======================== START TIME: Tue Jul 2 16:17:21 UTC 2024 python3 version = Python 3.10.14 ======================== The token has not been saved to the git credentials helper. Pass `add_to_git_credential=True` in this function directly or `--add-to-git-credential` if using via `huggingface-cli` if you want to set the git credential as well. Token is valid (permission: write). Your token has been saved to /admin/home/ferdinand_mom/.cache/huggingface/token Login successful Already on 'bench_cluster' M examples/config_tiny_llama.py M examples/config_tiny_llama.yaml M examples/train_tiny_llama.sh M src/nanotron/models/llama.py M src/nanotron/trainer.py Your branch is up to date with 'origin/bench_cluster'. Job status: RUNNING W0702 16:17:29.492000 140694550624064 torch/distributed/run.py:757] W0702 16:17:29.492000 140694550624064 torch/distributed/run.py:757] ***************************************** W0702 16:17:29.492000 140694550624064 torch/distributed/run.py:757] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. W0702 16:17:29.492000 140694550624064 torch/distributed/run.py:757] ***************************************** W0702 16:17:29.538000 140033536743232 torch/distributed/run.py:757] W0702 16:17:29.538000 140033536743232 torch/distributed/run.py:757] ***************************************** W0702 16:17:29.538000 140033536743232 torch/distributed/run.py:757] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. W0702 16:17:29.538000 140033536743232 torch/distributed/run.py:757] ***************************************** [default0]:07/02/2024 16:17:53 [WARNING|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Vocab Size Padding] Padded vocab (size: 50257) with 1 dummy tokens (new size: 50258) [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Config: [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Config(general=GeneralArgs(project='bench_cluster', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: run='%date_%jobid', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: seed=42, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: step=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: consumed_train_samples=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: benchmark_csv_path=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: ignore_sanity_checks=True), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: parallelism=ParallelismArgs(dp=4, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pp=2, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tp=2, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pp_engine=, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tp_mode=, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tp_linear_async_communication=False, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: expert_parallel_size=1), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: model=ModelArgs(model_config=LlamaConfig(bos_token_id=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: eos_token_id=2, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hidden_act='silu', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hidden_size=2048, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: initializer_range=0.02, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: intermediate_size=4096, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: is_llama_config=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: max_position_embeddings=4096, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_attention_heads=32, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_hidden_layers=24, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_key_value_heads=32, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pad_token_id=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pretraining_tp=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rms_norm_eps=1e-05, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rope_scaling=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rope_theta=10000.0, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tie_word_embeddings=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: use_cache=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: vocab_size=50258), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: init_method=RandomInit(std=0.025), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: dtype=torch.bfloat16, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: make_vocab_size_divisible_by=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: ddp_bucket_cap_mb=25), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tokenizer=TokenizerArgs(tokenizer_name_or_path='openai-community/gpt2', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tokenizer_revision=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tokenizer_max_length=None), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: checkpoints=CheckpointsArgs(checkpoints_path=Path('/dev/null'), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: checkpoint_interval=100000, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: save_initial_state=False, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: resume_checkpoint_path=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: checkpoints_path_is_shared_file_system=False), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: logging=LoggingArgs(log_level='info', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: log_level_replica='info', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: iteration_step_info_interval=1), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tokens=TokensArgs(sequence_length=4096, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: train_steps=20, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: micro_batch_size=2, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: batch_accumulation_per_replica=128, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: val_check_interval=-1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: limit_val_batches=0, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: limit_test_batches=0), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: optimizer=OptimizerArgs(optimizer_factory=AdamWOptimizerArgs(adam_eps=1e-08, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: adam_beta1=0.9, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: adam_beta2=0.95, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: torch_adam_is_fused=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: name='adamW'), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: zero_stage=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: weight_decay=0.01, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: clip_grad=1.0, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: accumulate_grad_in_fp32=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: learning_rate_scheduler=LRSchedulerArgs(learning_rate=0.0001, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lr_warmup_steps=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lr_warmup_style='linear', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lr_decay_style='linear', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lr_decay_steps=19, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lr_decay_starting_step=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: