486 INFO [train.py:811] (3/4) Start epoch 893
2023-11-15 02:07:32,130 INFO [train.py:811] (3/4) Start epoch 894
2023-11-15 02:08:30,723 INFO [train.py:467] (3/4) Epoch 894, batch 9, global_batch_idx: 33050, batch size: 76, loss[discriminator_loss=2.551, discriminator_real_loss=1.233, discriminator_fake_loss=1.317, generator_loss=30.61, generator_mel_loss=20.18, generator_kl_loss=2.08, generator_dur_loss=1.626, generator_adv_loss=2.127, generator_feat_match_loss=4.594, over 76.00 samples.], tot_loss[discriminator_loss=2.582, discriminator_real_loss=1.301, discriminator_fake_loss=1.282, generator_loss=30.4, generator_mel_loss=19.99, generator_kl_loss=2.023, generator_dur_loss=1.629, generator_adv_loss=2.17, generator_feat_match_loss=4.594, over 744.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:11:05,643 INFO [train.py:811] (3/4) Start epoch 895
2023-11-15 02:13:20,286 INFO [train.py:467] (3/4) Epoch 895, batch 22, global_batch_idx: 33100, batch size: 55, loss[discriminator_loss=2.641, discriminator_real_loss=1.242, discriminator_fake_loss=1.399, generator_loss=31.25, generator_mel_loss=20.45, generator_kl_loss=2.155, generator_dur_loss=1.692, generator_adv_loss=2.338, generator_feat_match_loss=4.621, over 55.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.295, discriminator_fake_loss=1.27, generator_loss=30.82, generator_mel_loss=20.21, generator_kl_loss=2.053, generator_dur_loss=1.633, generator_adv_loss=2.234, generator_feat_match_loss=4.686, over 1700.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:14:40,315 INFO [train.py:811] (3/4) Start epoch 896
2023-11-15 02:18:09,902 INFO [train.py:467] (3/4) Epoch 896, batch 35, global_batch_idx: 33150, batch size: 126, loss[discriminator_loss=2.57, discriminator_real_loss=1.239, discriminator_fake_loss=1.332, generator_loss=30.67, generator_mel_loss=19.92, generator_kl_loss=2.113, generator_dur_loss=1.613, generator_adv_loss=2.158, generator_feat_match_loss=4.867, over 126.00 samples.], tot_loss[discriminator_loss=2.564, discriminator_real_loss=1.291, discriminator_fake_loss=1.273, generator_loss=30.46, generator_mel_loss=19.82, generator_kl_loss=2, generator_dur_loss=1.629, generator_adv_loss=2.261, generator_feat_match_loss=4.752, over 2705.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:18:14,807 INFO [train.py:811] (3/4) Start epoch 897
2023-11-15 02:21:46,020 INFO [train.py:811] (3/4) Start epoch 898
2023-11-15 02:23:02,644 INFO [train.py:467] (3/4) Epoch 898, batch 11, global_batch_idx: 33200, batch size: 65, loss[discriminator_loss=2.678, discriminator_real_loss=1.344, discriminator_fake_loss=1.334, generator_loss=29.98, generator_mel_loss=20.07, generator_kl_loss=2.026, generator_dur_loss=1.619, generator_adv_loss=2.17, generator_feat_match_loss=4.094, over 65.00 samples.], tot_loss[discriminator_loss=2.568, discriminator_real_loss=1.29, discriminator_fake_loss=1.278, generator_loss=30.39, generator_mel_loss=19.87, generator_kl_loss=2.022, generator_dur_loss=1.632, generator_adv_loss=2.226, generator_feat_match_loss=4.643, over 774.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 02:23:02,646 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 02:23:13,663 INFO [train.py:517] (3/4) Epoch 898, validation: discriminator_loss=2.641, discriminator_real_loss=1.339, discriminator_fake_loss=1.302, generator_loss=31.78, generator_mel_loss=20.81, generator_kl_loss=2.316, generator_dur_loss=1.625, generator_adv_loss=2.175, generator_feat_match_loss=4.856, over 100.00 samples.
2023-11-15 02:23:13,665 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 02:25:27,278 INFO [train.py:811] (3/4) Start epoch 899
2023-11-15 02:27:58,744 INFO [train.py:467] (3/4) Epoch 899, batch 24, global_batch_idx: 33250, batch size: 58, loss[discriminator_loss=2.615, discriminator_real_loss=1.283, discriminator_fake_loss=1.332, generator_loss=30.13, generator_mel_loss=19.64, generator_kl_loss=1.992, generator_dur_loss=1.635, generator_adv_loss=2.15, generator_feat_match_loss=4.711, over 58.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.309, discriminator_fake_loss=1.269, generator_loss=30.51, generator_mel_loss=20.1, generator_kl_loss=2.024, generator_dur_loss=1.629, generator_adv_loss=2.166, generator_feat_match_loss=4.595, over 1968.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 02:29:04,394 INFO [train.py:811] (3/4) Start epoch 900
2023-11-15 02:32:36,637 INFO [train.py:811] (3/4) Start epoch 901
2023-11-15 02:32:54,377 INFO [train.py:467] (3/4) Epoch 901, batch 0, global_batch_idx: 33300, batch size: 71, loss[discriminator_loss=2.525, discriminator_real_loss=1.172, discriminator_fake_loss=1.354, generator_loss=30.83, generator_mel_loss=20.07, generator_kl_loss=1.979, generator_dur_loss=1.625, generator_adv_loss=2.23, generator_feat_match_loss=4.93, over 71.00 samples.], tot_loss[discriminator_loss=2.525, discriminator_real_loss=1.172, discriminator_fake_loss=1.354, generator_loss=30.83, generator_mel_loss=20.07, generator_kl_loss=1.979, generator_dur_loss=1.625, generator_adv_loss=2.23, generator_feat_match_loss=4.93, over 71.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 02:36:10,110 INFO [train.py:811] (3/4) Start epoch 902
2023-11-15 02:37:30,509 INFO [train.py:467] (3/4) Epoch 902, batch 13, global_batch_idx: 33350, batch size: 90, loss[discriminator_loss=2.617, discriminator_real_loss=1.394, discriminator_fake_loss=1.223, generator_loss=30.77, generator_mel_loss=20.04, generator_kl_loss=2.091, generator_dur_loss=1.631, generator_adv_loss=2.4, generator_feat_match_loss=4.605, over 90.00 samples.], tot_loss[discriminator_loss=2.513, discriminator_real_loss=1.269, discriminator_fake_loss=1.244, generator_loss=30.84, generator_mel_loss=20.06, generator_kl_loss=2.011, generator_dur_loss=1.633, generator_adv_loss=2.301, generator_feat_match_loss=4.835, over 1031.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 02:39:44,462 INFO [train.py:811] (3/4) Start epoch 903
2023-11-15 02:42:16,466 INFO [train.py:467] (3/4) Epoch 903, batch 26, global_batch_idx: 33400, batch size: 81, loss[discriminator_loss=2.395, discriminator_real_loss=1.119, discriminator_fake_loss=1.274, generator_loss=31.4, generator_mel_loss=19.85, generator_kl_loss=1.936, generator_dur_loss=1.628, generator_adv_loss=2.479, generator_feat_match_loss=5.508, over 81.00 samples.], tot_loss[discriminator_loss=2.463, discriminator_real_loss=1.238, discriminator_fake_loss=1.225, generator_loss=30.88, generator_mel_loss=19.67, generator_kl_loss=1.999, generator_dur_loss=1.625, generator_adv_loss=2.405, generator_feat_match_loss=5.181, over 2104.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:42:16,468 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 02:42:27,432 INFO [train.py:517] (3/4) Epoch 903, validation: discriminator_loss=2.411, discriminator_real_loss=1.1, discriminator_fake_loss=1.311, generator_loss=31.57, generator_mel_loss=20.23, generator_kl_loss=2.246, generator_dur_loss=1.631, generator_adv_loss=2.149, generator_feat_match_loss=5.312, over 100.00 samples.
2023-11-15 02:42:27,433 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 02:43:23,856 INFO [train.py:811] (3/4) Start epoch 904
2023-11-15 02:46:56,027 INFO [train.py:811] (3/4) Start epoch 905
2023-11-15 02:47:18,808 INFO [train.py:467] (3/4) Epoch 905, batch 2, global_batch_idx: 33450, batch size: 52, loss[discriminator_loss=2.418, discriminator_real_loss=1.293, discriminator_fake_loss=1.126, generator_loss=31.49, generator_mel_loss=20.17, generator_kl_loss=2.104, generator_dur_loss=1.64, generator_adv_loss=2.395, generator_feat_match_loss=5.18, over 52.00 samples.], tot_loss[discriminator_loss=2.543, discriminator_real_loss=1.341, discriminator_fake_loss=1.202, generator_loss=31.04, generator_mel_loss=20.06, generator_kl_loss=2.07, generator_dur_loss=1.642, generator_adv_loss=2.373, generator_feat_match_loss=4.892, over 170.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:50:24,479 INFO [train.py:811] (3/4) Start epoch 906
2023-11-15 02:52:02,441 INFO [train.py:467] (3/4) Epoch 906, batch 15, global_batch_idx: 33500, batch size: 59, loss[discriminator_loss=2.551, discriminator_real_loss=1.402, discriminator_fake_loss=1.149, generator_loss=30.75, generator_mel_loss=20.09, generator_kl_loss=2.031, generator_dur_loss=1.624, generator_adv_loss=2.254, generator_feat_match_loss=4.75, over 59.00 samples.], tot_loss[discriminator_loss=2.531, discriminator_real_loss=1.269, discriminator_fake_loss=1.262, generator_loss=30.68, generator_mel_loss=19.94, generator_kl_loss=2.02, generator_dur_loss=1.63, generator_adv_loss=2.291, generator_feat_match_loss=4.795, over 1227.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:53:54,395 INFO [train.py:811] (3/4) Start epoch 907
2023-11-15 02:56:48,354 INFO [train.py:467] (3/4) Epoch 907, batch 28, global_batch_idx: 33550, batch size: 55, loss[discriminator_loss=2.605, discriminator_real_loss=1.29, discriminator_fake_loss=1.316, generator_loss=30.33, generator_mel_loss=19.99, generator_kl_loss=2.021, generator_dur_loss=1.637, generator_adv_loss=2.332, generator_feat_match_loss=4.352, over 55.00 samples.], tot_loss[discriminator_loss=2.6, discriminator_real_loss=1.312, discriminator_fake_loss=1.288, generator_loss=30.59, generator_mel_loss=20.15, generator_kl_loss=2.032, generator_dur_loss=1.627, generator_adv_loss=2.19, generator_feat_match_loss=4.581, over 2296.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 8.0
2023-11-15 02:57:26,622 INFO [train.py:811] (3/4) Start epoch 908
2023-11-15 03:00:55,857 INFO [train.py:811] (3/4) Start epoch 909
2023-11-15 03:01:33,973 INFO [train.py:467] (3/4) Epoch 909, batch 4, global_batch_idx: 33600, batch size: 61, loss[discriminator_loss=2.516, discriminator_real_loss=1.247, discriminator_fake_loss=1.268, generator_loss=29.71, generator_mel_loss=19.56, generator_kl_loss=2.004, generator_dur_loss=1.637, generator_adv_loss=2.035, generator_feat_match_loss=4.469, over 61.00 samples.], tot_loss[discriminator_loss=2.504, discriminator_real_loss=1.268, discriminator_fake_loss=1.235, generator_loss=30.45, generator_mel_loss=19.89, generator_kl_loss=2.005, generator_dur_loss=1.636, generator_adv_loss=2.211, generator_feat_match_loss=4.705, over 337.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 03:01:33,975 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 03:01:45,803 INFO [train.py:517] (3/4) Epoch 909, validation: discriminator_loss=2.46, discriminator_real_loss=1.126, discriminator_fake_loss=1.334, generator_loss=30.96, generator_mel_loss=20.45, generator_kl_loss=2.217, generator_dur_loss=1.627, generator_adv_loss=1.928, generator_feat_match_loss=4.737, over 100.00 samples.
2023-11-15 03:01:45,804 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 03:04:43,867 INFO [train.py:811] (3/4) Start epoch 910
2023-11-15 03:06:34,432 INFO [train.py:467] (3/4) Epoch 910, batch 17, global_batch_idx: 33650, batch size: 90, loss[discriminator_loss=2.406, discriminator_real_loss=1.172, discriminator_fake_loss=1.235, generator_loss=30.76, generator_mel_loss=19.7, generator_kl_loss=1.972, generator_dur_loss=1.602, generator_adv_loss=2.293, generator_feat_match_loss=5.195, over 90.00 samples.], tot_loss[discriminator_loss=2.504, discriminator_real_loss=1.274, discriminator_fake_loss=1.23, generator_loss=30.84, generator_mel_loss=19.88, generator_kl_loss=2.021, generator_dur_loss=1.628, generator_adv_loss=2.331, generator_feat_match_loss=4.978, over 1426.00 samples.], cur_lr_g: 1.79e-04, cur_lr_d: 1.79e-04, grad_scale: 16.0
2023-11-15 03:08:17,633 INFO [train.py:811] (3/4) Start epoch 911
2023-11-15 03:11:20,829 INFO [train.py:467] (3/4) Epoch 911, batch 30, global_batch_idx: 33700, batch size: 90, loss[discriminator_loss=2.299, discriminator_real_loss=1.191, discriminator_fake_loss=1.107, generator_loss=31.62, generator_mel_loss=19.64, generator_kl_loss=2.044, generator_dur_loss=1.638, generator_adv_loss=2.6, generator_feat_match_loss=5.695, over 90.00 samples.], tot_loss[discriminator_loss=2.473, discriminator_real_loss=1.248, discriminator_fake_loss=1.225, generator_loss=30.86, generator_mel_loss=19.8, generator_kl_loss=2.007, generator_dur_loss=1.63, generator_adv_loss=2.383, generator_feat_match_loss=5.034, over 2142.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:11:55,064 INFO [train.py:811] (3/4) Start epoch 912
2023-11-15 03:15:23,974 INFO [train.py:811] (3/4) Start epoch 913
2023-11-15 03:16:10,075 INFO [train.py:467] (3/4) Epoch 913, batch 6, global_batch_idx: 33750, batch size: 52, loss[discriminator_loss=2.572, discriminator_real_loss=1.443, discriminator_fake_loss=1.129, generator_loss=29.59, generator_mel_loss=19.33, generator_kl_loss=2.061, generator_dur_loss=1.621, generator_adv_loss=2.145, generator_feat_match_loss=4.426, over 52.00 samples.], tot_loss[discriminator_loss=2.52, discriminator_real_loss=1.276, discriminator_fake_loss=1.244, generator_loss=30.39, generator_mel_loss=19.92, generator_kl_loss=2.048, generator_dur_loss=1.625, generator_adv_loss=2.212, generator_feat_match_loss=4.584, over 492.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:18:52,872 INFO [train.py:811] (3/4) Start epoch 914
2023-11-15 03:20:58,332 INFO [train.py:467] (3/4) Epoch 914, batch 19, global_batch_idx: 33800, batch size: 58, loss[discriminator_loss=2.574, discriminator_real_loss=1.389, discriminator_fake_loss=1.187, generator_loss=30.64, generator_mel_loss=20.06, generator_kl_loss=2.037, generator_dur_loss=1.63, generator_adv_loss=2.197, generator_feat_match_loss=4.723, over 58.00 samples.], tot_loss[discriminator_loss=2.565, discriminator_real_loss=1.309, discriminator_fake_loss=1.256, generator_loss=30.83, generator_mel_loss=20.18, generator_kl_loss=2.029, generator_dur_loss=1.62, generator_adv_loss=2.239, generator_feat_match_loss=4.758, over 1594.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:20:58,334 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 03:21:09,508 INFO [train.py:517] (3/4) Epoch 914, validation: discriminator_loss=2.618, discriminator_real_loss=1.258, discriminator_fake_loss=1.361, generator_loss=31.84, generator_mel_loss=21.04, generator_kl_loss=2.192, generator_dur_loss=1.622, generator_adv_loss=2.014, generator_feat_match_loss=4.965, over 100.00 samples.
2023-11-15 03:21:09,509 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 03:22:36,852 INFO [train.py:811] (3/4) Start epoch 915
2023-11-15 03:25:40,315 INFO [train.py:467] (3/4) Epoch 915, batch 32, global_batch_idx: 33850, batch size: 50, loss[discriminator_loss=2.621, discriminator_real_loss=1.298, discriminator_fake_loss=1.324, generator_loss=30.52, generator_mel_loss=19.66, generator_kl_loss=1.984, generator_dur_loss=1.634, generator_adv_loss=2.465, generator_feat_match_loss=4.777, over 50.00 samples.], tot_loss[discriminator_loss=2.636, discriminator_real_loss=1.361, discriminator_fake_loss=1.276, generator_loss=30.72, generator_mel_loss=19.74, generator_kl_loss=2.02, generator_dur_loss=1.627, generator_adv_loss=2.363, generator_feat_match_loss=4.962, over 2251.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:26:08,758 INFO [train.py:811] (3/4) Start epoch 916
2023-11-15 03:29:46,099 INFO [train.py:811] (3/4) Start epoch 917
2023-11-15 03:30:40,659 INFO [train.py:467] (3/4) Epoch 917, batch 8, global_batch_idx: 33900, batch size: 90, loss[discriminator_loss=2.562, discriminator_real_loss=1.366, discriminator_fake_loss=1.197, generator_loss=30.45, generator_mel_loss=20.09, generator_kl_loss=1.99, generator_dur_loss=1.62, generator_adv_loss=2.141, generator_feat_match_loss=4.609, over 90.00 samples.], tot_loss[discriminator_loss=2.598, discriminator_real_loss=1.329, discriminator_fake_loss=1.269, generator_loss=30.6, generator_mel_loss=20.17, generator_kl_loss=2.046, generator_dur_loss=1.636, generator_adv_loss=2.209, generator_feat_match_loss=4.541, over 607.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:33:14,261 INFO [train.py:811] (3/4) Start epoch 918
2023-11-15 03:35:26,041 INFO [train.py:467] (3/4) Epoch 918, batch 21, global_batch_idx: 33950, batch size: 153, loss[discriminator_loss=2.559, discriminator_real_loss=1.232, discriminator_fake_loss=1.326, generator_loss=30.73, generator_mel_loss=20, generator_kl_loss=2.02, generator_dur_loss=1.607, generator_adv_loss=2.236, generator_feat_match_loss=4.863, over 153.00 samples.], tot_loss[discriminator_loss=2.615, discriminator_real_loss=1.323, discriminator_fake_loss=1.292, generator_loss=30.48, generator_mel_loss=20.15, generator_kl_loss=2.018, generator_dur_loss=1.627, generator_adv_loss=2.162, generator_feat_match_loss=4.525, over 1721.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:36:48,664 INFO [train.py:811] (3/4) Start epoch 919
2023-11-15 03:40:07,615 INFO [train.py:467] (3/4) Epoch 919, batch 34, global_batch_idx: 34000, batch size: 126, loss[discriminator_loss=2.428, discriminator_real_loss=1.168, discriminator_fake_loss=1.26, generator_loss=30.63, generator_mel_loss=19.61, generator_kl_loss=1.909, generator_dur_loss=1.631, generator_adv_loss=2.199, generator_feat_match_loss=5.281, over 126.00 samples.], tot_loss[discriminator_loss=2.595, discriminator_real_loss=1.295, discriminator_fake_loss=1.3, generator_loss=30.59, generator_mel_loss=19.98, generator_kl_loss=2.013, generator_dur_loss=1.63, generator_adv_loss=2.245, generator_feat_match_loss=4.719, over 2656.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 03:40:07,617 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 03:40:18,697 INFO [train.py:517] (3/4) Epoch 919, validation: discriminator_loss=2.484, discriminator_real_loss=1.073, discriminator_fake_loss=1.411, generator_loss=31.27, generator_mel_loss=20.37, generator_kl_loss=2.23, generator_dur_loss=1.627, generator_adv_loss=1.906, generator_feat_match_loss=5.136, over 100.00 samples.
2023-11-15 03:40:18,698 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 03:40:30,315 INFO [train.py:811] (3/4) Start epoch 920
2023-11-15 03:44:00,678 INFO [train.py:811] (3/4) Start epoch 921
2023-11-15 03:45:12,072 INFO [train.py:467] (3/4) Epoch 921, batch 10, global_batch_idx: 34050, batch size: 76, loss[discriminator_loss=2.613, discriminator_real_loss=1.197, discriminator_fake_loss=1.417, generator_loss=30.69, generator_mel_loss=20.04, generator_kl_loss=2.054, generator_dur_loss=1.622, generator_adv_loss=2.234, generator_feat_match_loss=4.742, over 76.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.291, discriminator_fake_loss=1.283, generator_loss=30.92, generator_mel_loss=20.06, generator_kl_loss=2.029, generator_dur_loss=1.625, generator_adv_loss=2.293, generator_feat_match_loss=4.91, over 869.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 03:47:33,987 INFO [train.py:811] (3/4) Start epoch 922
2023-11-15 03:49:50,858 INFO [train.py:467] (3/4) Epoch 922, batch 23, global_batch_idx: 34100, batch size: 61, loss[discriminator_loss=2.648, discriminator_real_loss=1.396, discriminator_fake_loss=1.253, generator_loss=29.92, generator_mel_loss=19.62, generator_kl_loss=1.984, generator_dur_loss=1.651, generator_adv_loss=2.297, generator_feat_match_loss=4.367, over 61.00 samples.], tot_loss[discriminator_loss=2.59, discriminator_real_loss=1.324, discriminator_fake_loss=1.266, generator_loss=30.55, generator_mel_loss=19.7, generator_kl_loss=2.004, generator_dur_loss=1.63, generator_adv_loss=2.326, generator_feat_match_loss=4.887, over 1754.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:51:01,263 INFO [train.py:811] (3/4) Start epoch 923
2023-11-15 03:54:32,361 INFO [train.py:467] (3/4) Epoch 923, batch 36, global_batch_idx: 34150, batch size: 71, loss[discriminator_loss=2.504, discriminator_real_loss=1.386, discriminator_fake_loss=1.117, generator_loss=30.25, generator_mel_loss=19.61, generator_kl_loss=2.061, generator_dur_loss=1.629, generator_adv_loss=2.268, generator_feat_match_loss=4.68, over 71.00 samples.], tot_loss[discriminator_loss=2.558, discriminator_real_loss=1.301, discriminator_fake_loss=1.257, generator_loss=30.58, generator_mel_loss=19.97, generator_kl_loss=2.016, generator_dur_loss=1.626, generator_adv_loss=2.243, generator_feat_match_loss=4.724, over 2892.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:54:33,027 INFO [train.py:811] (3/4) Start epoch 924
2023-11-15 03:58:05,046 INFO [train.py:811] (3/4) Start epoch 925
2023-11-15 03:59:25,747 INFO [train.py:467] (3/4) Epoch 925, batch 12, global_batch_idx: 34200, batch size: 50, loss[discriminator_loss=2.426, discriminator_real_loss=1.303, discriminator_fake_loss=1.123, generator_loss=31.07, generator_mel_loss=19.69, generator_kl_loss=2.028, generator_dur_loss=1.619, generator_adv_loss=2.418, generator_feat_match_loss=5.32, over 50.00 samples.], tot_loss[discriminator_loss=2.491, discriminator_real_loss=1.278, discriminator_fake_loss=1.212, generator_loss=30.98, generator_mel_loss=19.91, generator_kl_loss=2.034, generator_dur_loss=1.628, generator_adv_loss=2.365, generator_feat_match_loss=5.04, over 889.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 03:59:25,748 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 03:59:36,632 INFO [train.py:517] (3/4) Epoch 925, validation: discriminator_loss=2.463, discriminator_real_loss=1.216, discriminator_fake_loss=1.248, generator_loss=31.49, generator_mel_loss=20.59, generator_kl_loss=2.14, generator_dur_loss=1.643, generator_adv_loss=2.13, generator_feat_match_loss=4.99, over 100.00 samples.
2023-11-15 03:59:36,633 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 04:01:46,011 INFO [train.py:811] (3/4) Start epoch 926
2023-11-15 04:04:18,713 INFO [train.py:467] (3/4) Epoch 926, batch 25, global_batch_idx: 34250, batch size: 153, loss[discriminator_loss=2.539, discriminator_real_loss=1.344, discriminator_fake_loss=1.196, generator_loss=30.69, generator_mel_loss=19.81, generator_kl_loss=2.142, generator_dur_loss=1.591, generator_adv_loss=2.398, generator_feat_match_loss=4.758, over 153.00 samples.], tot_loss[discriminator_loss=2.542, discriminator_real_loss=1.277, discriminator_fake_loss=1.265, generator_loss=30.79, generator_mel_loss=19.91, generator_kl_loss=2.032, generator_dur_loss=1.626, generator_adv_loss=2.307, generator_feat_match_loss=4.921, over 2041.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:05:19,674 INFO [train.py:811] (3/4) Start epoch 927
2023-11-15 04:08:52,730 INFO [train.py:811] (3/4) Start epoch 928
2023-11-15 04:09:15,310 INFO [train.py:467] (3/4) Epoch 928, batch 1, global_batch_idx: 34300, batch size: 76, loss[discriminator_loss=2.461, discriminator_real_loss=1.22, discriminator_fake_loss=1.241, generator_loss=30.7, generator_mel_loss=19.91, generator_kl_loss=1.989, generator_dur_loss=1.638, generator_adv_loss=2.254, generator_feat_match_loss=4.902, over 76.00 samples.], tot_loss[discriminator_loss=2.466, discriminator_real_loss=1.252, discriminator_fake_loss=1.213, generator_loss=31.08, generator_mel_loss=20.06, generator_kl_loss=2.022, generator_dur_loss=1.631, generator_adv_loss=2.327, generator_feat_match_loss=5.041, over 186.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:12:24,470 INFO [train.py:811] (3/4) Start epoch 929
2023-11-15 04:13:53,612 INFO [train.py:467] (3/4) Epoch 929, batch 14, global_batch_idx: 34350, batch size: 73, loss[discriminator_loss=2.568, discriminator_real_loss=1.176, discriminator_fake_loss=1.393, generator_loss=30.05, generator_mel_loss=19.47, generator_kl_loss=2.003, generator_dur_loss=1.612, generator_adv_loss=2.232, generator_feat_match_loss=4.734, over 73.00 samples.], tot_loss[discriminator_loss=2.47, discriminator_real_loss=1.221, discriminator_fake_loss=1.248, generator_loss=30.64, generator_mel_loss=19.55, generator_kl_loss=2.011, generator_dur_loss=1.625, generator_adv_loss=2.359, generator_feat_match_loss=5.1, over 996.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:15:57,093 INFO [train.py:811] (3/4) Start epoch 930
2023-11-15 04:18:36,052 INFO [train.py:467] (3/4) Epoch 930, batch 27, global_batch_idx: 34400, batch size: 60, loss[discriminator_loss=2.566, discriminator_real_loss=1.348, discriminator_fake_loss=1.218, generator_loss=29.84, generator_mel_loss=19.69, generator_kl_loss=2.023, generator_dur_loss=1.64, generator_adv_loss=2.137, generator_feat_match_loss=4.348, over 60.00 samples.], tot_loss[discriminator_loss=2.548, discriminator_real_loss=1.303, discriminator_fake_loss=1.245, generator_loss=30.52, generator_mel_loss=19.99, generator_kl_loss=2.004, generator_dur_loss=1.623, generator_adv_loss=2.2, generator_feat_match_loss=4.707, over 2049.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 04:18:36,054 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 04:18:47,234 INFO [train.py:517] (3/4) Epoch 930, validation: discriminator_loss=2.608, discriminator_real_loss=1.257, discriminator_fake_loss=1.352, generator_loss=31.26, generator_mel_loss=20.85, generator_kl_loss=2.339, generator_dur_loss=1.624, generator_adv_loss=1.921, generator_feat_match_loss=4.522, over 100.00 samples.
2023-11-15 04:18:47,235 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 04:19:39,986 INFO [train.py:811] (3/4) Start epoch 931
2023-11-15 04:23:08,405 INFO [train.py:811] (3/4) Start epoch 932
2023-11-15 04:23:37,121 INFO [train.py:467] (3/4) Epoch 932, batch 3, global_batch_idx: 34450, batch size: 73, loss[discriminator_loss=2.516, discriminator_real_loss=1.322, discriminator_fake_loss=1.192, generator_loss=30.94, generator_mel_loss=19.99, generator_kl_loss=2.029, generator_dur_loss=1.617, generator_adv_loss=2.369, generator_feat_match_loss=4.941, over 73.00 samples.], tot_loss[discriminator_loss=2.622, discriminator_real_loss=1.301, discriminator_fake_loss=1.321, generator_loss=30.62, generator_mel_loss=20.08, generator_kl_loss=1.979, generator_dur_loss=1.62, generator_adv_loss=2.244, generator_feat_match_loss=4.698, over 251.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 04:26:32,041 INFO [train.py:811] (3/4) Start epoch 933
2023-11-15 04:28:17,118 INFO [train.py:467] (3/4) Epoch 933, batch 16, global_batch_idx: 34500, batch size: 51, loss[discriminator_loss=2.469, discriminator_real_loss=1.33, discriminator_fake_loss=1.14, generator_loss=31.2, generator_mel_loss=20.25, generator_kl_loss=2.161, generator_dur_loss=1.635, generator_adv_loss=2.299, generator_feat_match_loss=4.863, over 51.00 samples.], tot_loss[discriminator_loss=2.499, discriminator_real_loss=1.261, discriminator_fake_loss=1.238, generator_loss=30.62, generator_mel_loss=19.75, generator_kl_loss=2.024, generator_dur_loss=1.618, generator_adv_loss=2.309, generator_feat_match_loss=4.916, over 1254.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:30:00,583 INFO [train.py:811] (3/4) Start epoch 934
2023-11-15 04:32:51,934 INFO [train.py:467] (3/4) Epoch 934, batch 29, global_batch_idx: 34550, batch size: 153, loss[discriminator_loss=2.426, discriminator_real_loss=1.193, discriminator_fake_loss=1.231, generator_loss=31.29, generator_mel_loss=19.48, generator_kl_loss=2.081, generator_dur_loss=1.599, generator_adv_loss=2.611, generator_feat_match_loss=5.523, over 153.00 samples.], tot_loss[discriminator_loss=2.495, discriminator_real_loss=1.26, discriminator_fake_loss=1.235, generator_loss=30.97, generator_mel_loss=19.75, generator_kl_loss=2, generator_dur_loss=1.626, generator_adv_loss=2.411, generator_feat_match_loss=5.182, over 2200.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:33:31,336 INFO [train.py:811] (3/4) Start epoch 935
2023-11-15 04:36:59,727 INFO [train.py:811] (3/4) Start epoch 936
2023-11-15 04:37:38,993 INFO [train.py:467] (3/4) Epoch 936, batch 5, global_batch_idx: 34600, batch size: 53, loss[discriminator_loss=2.549, discriminator_real_loss=1.355, discriminator_fake_loss=1.193, generator_loss=29.75, generator_mel_loss=19.69, generator_kl_loss=1.963, generator_dur_loss=1.641, generator_adv_loss=2.115, generator_feat_match_loss=4.348, over 53.00 samples.], tot_loss[discriminator_loss=2.588, discriminator_real_loss=1.302, discriminator_fake_loss=1.286, generator_loss=30.33, generator_mel_loss=19.95, generator_kl_loss=2.037, generator_dur_loss=1.638, generator_adv_loss=2.14, generator_feat_match_loss=4.564, over 382.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:37:38,994 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 04:37:51,275 INFO [train.py:517] (3/4) Epoch 936, validation: discriminator_loss=2.564, discriminator_real_loss=1.232, discriminator_fake_loss=1.332, generator_loss=31.25, generator_mel_loss=20.65, generator_kl_loss=2.279, generator_dur_loss=1.634, generator_adv_loss=1.962, generator_feat_match_loss=4.719, over 100.00 samples.
2023-11-15 04:37:51,276 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 04:40:43,420 INFO [train.py:811] (3/4) Start epoch 937
2023-11-15 04:42:26,959 INFO [train.py:467] (3/4) Epoch 937, batch 18, global_batch_idx: 34650, batch size: 54, loss[discriminator_loss=2.592, discriminator_real_loss=1.314, discriminator_fake_loss=1.277, generator_loss=30.48, generator_mel_loss=20.01, generator_kl_loss=2.031, generator_dur_loss=1.632, generator_adv_loss=2.131, generator_feat_match_loss=4.668, over 54.00 samples.], tot_loss[discriminator_loss=2.586, discriminator_real_loss=1.297, discriminator_fake_loss=1.289, generator_loss=30.34, generator_mel_loss=20, generator_kl_loss=1.996, generator_dur_loss=1.631, generator_adv_loss=2.163, generator_feat_match_loss=4.547, over 1222.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:44:17,007 INFO [train.py:811] (3/4) Start epoch 938
2023-11-15 04:47:22,794 INFO [train.py:467] (3/4) Epoch 938, batch 31, global_batch_idx: 34700, batch size: 69, loss[discriminator_loss=2.441, discriminator_real_loss=1.35, discriminator_fake_loss=1.093, generator_loss=31.62, generator_mel_loss=20.09, generator_kl_loss=2.114, generator_dur_loss=1.613, generator_adv_loss=2.449, generator_feat_match_loss=5.352, over 69.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.298, discriminator_fake_loss=1.241, generator_loss=30.72, generator_mel_loss=19.86, generator_kl_loss=2.004, generator_dur_loss=1.628, generator_adv_loss=2.332, generator_feat_match_loss=4.898, over 2326.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:47:49,225 INFO [train.py:811] (3/4) Start epoch 939
2023-11-15 04:51:21,440 INFO [train.py:811] (3/4) Start epoch 940
2023-11-15 04:52:17,835 INFO [train.py:467] (3/4) Epoch 940, batch 7, global_batch_idx: 34750, batch size: 56, loss[discriminator_loss=2.539, discriminator_real_loss=1.311, discriminator_fake_loss=1.229, generator_loss=30.4, generator_mel_loss=19.62, generator_kl_loss=2.079, generator_dur_loss=1.658, generator_adv_loss=2.396, generator_feat_match_loss=4.652, over 56.00 samples.], tot_loss[discriminator_loss=2.523, discriminator_real_loss=1.25, discriminator_fake_loss=1.273, generator_loss=30.84, generator_mel_loss=20.04, generator_kl_loss=2.037, generator_dur_loss=1.627, generator_adv_loss=2.255, generator_feat_match_loss=4.879, over 613.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 04:54:54,434 INFO [train.py:811] (3/4) Start epoch 941
2023-11-15 04:57:01,270 INFO [train.py:467] (3/4) Epoch 941, batch 20, global_batch_idx: 34800, batch size: 65, loss[discriminator_loss=2.816, discriminator_real_loss=1.468, discriminator_fake_loss=1.348, generator_loss=30.23, generator_mel_loss=19.53, generator_kl_loss=1.951, generator_dur_loss=1.601, generator_adv_loss=2.496, generator_feat_match_loss=4.66, over 65.00 samples.], tot_loss[discriminator_loss=2.562, discriminator_real_loss=1.276, discriminator_fake_loss=1.285, generator_loss=30.77, generator_mel_loss=19.66, generator_kl_loss=1.977, generator_dur_loss=1.628, generator_adv_loss=2.368, generator_feat_match_loss=5.137, over 1566.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 04:57:01,272 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 04:57:12,268 INFO [train.py:517] (3/4) Epoch 941, validation: discriminator_loss=2.681, discriminator_real_loss=1.534, discriminator_fake_loss=1.147, generator_loss=31.02, generator_mel_loss=20.36, generator_kl_loss=2.139, generator_dur_loss=1.62, generator_adv_loss=2.381, generator_feat_match_loss=4.521, over 100.00 samples.
2023-11-15 04:57:12,269 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 04:58:42,422 INFO [train.py:811] (3/4) Start epoch 942
2023-11-15 05:02:02,345 INFO [train.py:467] (3/4) Epoch 942, batch 33, global_batch_idx: 34850, batch size: 79, loss[discriminator_loss=2.607, discriminator_real_loss=1.317, discriminator_fake_loss=1.29, generator_loss=30.18, generator_mel_loss=19.74, generator_kl_loss=1.949, generator_dur_loss=1.629, generator_adv_loss=2.25, generator_feat_match_loss=4.613, over 79.00 samples.], tot_loss[discriminator_loss=2.583, discriminator_real_loss=1.307, discriminator_fake_loss=1.277, generator_loss=30.46, generator_mel_loss=20, generator_kl_loss=2.016, generator_dur_loss=1.619, generator_adv_loss=2.187, generator_feat_match_loss=4.641, over 2764.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:02:16,222 INFO [train.py:811] (3/4) Start epoch 943
2023-11-15 05:05:42,541 INFO [train.py:811] (3/4) Start epoch 944
2023-11-15 05:06:41,038 INFO [train.py:467] (3/4) Epoch 944, batch 9, global_batch_idx: 34900, batch size: 81, loss[discriminator_loss=2.561, discriminator_real_loss=1.259, discriminator_fake_loss=1.302, generator_loss=30.94, generator_mel_loss=20.13, generator_kl_loss=2.02, generator_dur_loss=1.633, generator_adv_loss=2.236, generator_feat_match_loss=4.914, over 81.00 samples.], tot_loss[discriminator_loss=2.591, discriminator_real_loss=1.308, discriminator_fake_loss=1.282, generator_loss=30.7, generator_mel_loss=20.14, generator_kl_loss=2.019, generator_dur_loss=1.634, generator_adv_loss=2.216, generator_feat_match_loss=4.691, over 704.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:09:15,045 INFO [train.py:811] (3/4) Start epoch 945
2023-11-15 05:11:26,089 INFO [train.py:467] (3/4) Epoch 945, batch 22, global_batch_idx: 34950, batch size: 51, loss[discriminator_loss=2.555, discriminator_real_loss=1.279, discriminator_fake_loss=1.276, generator_loss=29.94, generator_mel_loss=19.68, generator_kl_loss=2.092, generator_dur_loss=1.599, generator_adv_loss=1.998, generator_feat_match_loss=4.574, over 51.00 samples.], tot_loss[discriminator_loss=2.573, discriminator_real_loss=1.317, discriminator_fake_loss=1.256, generator_loss=30.58, generator_mel_loss=19.96, generator_kl_loss=1.986, generator_dur_loss=1.633, generator_adv_loss=2.249, generator_feat_match_loss=4.744, over 1604.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:12:43,592 INFO [train.py:811] (3/4) Start epoch 946
2023-11-15 05:16:11,576 INFO [train.py:467] (3/4) Epoch 946, batch 35, global_batch_idx: 35000, batch size: 126, loss[discriminator_loss=2.504, discriminator_real_loss=1.292, discriminator_fake_loss=1.211, generator_loss=30.58, generator_mel_loss=20, generator_kl_loss=2.002, generator_dur_loss=1.605, generator_adv_loss=2.172, generator_feat_match_loss=4.805, over 126.00 samples.], tot_loss[discriminator_loss=2.578, discriminator_real_loss=1.308, discriminator_fake_loss=1.269, generator_loss=30.46, generator_mel_loss=19.92, generator_kl_loss=1.993, generator_dur_loss=1.623, generator_adv_loss=2.219, generator_feat_match_loss=4.707, over 2798.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:16:11,579 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 05:16:22,447 INFO [train.py:517] (3/4) Epoch 946, validation: discriminator_loss=2.501, discriminator_real_loss=1.169, discriminator_fake_loss=1.332, generator_loss=31.78, generator_mel_loss=20.75, generator_kl_loss=2.266, generator_dur_loss=1.629, generator_adv_loss=2.086, generator_feat_match_loss=5.049, over 100.00 samples.
2023-11-15 05:16:22,447 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 05:16:29,196 INFO [train.py:811] (3/4) Start epoch 947
2023-11-15 05:20:00,066 INFO [train.py:811] (3/4) Start epoch 948
2023-11-15 05:21:12,340 INFO [train.py:467] (3/4) Epoch 948, batch 11, global_batch_idx: 35050, batch size: 79, loss[discriminator_loss=2.656, discriminator_real_loss=1.296, discriminator_fake_loss=1.361, generator_loss=30.07, generator_mel_loss=19.96, generator_kl_loss=2.105, generator_dur_loss=1.623, generator_adv_loss=2.064, generator_feat_match_loss=4.324, over 79.00 samples.], tot_loss[discriminator_loss=2.529, discriminator_real_loss=1.271, discriminator_fake_loss=1.259, generator_loss=30.48, generator_mel_loss=19.69, generator_kl_loss=2.023, generator_dur_loss=1.62, generator_adv_loss=2.281, generator_feat_match_loss=4.868, over 884.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:23:30,614 INFO [train.py:811] (3/4) Start epoch 949
2023-11-15 05:26:02,788 INFO [train.py:467] (3/4) Epoch 949, batch 24, global_batch_idx: 35100, batch size: 90, loss[discriminator_loss=2.537, discriminator_real_loss=1.311, discriminator_fake_loss=1.227, generator_loss=30.68, generator_mel_loss=19.79, generator_kl_loss=2.017, generator_dur_loss=1.634, generator_adv_loss=2.264, generator_feat_match_loss=4.977, over 90.00 samples.], tot_loss[discriminator_loss=2.554, discriminator_real_loss=1.295, discriminator_fake_loss=1.258, generator_loss=30.76, generator_mel_loss=19.95, generator_kl_loss=1.999, generator_dur_loss=1.624, generator_adv_loss=2.286, generator_feat_match_loss=4.894, over 1948.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:27:06,792 INFO [train.py:811] (3/4) Start epoch 950
2023-11-15 05:30:42,796 INFO [train.py:811] (3/4) Start epoch 951
2023-11-15 05:30:58,140 INFO [train.py:467] (3/4) Epoch 951, batch 0, global_batch_idx: 35150, batch size: 52, loss[discriminator_loss=2.531, discriminator_real_loss=1.295, discriminator_fake_loss=1.237, generator_loss=31.06, generator_mel_loss=20.11, generator_kl_loss=2.09, generator_dur_loss=1.641, generator_adv_loss=2.285, generator_feat_match_loss=4.926, over 52.00 samples.], tot_loss[discriminator_loss=2.531, discriminator_real_loss=1.295, discriminator_fake_loss=1.237, generator_loss=31.06, generator_mel_loss=20.11, generator_kl_loss=2.09, generator_dur_loss=1.641, generator_adv_loss=2.285, generator_feat_match_loss=4.926, over 52.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 05:34:08,002 INFO [train.py:811] (3/4) Start epoch 952
2023-11-15 05:35:39,373 INFO [train.py:467] (3/4) Epoch 952, batch 13, global_batch_idx: 35200, batch size: 52, loss[discriminator_loss=2.371, discriminator_real_loss=1.122, discriminator_fake_loss=1.248, generator_loss=31.68, generator_mel_loss=19.87, generator_kl_loss=2.108, generator_dur_loss=1.628, generator_adv_loss=2.375, generator_feat_match_loss=5.703, over 52.00 samples.], tot_loss[discriminator_loss=2.469, discriminator_real_loss=1.265, discriminator_fake_loss=1.204, generator_loss=31.13, generator_mel_loss=19.81, generator_kl_loss=2.023, generator_dur_loss=1.63, generator_adv_loss=2.418, generator_feat_match_loss=5.252, over 975.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 16.0
2023-11-15 05:35:39,375 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 05:35:50,390 INFO [train.py:517] (3/4) Epoch 952, validation: discriminator_loss=2.606, discriminator_real_loss=1.169, discriminator_fake_loss=1.437, generator_loss=30.7, generator_mel_loss=20.22, generator_kl_loss=2.176, generator_dur_loss=1.628, generator_adv_loss=1.853, generator_feat_match_loss=4.83, over 100.00 samples.
2023-11-15 05:35:50,391 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 05:37:47,734 INFO [train.py:811] (3/4) Start epoch 953
2023-11-15 05:40:21,310 INFO [train.py:467] (3/4) Epoch 953, batch 26, global_batch_idx: 35250, batch size: 59, loss[discriminator_loss=2.598, discriminator_real_loss=1.309, discriminator_fake_loss=1.289, generator_loss=30.36, generator_mel_loss=20.03, generator_kl_loss=2.02, generator_dur_loss=1.619, generator_adv_loss=2.08, generator_feat_match_loss=4.613, over 59.00 samples.], tot_loss[discriminator_loss=2.546, discriminator_real_loss=1.298, discriminator_fake_loss=1.248, generator_loss=30.68, generator_mel_loss=19.85, generator_kl_loss=2.035, generator_dur_loss=1.631, generator_adv_loss=2.302, generator_feat_match_loss=4.858, over 1960.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 05:41:16,404 INFO [train.py:811] (3/4) Start epoch 954
2023-11-15 05:44:52,862 INFO [train.py:811] (3/4) Start epoch 955
2023-11-15 05:45:18,790 INFO [train.py:467] (3/4) Epoch 955, batch 2, global_batch_idx: 35300, batch size: 49, loss[discriminator_loss=2.602, discriminator_real_loss=1.455, discriminator_fake_loss=1.146, generator_loss=30.65, generator_mel_loss=19.71, generator_kl_loss=2.071, generator_dur_loss=1.623, generator_adv_loss=2.303, generator_feat_match_loss=4.938, over 49.00 samples.], tot_loss[discriminator_loss=2.64, discriminator_real_loss=1.391, discriminator_fake_loss=1.249, generator_loss=30.36, generator_mel_loss=19.86, generator_kl_loss=2.019, generator_dur_loss=1.632, generator_adv_loss=2.268, generator_feat_match_loss=4.581, over 193.00 samples.], cur_lr_g: 1.78e-04, cur_lr_d: 1.78e-04, grad_scale: 8.0
2023-11-15 05:48:20,762 INFO [train.py:811] (3/4) Start epoch 956
2023-11-15 05:50:03,043 INFO [train.py:467] (3/4) Epoch 956, batch 15, global_batch_idx: 35350, batch size: 71, loss[discriminator_loss=2.434, discriminator_real_loss=1.268, discriminator_fake_loss=1.167, generator_loss=31.95, generator_mel_loss=19.97, generator_kl_loss=2.042, generator_dur_loss=1.637, generator_adv_loss=2.582, generator_feat_match_loss=5.719, over 71.00 samples.], tot_loss[discriminator_loss=2.495, discriminator_real_loss=1.257, discriminator_fake_loss=1.238, generator_loss=31.21, generator_mel_loss=19.95, generator_kl_loss=2.05, generator_dur_loss=1.627, generator_adv_loss=2.382, generator_feat_match_loss=5.197, over 1424.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 05:51:59,122 INFO [train.py:811] (3/4) Start epoch 957
2023-11-15 05:54:48,252 INFO [train.py:467] (3/4) Epoch 957, batch 28, global_batch_idx: 35400, batch size: 61, loss[discriminator_loss=2.559, discriminator_real_loss=1.259, discriminator_fake_loss=1.299, generator_loss=29.77, generator_mel_loss=19.34, generator_kl_loss=2.02, generator_dur_loss=1.618, generator_adv_loss=2.229, generator_feat_match_loss=4.562, over 61.00 samples.], tot_loss[discriminator_loss=2.477, discriminator_real_loss=1.254, discriminator_fake_loss=1.223, generator_loss=30.77, generator_mel_loss=19.56, generator_kl_loss=1.999, generator_dur_loss=1.627, generator_adv_loss=2.394, generator_feat_match_loss=5.185, over 2129.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 05:54:48,254 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 05:54:59,150 INFO [train.py:517] (3/4) Epoch 957, validation: discriminator_loss=2.546, discriminator_real_loss=1.16, discriminator_fake_loss=1.386, generator_loss=31.21, generator_mel_loss=20.37, generator_kl_loss=2.276, generator_dur_loss=1.623, generator_adv_loss=1.948, generator_feat_match_loss=4.993, over 100.00 samples.
2023-11-15 05:54:59,152 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 05:55:44,642 INFO [train.py:811] (3/4) Start epoch 958
2023-11-15 05:59:12,255 INFO [train.py:811] (3/4) Start epoch 959
2023-11-15 05:59:49,567 INFO [train.py:467] (3/4) Epoch 959, batch 4, global_batch_idx: 35450, batch size: 55, loss[discriminator_loss=2.506, discriminator_real_loss=1.299, discriminator_fake_loss=1.207, generator_loss=30.48, generator_mel_loss=19.77, generator_kl_loss=2.02, generator_dur_loss=1.65, generator_adv_loss=2.162, generator_feat_match_loss=4.879, over 55.00 samples.], tot_loss[discriminator_loss=2.539, discriminator_real_loss=1.295, discriminator_fake_loss=1.243, generator_loss=30.7, generator_mel_loss=19.95, generator_kl_loss=2.031, generator_dur_loss=1.636, generator_adv_loss=2.252, generator_feat_match_loss=4.823, over 301.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:02:43,763 INFO [train.py:811] (3/4) Start epoch 960
2023-11-15 06:04:35,385 INFO [train.py:467] (3/4) Epoch 960, batch 17, global_batch_idx: 35500, batch size: 79, loss[discriminator_loss=2.535, discriminator_real_loss=1.336, discriminator_fake_loss=1.199, generator_loss=30.51, generator_mel_loss=19.84, generator_kl_loss=1.942, generator_dur_loss=1.619, generator_adv_loss=2.289, generator_feat_match_loss=4.812, over 79.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.319, discriminator_fake_loss=1.266, generator_loss=30.22, generator_mel_loss=19.76, generator_kl_loss=1.987, generator_dur_loss=1.617, generator_adv_loss=2.233, generator_feat_match_loss=4.622, over 1284.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:06:19,984 INFO [train.py:811] (3/4) Start epoch 961
2023-11-15 06:09:11,917 INFO [train.py:467] (3/4) Epoch 961, batch 30, global_batch_idx: 35550, batch size: 73, loss[discriminator_loss=2.555, discriminator_real_loss=1.22, discriminator_fake_loss=1.336, generator_loss=30.09, generator_mel_loss=19.59, generator_kl_loss=2.087, generator_dur_loss=1.617, generator_adv_loss=2.367, generator_feat_match_loss=4.422, over 73.00 samples.], tot_loss[discriminator_loss=2.542, discriminator_real_loss=1.293, discriminator_fake_loss=1.249, generator_loss=30.64, generator_mel_loss=19.83, generator_kl_loss=2.036, generator_dur_loss=1.626, generator_adv_loss=2.279, generator_feat_match_loss=4.87, over 2350.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:09:50,302 INFO [train.py:811] (3/4) Start epoch 962
2023-11-15 06:13:27,428 INFO [train.py:811] (3/4) Start epoch 963
2023-11-15 06:14:18,877 INFO [train.py:467] (3/4) Epoch 963, batch 6, global_batch_idx: 35600, batch size: 110, loss[discriminator_loss=2.523, discriminator_real_loss=1.338, discriminator_fake_loss=1.187, generator_loss=31, generator_mel_loss=20.38, generator_kl_loss=2.045, generator_dur_loss=1.643, generator_adv_loss=2.107, generator_feat_match_loss=4.824, over 110.00 samples.], tot_loss[discriminator_loss=2.573, discriminator_real_loss=1.313, discriminator_fake_loss=1.26, generator_loss=30.49, generator_mel_loss=19.97, generator_kl_loss=2.002, generator_dur_loss=1.63, generator_adv_loss=2.213, generator_feat_match_loss=4.678, over 545.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 06:14:18,878 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 06:14:31,163 INFO [train.py:517] (3/4) Epoch 963, validation: discriminator_loss=2.487, discriminator_real_loss=1.174, discriminator_fake_loss=1.313, generator_loss=31.52, generator_mel_loss=20.6, generator_kl_loss=2.213, generator_dur_loss=1.629, generator_adv_loss=2.048, generator_feat_match_loss=5.035, over 100.00 samples.
2023-11-15 06:14:31,164 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 06:17:12,292 INFO [train.py:811] (3/4) Start epoch 964
2023-11-15 06:19:10,299 INFO [train.py:467] (3/4) Epoch 964, batch 19, global_batch_idx: 35650, batch size: 69, loss[discriminator_loss=2.504, discriminator_real_loss=1.217, discriminator_fake_loss=1.288, generator_loss=30.01, generator_mel_loss=19.55, generator_kl_loss=1.988, generator_dur_loss=1.625, generator_adv_loss=2.312, generator_feat_match_loss=4.527, over 69.00 samples.], tot_loss[discriminator_loss=2.546, discriminator_real_loss=1.296, discriminator_fake_loss=1.251, generator_loss=30.79, generator_mel_loss=19.84, generator_kl_loss=2.005, generator_dur_loss=1.635, generator_adv_loss=2.336, generator_feat_match_loss=4.976, over 1253.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:20:45,338 INFO [train.py:811] (3/4) Start epoch 965
2023-11-15 06:24:00,965 INFO [train.py:467] (3/4) Epoch 965, batch 32, global_batch_idx: 35700, batch size: 59, loss[discriminator_loss=2.553, discriminator_real_loss=1.216, discriminator_fake_loss=1.337, generator_loss=30.41, generator_mel_loss=20.12, generator_kl_loss=2.03, generator_dur_loss=1.653, generator_adv_loss=1.95, generator_feat_match_loss=4.66, over 59.00 samples.], tot_loss[discriminator_loss=2.54, discriminator_real_loss=1.278, discriminator_fake_loss=1.263, generator_loss=30.53, generator_mel_loss=19.85, generator_kl_loss=2.005, generator_dur_loss=1.623, generator_adv_loss=2.227, generator_feat_match_loss=4.832, over 2437.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:24:20,475 INFO [train.py:811] (3/4) Start epoch 966
2023-11-15 06:27:52,231 INFO [train.py:811] (3/4) Start epoch 967
2023-11-15 06:28:53,392 INFO [train.py:467] (3/4) Epoch 967, batch 8, global_batch_idx: 35750, batch size: 49, loss[discriminator_loss=2.602, discriminator_real_loss=1.304, discriminator_fake_loss=1.297, generator_loss=30.3, generator_mel_loss=19.7, generator_kl_loss=2.003, generator_dur_loss=1.624, generator_adv_loss=2.301, generator_feat_match_loss=4.672, over 49.00 samples.], tot_loss[discriminator_loss=2.542, discriminator_real_loss=1.281, discriminator_fake_loss=1.261, generator_loss=30.89, generator_mel_loss=20.2, generator_kl_loss=2.02, generator_dur_loss=1.634, generator_adv_loss=2.212, generator_feat_match_loss=4.82, over 667.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:31:29,526 INFO [train.py:811] (3/4) Start epoch 968
2023-11-15 06:33:36,851 INFO [train.py:467] (3/4) Epoch 968, batch 21, global_batch_idx: 35800, batch size: 81, loss[discriminator_loss=2.418, discriminator_real_loss=1.157, discriminator_fake_loss=1.261, generator_loss=31.34, generator_mel_loss=19.83, generator_kl_loss=1.986, generator_dur_loss=1.613, generator_adv_loss=2.422, generator_feat_match_loss=5.484, over 81.00 samples.], tot_loss[discriminator_loss=2.511, discriminator_real_loss=1.268, discriminator_fake_loss=1.243, generator_loss=30.92, generator_mel_loss=19.84, generator_kl_loss=2.014, generator_dur_loss=1.624, generator_adv_loss=2.377, generator_feat_match_loss=5.065, over 1751.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:33:36,853 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 06:33:47,703 INFO [train.py:517] (3/4) Epoch 968, validation: discriminator_loss=2.741, discriminator_real_loss=1.128, discriminator_fake_loss=1.613, generator_loss=30.96, generator_mel_loss=20.77, generator_kl_loss=2.193, generator_dur_loss=1.637, generator_adv_loss=1.671, generator_feat_match_loss=4.682, over 100.00 samples.
2023-11-15 06:33:47,705 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 06:35:06,123 INFO [train.py:811] (3/4) Start epoch 969
2023-11-15 06:38:27,692 INFO [train.py:467] (3/4) Epoch 969, batch 34, global_batch_idx: 35850, batch size: 90, loss[discriminator_loss=2.492, discriminator_real_loss=1.358, discriminator_fake_loss=1.133, generator_loss=31, generator_mel_loss=19.98, generator_kl_loss=2.02, generator_dur_loss=1.635, generator_adv_loss=2.342, generator_feat_match_loss=5.031, over 90.00 samples.], tot_loss[discriminator_loss=2.482, discriminator_real_loss=1.253, discriminator_fake_loss=1.229, generator_loss=30.94, generator_mel_loss=19.92, generator_kl_loss=2.039, generator_dur_loss=1.622, generator_adv_loss=2.315, generator_feat_match_loss=5.04, over 2876.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:38:40,300 INFO [train.py:811] (3/4) Start epoch 970
2023-11-15 06:42:10,454 INFO [train.py:811] (3/4) Start epoch 971
2023-11-15 06:43:22,047 INFO [train.py:467] (3/4) Epoch 971, batch 10, global_batch_idx: 35900, batch size: 67, loss[discriminator_loss=2.578, discriminator_real_loss=1.26, discriminator_fake_loss=1.319, generator_loss=30.41, generator_mel_loss=19.67, generator_kl_loss=2.041, generator_dur_loss=1.644, generator_adv_loss=2.061, generator_feat_match_loss=4.992, over 67.00 samples.], tot_loss[discriminator_loss=2.521, discriminator_real_loss=1.311, discriminator_fake_loss=1.211, generator_loss=30.96, generator_mel_loss=19.51, generator_kl_loss=1.99, generator_dur_loss=1.623, generator_adv_loss=2.493, generator_feat_match_loss=5.345, over 947.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:45:42,006 INFO [train.py:811] (3/4) Start epoch 972
2023-11-15 06:48:04,620 INFO [train.py:467] (3/4) Epoch 972, batch 23, global_batch_idx: 35950, batch size: 126, loss[discriminator_loss=2.529, discriminator_real_loss=1.331, discriminator_fake_loss=1.198, generator_loss=31.36, generator_mel_loss=20.28, generator_kl_loss=2.001, generator_dur_loss=1.623, generator_adv_loss=2.277, generator_feat_match_loss=5.172, over 126.00 samples.], tot_loss[discriminator_loss=2.557, discriminator_real_loss=1.305, discriminator_fake_loss=1.253, generator_loss=30.61, generator_mel_loss=19.95, generator_kl_loss=2.045, generator_dur_loss=1.625, generator_adv_loss=2.227, generator_feat_match_loss=4.766, over 1796.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 06:49:16,028 INFO [train.py:811] (3/4) Start epoch 973
2023-11-15 06:52:45,984 INFO [train.py:467] (3/4) Epoch 973, batch 36, global_batch_idx: 36000, batch size: 90, loss[discriminator_loss=2.518, discriminator_real_loss=1.236, discriminator_fake_loss=1.281, generator_loss=30.1, generator_mel_loss=19.5, generator_kl_loss=2.02, generator_dur_loss=1.629, generator_adv_loss=2.266, generator_feat_match_loss=4.684, over 90.00 samples.], tot_loss[discriminator_loss=2.563, discriminator_real_loss=1.288, discriminator_fake_loss=1.275, generator_loss=30.72, generator_mel_loss=19.86, generator_kl_loss=2.003, generator_dur_loss=1.63, generator_adv_loss=2.31, generator_feat_match_loss=4.92, over 2487.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 06:52:45,986 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 06:52:57,018 INFO [train.py:517] (3/4) Epoch 973, validation: discriminator_loss=2.505, discriminator_real_loss=1.19, discriminator_fake_loss=1.315, generator_loss=31.28, generator_mel_loss=20.59, generator_kl_loss=2.157, generator_dur_loss=1.638, generator_adv_loss=1.969, generator_feat_match_loss=4.919, over 100.00 samples.
2023-11-15 06:52:57,019 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 06:52:57,755 INFO [train.py:811] (3/4) Start epoch 974
2023-11-15 06:56:32,572 INFO [train.py:811] (3/4) Start epoch 975
2023-11-15 06:57:50,210 INFO [train.py:467] (3/4) Epoch 975, batch 12, global_batch_idx: 36050, batch size: 69, loss[discriminator_loss=2.512, discriminator_real_loss=1.276, discriminator_fake_loss=1.236, generator_loss=31.19, generator_mel_loss=20.16, generator_kl_loss=2.006, generator_dur_loss=1.628, generator_adv_loss=2.357, generator_feat_match_loss=5.031, over 69.00 samples.], tot_loss[discriminator_loss=2.552, discriminator_real_loss=1.304, discriminator_fake_loss=1.248, generator_loss=30.83, generator_mel_loss=20.02, generator_kl_loss=2.056, generator_dur_loss=1.628, generator_adv_loss=2.282, generator_feat_match_loss=4.845, over 1026.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:00:06,054 INFO [train.py:811] (3/4) Start epoch 976
2023-11-15 07:02:39,589 INFO [train.py:467] (3/4) Epoch 976, batch 25, global_batch_idx: 36100, batch size: 55, loss[discriminator_loss=2.77, discriminator_real_loss=1.213, discriminator_fake_loss=1.558, generator_loss=29.79, generator_mel_loss=19.62, generator_kl_loss=2.018, generator_dur_loss=1.656, generator_adv_loss=2.139, generator_feat_match_loss=4.355, over 55.00 samples.], tot_loss[discriminator_loss=2.521, discriminator_real_loss=1.271, discriminator_fake_loss=1.25, generator_loss=31.05, generator_mel_loss=19.85, generator_kl_loss=2.02, generator_dur_loss=1.626, generator_adv_loss=2.397, generator_feat_match_loss=5.156, over 2015.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:03:40,887 INFO [train.py:811] (3/4) Start epoch 977
2023-11-15 07:07:13,370 INFO [train.py:811] (3/4) Start epoch 978
2023-11-15 07:07:35,420 INFO [train.py:467] (3/4) Epoch 978, batch 1, global_batch_idx: 36150, batch size: 56, loss[discriminator_loss=2.537, discriminator_real_loss=1.264, discriminator_fake_loss=1.273, generator_loss=30.59, generator_mel_loss=19.75, generator_kl_loss=2.025, generator_dur_loss=1.623, generator_adv_loss=2.291, generator_feat_match_loss=4.902, over 56.00 samples.], tot_loss[discriminator_loss=2.531, discriminator_real_loss=1.27, discriminator_fake_loss=1.262, generator_loss=30.59, generator_mel_loss=19.78, generator_kl_loss=1.974, generator_dur_loss=1.631, generator_adv_loss=2.266, generator_feat_match_loss=4.936, over 114.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:10:43,225 INFO [train.py:811] (3/4) Start epoch 979
2023-11-15 07:12:11,736 INFO [train.py:467] (3/4) Epoch 979, batch 14, global_batch_idx: 36200, batch size: 53, loss[discriminator_loss=2.48, discriminator_real_loss=1.24, discriminator_fake_loss=1.239, generator_loss=30.61, generator_mel_loss=20.04, generator_kl_loss=1.931, generator_dur_loss=1.627, generator_adv_loss=1.961, generator_feat_match_loss=5.051, over 53.00 samples.], tot_loss[discriminator_loss=2.618, discriminator_real_loss=1.335, discriminator_fake_loss=1.283, generator_loss=30.75, generator_mel_loss=20.01, generator_kl_loss=2.024, generator_dur_loss=1.632, generator_adv_loss=2.256, generator_feat_match_loss=4.819, over 1086.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:12:11,738 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 07:12:23,049 INFO [train.py:517] (3/4) Epoch 979, validation: discriminator_loss=2.618, discriminator_real_loss=1.056, discriminator_fake_loss=1.562, generator_loss=30.72, generator_mel_loss=20.33, generator_kl_loss=2.276, generator_dur_loss=1.629, generator_adv_loss=1.682, generator_feat_match_loss=4.806, over 100.00 samples.
2023-11-15 07:12:23,050 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 07:14:23,814 INFO [train.py:811] (3/4) Start epoch 980
2023-11-15 07:17:09,469 INFO [train.py:467] (3/4) Epoch 980, batch 27, global_batch_idx: 36250, batch size: 63, loss[discriminator_loss=2.49, discriminator_real_loss=1.3, discriminator_fake_loss=1.19, generator_loss=31.12, generator_mel_loss=19.92, generator_kl_loss=2.124, generator_dur_loss=1.649, generator_adv_loss=2.434, generator_feat_match_loss=4.996, over 63.00 samples.], tot_loss[discriminator_loss=2.524, discriminator_real_loss=1.27, discriminator_fake_loss=1.253, generator_loss=30.81, generator_mel_loss=19.89, generator_kl_loss=2.009, generator_dur_loss=1.626, generator_adv_loss=2.319, generator_feat_match_loss=4.973, over 2036.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:17:56,545 INFO [train.py:811] (3/4) Start epoch 981
2023-11-15 07:21:34,018 INFO [train.py:811] (3/4) Start epoch 982
2023-11-15 07:22:04,861 INFO [train.py:467] (3/4) Epoch 982, batch 3, global_batch_idx: 36300, batch size: 64, loss[discriminator_loss=2.523, discriminator_real_loss=1.277, discriminator_fake_loss=1.246, generator_loss=30.4, generator_mel_loss=19.98, generator_kl_loss=1.971, generator_dur_loss=1.628, generator_adv_loss=2.205, generator_feat_match_loss=4.617, over 64.00 samples.], tot_loss[discriminator_loss=2.555, discriminator_real_loss=1.306, discriminator_fake_loss=1.249, generator_loss=30.58, generator_mel_loss=19.93, generator_kl_loss=2.02, generator_dur_loss=1.629, generator_adv_loss=2.27, generator_feat_match_loss=4.736, over 285.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:25:06,678 INFO [train.py:811] (3/4) Start epoch 983
2023-11-15 07:26:50,763 INFO [train.py:467] (3/4) Epoch 983, batch 16, global_batch_idx: 36350, batch size: 53, loss[discriminator_loss=2.574, discriminator_real_loss=1.167, discriminator_fake_loss=1.406, generator_loss=30.36, generator_mel_loss=19.87, generator_kl_loss=1.966, generator_dur_loss=1.621, generator_adv_loss=2.324, generator_feat_match_loss=4.578, over 53.00 samples.], tot_loss[discriminator_loss=2.567, discriminator_real_loss=1.285, discriminator_fake_loss=1.281, generator_loss=30.5, generator_mel_loss=19.73, generator_kl_loss=2.013, generator_dur_loss=1.627, generator_adv_loss=2.24, generator_feat_match_loss=4.884, over 1275.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 07:28:40,833 INFO [train.py:811] (3/4) Start epoch 984
2023-11-15 07:31:33,914 INFO [train.py:467] (3/4) Epoch 984, batch 29, global_batch_idx: 36400, batch size: 53, loss[discriminator_loss=2.734, discriminator_real_loss=1.356, discriminator_fake_loss=1.378, generator_loss=30.98, generator_mel_loss=20.62, generator_kl_loss=2.066, generator_dur_loss=1.658, generator_adv_loss=2.096, generator_feat_match_loss=4.535, over 53.00 samples.], tot_loss[discriminator_loss=2.56, discriminator_real_loss=1.294, discriminator_fake_loss=1.266, generator_loss=30.52, generator_mel_loss=19.96, generator_kl_loss=2.034, generator_dur_loss=1.623, generator_adv_loss=2.197, generator_feat_match_loss=4.705, over 2305.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:31:33,916 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 07:31:44,573 INFO [train.py:517] (3/4) Epoch 984, validation: discriminator_loss=2.678, discriminator_real_loss=1.31, discriminator_fake_loss=1.368, generator_loss=31.65, generator_mel_loss=20.98, generator_kl_loss=2.292, generator_dur_loss=1.636, generator_adv_loss=2.001, generator_feat_match_loss=4.745, over 100.00 samples.
2023-11-15 07:31:44,575 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 07:32:24,744 INFO [train.py:811] (3/4) Start epoch 985
2023-11-15 07:36:00,958 INFO [train.py:811] (3/4) Start epoch 986
2023-11-15 07:36:45,103 INFO [train.py:467] (3/4) Epoch 986, batch 5, global_batch_idx: 36450, batch size: 64, loss[discriminator_loss=2.438, discriminator_real_loss=1.175, discriminator_fake_loss=1.264, generator_loss=30.84, generator_mel_loss=19.68, generator_kl_loss=2.035, generator_dur_loss=1.599, generator_adv_loss=2.145, generator_feat_match_loss=5.379, over 64.00 samples.], tot_loss[discriminator_loss=2.592, discriminator_real_loss=1.302, discriminator_fake_loss=1.29, generator_loss=30.69, generator_mel_loss=19.82, generator_kl_loss=2.053, generator_dur_loss=1.626, generator_adv_loss=2.276, generator_feat_match_loss=4.917, over 485.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:39:28,552 INFO [train.py:811] (3/4) Start epoch 987
2023-11-15 07:41:29,278 INFO [train.py:467] (3/4) Epoch 987, batch 18, global_batch_idx: 36500, batch size: 153, loss[discriminator_loss=2.59, discriminator_real_loss=1.249, discriminator_fake_loss=1.342, generator_loss=30.88, generator_mel_loss=20.18, generator_kl_loss=2.01, generator_dur_loss=1.588, generator_adv_loss=2.279, generator_feat_match_loss=4.82, over 153.00 samples.], tot_loss[discriminator_loss=2.587, discriminator_real_loss=1.308, discriminator_fake_loss=1.28, generator_loss=30.59, generator_mel_loss=20.06, generator_kl_loss=1.997, generator_dur_loss=1.618, generator_adv_loss=2.189, generator_feat_match_loss=4.732, over 1418.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:43:01,332 INFO [train.py:811] (3/4) Start epoch 988
2023-11-15 07:46:11,555 INFO [train.py:467] (3/4) Epoch 988, batch 31, global_batch_idx: 36550, batch size: 85, loss[discriminator_loss=2.688, discriminator_real_loss=1.359, discriminator_fake_loss=1.329, generator_loss=30.28, generator_mel_loss=20.1, generator_kl_loss=2.019, generator_dur_loss=1.622, generator_adv_loss=2.141, generator_feat_match_loss=4.395, over 85.00 samples.], tot_loss[discriminator_loss=2.612, discriminator_real_loss=1.336, discriminator_fake_loss=1.275, generator_loss=30.67, generator_mel_loss=20.14, generator_kl_loss=2.04, generator_dur_loss=1.624, generator_adv_loss=2.192, generator_feat_match_loss=4.679, over 2485.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:46:36,111 INFO [train.py:811] (3/4) Start epoch 989
2023-11-15 07:50:10,964 INFO [train.py:811] (3/4) Start epoch 990
2023-11-15 07:51:03,430 INFO [train.py:467] (3/4) Epoch 990, batch 7, global_batch_idx: 36600, batch size: 54, loss[discriminator_loss=2.555, discriminator_real_loss=1.35, discriminator_fake_loss=1.206, generator_loss=29.96, generator_mel_loss=19.53, generator_kl_loss=1.933, generator_dur_loss=1.642, generator_adv_loss=2.23, generator_feat_match_loss=4.633, over 54.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.311, discriminator_fake_loss=1.274, generator_loss=30.51, generator_mel_loss=19.92, generator_kl_loss=2.015, generator_dur_loss=1.623, generator_adv_loss=2.197, generator_feat_match_loss=4.746, over 623.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:51:03,432 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 07:51:15,413 INFO [train.py:517] (3/4) Epoch 990, validation: discriminator_loss=2.502, discriminator_real_loss=1.136, discriminator_fake_loss=1.367, generator_loss=31.61, generator_mel_loss=20.82, generator_kl_loss=2.262, generator_dur_loss=1.631, generator_adv_loss=1.992, generator_feat_match_loss=4.902, over 100.00 samples.
2023-11-15 07:51:15,414 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 07:53:57,953 INFO [train.py:811] (3/4) Start epoch 991
2023-11-15 07:56:07,124 INFO [train.py:467] (3/4) Epoch 991, batch 20, global_batch_idx: 36650, batch size: 154, loss[discriminator_loss=2.574, discriminator_real_loss=1.33, discriminator_fake_loss=1.244, generator_loss=30.95, generator_mel_loss=20.27, generator_kl_loss=1.892, generator_dur_loss=1.623, generator_adv_loss=2.234, generator_feat_match_loss=4.93, over 154.00 samples.], tot_loss[discriminator_loss=2.609, discriminator_real_loss=1.32, discriminator_fake_loss=1.289, generator_loss=30.79, generator_mel_loss=20.21, generator_kl_loss=2.004, generator_dur_loss=1.629, generator_adv_loss=2.202, generator_feat_match_loss=4.748, over 1573.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 07:57:34,006 INFO [train.py:811] (3/4) Start epoch 992
2023-11-15 08:00:41,890 INFO [train.py:467] (3/4) Epoch 992, batch 33, global_batch_idx: 36700, batch size: 63, loss[discriminator_loss=2.588, discriminator_real_loss=1.381, discriminator_fake_loss=1.207, generator_loss=30.13, generator_mel_loss=19.82, generator_kl_loss=2.052, generator_dur_loss=1.626, generator_adv_loss=2.217, generator_feat_match_loss=4.418, over 63.00 samples.], tot_loss[discriminator_loss=2.585, discriminator_real_loss=1.312, discriminator_fake_loss=1.273, generator_loss=30.64, generator_mel_loss=20.06, generator_kl_loss=2.033, generator_dur_loss=1.628, generator_adv_loss=2.219, generator_feat_match_loss=4.697, over 2501.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 08:01:03,129 INFO [train.py:811] (3/4) Start epoch 993
2023-11-15 08:04:36,987 INFO [train.py:811] (3/4) Start epoch 994
2023-11-15 08:05:43,135 INFO [train.py:467] (3/4) Epoch 994, batch 9, global_batch_idx: 36750, batch size: 110, loss[discriminator_loss=2.504, discriminator_real_loss=1.322, discriminator_fake_loss=1.181, generator_loss=31.08, generator_mel_loss=19.83, generator_kl_loss=2.063, generator_dur_loss=1.606, generator_adv_loss=2.426, generator_feat_match_loss=5.156, over 110.00 samples.], tot_loss[discriminator_loss=2.664, discriminator_real_loss=1.361, discriminator_fake_loss=1.303, generator_loss=30.82, generator_mel_loss=19.71, generator_kl_loss=1.994, generator_dur_loss=1.624, generator_adv_loss=2.383, generator_feat_match_loss=5.107, over 879.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 08:08:07,109 INFO [train.py:811] (3/4) Start epoch 995
2023-11-15 08:10:27,348 INFO [train.py:467] (3/4) Epoch 995, batch 22, global_batch_idx: 36800, batch size: 71, loss[discriminator_loss=2.59, discriminator_real_loss=1.4, discriminator_fake_loss=1.19, generator_loss=30.62, generator_mel_loss=20.17, generator_kl_loss=2.029, generator_dur_loss=1.605, generator_adv_loss=2.059, generator_feat_match_loss=4.754, over 71.00 samples.], tot_loss[discriminator_loss=2.579, discriminator_real_loss=1.308, discriminator_fake_loss=1.271, generator_loss=30.58, generator_mel_loss=20, generator_kl_loss=2.03, generator_dur_loss=1.62, generator_adv_loss=2.171, generator_feat_match_loss=4.76, over 1849.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 08:10:27,350 INFO [train.py:508] (3/4) Computing validation loss
2023-11-15 08:10:38,606 INFO [train.py:517] (3/4) Epoch 995, validation: discriminator_loss=2.653, discriminator_real_loss=1.189, discriminator_fake_loss=1.464, generator_loss=30.64, generator_mel_loss=20.53, generator_kl_loss=2.269, generator_dur_loss=1.626, generator_adv_loss=1.772, generator_feat_match_loss=4.444, over 100.00 samples.
2023-11-15 08:10:38,607 INFO [train.py:518] (3/4) Maximum memory allocated so far is 27302MB
2023-11-15 08:11:49,631 INFO [train.py:811] (3/4) Start epoch 996
2023-11-15 08:15:19,984 INFO [train.py:467] (3/4) Epoch 996, batch 35, global_batch_idx: 36850, batch size: 73, loss[discriminator_loss=2.547, discriminator_real_loss=1.26, discriminator_fake_loss=1.286, generator_loss=30.55, generator_mel_loss=19.85, generator_kl_loss=2.026, generator_dur_loss=1.642, generator_adv_loss=2.266, generator_feat_match_loss=4.766, over 73.00 samples.], tot_loss[discriminator_loss=2.586, discriminator_real_loss=1.309, discriminator_fake_loss=1.277, generator_loss=30.59, generator_mel_loss=19.93, generator_kl_loss=2.022, generator_dur_loss=1.627, generator_adv_loss=2.235, generator_feat_match_loss=4.775, over 2726.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 16.0
2023-11-15 08:15:25,711 INFO [train.py:811] (3/4) Start epoch 997
2023-11-15 08:18:59,889 INFO [train.py:811] (3/4) Start epoch 998
2023-11-15 08:20:22,857 INFO [train.py:467] (3/4) Epoch 998, batch 11, global_batch_idx: 36900, batch size: 65, loss[discriminator_loss=2.654, discriminator_real_loss=1.318, discriminator_fake_loss=1.336, generator_loss=30.54, generator_mel_loss=19.99, generator_kl_loss=2.104, generator_dur_loss=1.653, generator_adv_loss=2.031, generator_feat_match_loss=4.762, over 65.00 samples.], tot_loss[discriminator_loss=2.574, discriminator_real_loss=1.309, discriminator_fake_loss=1.265, generator_loss=30.92, generator_mel_loss=20.07, generator_kl_loss=2.048, generator_dur_loss=1.636, generator_adv_loss=2.276, generator_feat_match_loss=4.887, over 933.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 08:22:38,780 INFO [train.py:811] (3/4) Start epoch 999
2023-11-15 08:25:03,030 INFO [train.py:467] (3/4) Epoch 999, batch 24, global_batch_idx: 36950, batch size: 69, loss[discriminator_loss=2.635, discriminator_real_loss=1.391, discriminator_fake_loss=1.244, generator_loss=30.11, generator_mel_loss=19.99, generator_kl_loss=2.016, generator_dur_loss=1.613, generator_adv_loss=2.033, generator_feat_match_loss=4.457, over 69.00 samples.], tot_loss[discriminator_loss=2.563, discriminator_real_loss=1.294, discriminator_fake_loss=1.269, generator_loss=30.54, generator_mel_loss=19.88, generator_kl_loss=2.01, generator_dur_loss=1.627, generator_adv_loss=2.229, generator_feat_match_loss=4.798, over 1752.00 samples.], cur_lr_g: 1.77e-04, cur_lr_d: 1.77e-04, grad_scale: 8.0
2023-11-15 08:26:10,777 INFO [train.py:811] (3/4) Start epoch 1000
2023-11-15 08:29:43,421 INFO [train.py:868] (3/4) Done!