min_decay_lr=1e-05)), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: data_stages=[DatasetStageArgs(name='Training Stage', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: start_training_step=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: data=DataArgs(dataset=PretrainDatasetsArgs(hf_dataset_or_datasets='roneneldan/TinyStories', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hf_dataset_splits='train', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hf_dataset_config_name=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: dataset_processing_num_proc_per_process=64, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: dataset_overwrite_cache=False, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: text_column_name='text'), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: seed=42, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_loading_workers=32))], [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: profiler=ProfilerArgs(profiler_export_path=Path('/fsx/ferdinandmom/ferdinand-hf/bench_cluster/results/llama-1B/16_GPUS/dp-4_tp-2_pp-2_mbz-2')), [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: lighteval=None) [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Model Config: [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: LlamaConfig(bos_token_id=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: eos_token_id=2, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hidden_act='silu', [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: hidden_size=2048, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: initializer_range=0.02, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: intermediate_size=4096, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: is_llama_config=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: max_position_embeddings=4096, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_attention_heads=32, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_hidden_layers=24, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: num_key_value_heads=32, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pad_token_id=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: pretraining_tp=1, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rms_norm_eps=1e-05, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rope_scaling=None, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: rope_theta=10000.0, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: tie_word_embeddings=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: use_cache=True, [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: vocab_size=50258) [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Building model.. [default0]:07/02/2024 16:17:53 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Setting PP block ranks... [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=0|ip-26-0-163-147]: Local number of parameters: 261M (498.24MiB) [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=1|ip-26-0-163-147]: Local number of parameters: 261M (498.24MiB) [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=1|ip-26-0-163-147]: [After model building] Memory usage: 508.26MiB. Peak allocated: 510.29MiB Peak reserved: 526.00MiB [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=1|ip-26-0-163-147]: No checkpoint path provided. [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=0|ip-26-0-163-147]: [After model building] Memory usage: 508.26MiB. Peak allocated: 510.29MiB Peak reserved: 526.00MiB [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=1|TP=0|ip-26-0-163-147]: No checkpoint path provided. [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Total number of parameters: 1.21G (2313.02MiB) [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Local number of parameters: 345M (658.27MiB) [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [After model building] Memory usage: 672.29MiB. Peak allocated: 674.32MiB Peak reserved: 690.00MiB [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: No checkpoint path provided. [default0]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Parametrizing model parameters using StandardParametrizator [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=1|ip-26-0-163-134]: Local number of parameters: 345M (658.27MiB) [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=1|ip-26-0-163-134]: [After model building] Memory usage: 672.29MiB. Peak allocated: 674.32MiB Peak reserved: 690.00MiB [default1]:07/02/2024 16:18:05 [INFO|DP=0|PP=0|TP=1|ip-26-0-163-134]: No checkpoint path provided. [default3]:07/02/2024 16:18:05 [INFO|DP=1|PP=0|TP=1|ip-26-0-163-134]: No checkpoint path provided. [default3]:07/02/2024 16:18:05 [INFO|DP=1|PP=1|TP=1|ip-26-0-163-147]: No checkpoint path provided. [default2]:07/02/2024 16:18:05 [INFO|DP=1|PP=1|TP=0|ip-26-0-163-147]: No checkpoint path provided. [default2]:07/02/2024 16:18:05 [INFO|DP=1|PP=0|TP=0|ip-26-0-163-134]: No checkpoint path provided. [default5]:07/02/2024 16:18:05 [INFO|DP=2|PP=0|TP=1|ip-26-0-163-134]: No checkpoint path provided. [default4]:07/02/2024 16:18:05 [INFO|DP=2|PP=0|TP=0|ip-26-0-163-134]: No checkpoint path provided. [default5]:07/02/2024 16:18:05 [INFO|DP=2|PP=1|TP=1|ip-26-0-163-147]: No checkpoint path provided. [default4]:07/02/2024 16:18:05 [INFO|DP=2|PP=1|TP=0|ip-26-0-163-147]: No checkpoint path provided. [default7]:07/02/2024 16:18:05 [INFO|DP=3|PP=0|TP=1|ip-26-0-163-134]: No checkpoint path provided. [default6]:07/02/2024 16:18:05 [INFO|DP=3|PP=1|TP=0|ip-26-0-163-147]: No checkpoint path provided. [default7]:07/02/2024 16:18:05 [INFO|DP=3|PP=1|TP=1|ip-26-0-163-147]: No checkpoint path provided. [default6]:07/02/2024 16:18:05 [INFO|DP=3|PP=0|TP=0|ip-26-0-163-134]: No checkpoint path provided. [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Optimizer Building] Using LearningRateForSP as learning rate [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [ZeRO sharding] Size of optimizer params per rank: [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [ZeRO sharding] DP Rank 0 has 86.3M out of 345M (25.00%) params' optimizer states [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [ZeRO sharding] DP Rank 1 has 86.3M out of 345M (25.00%) params' optimizer states [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [ZeRO sharding] DP Rank 2 has 86.3M out of 345M (25.00%) params' optimizer states [default0]:07/02/2024 16:18:10 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [ZeRO sharding] DP Rank 3 has 86.3M out of 345M (25.00%) params' optimizer states [default0]:07/02/2024 16:18:11 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Training Plan] Stage Training Stage has 19 remaining training steps and has consumed 0 samples [default0]:07/02/2024 16:18:11 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Using `datasets` library [default0]:07/02/2024 16:18:11 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Loading tokenizer from openai-community/gpt2 and transformers/hf_hub versions ('4.41.2', '0.23.4') [default0]:07/02/2024 16:18:11 [WARNING|DP=0|PP=0|TP=0|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default0]:Repo card metadata block was not found. Setting CardData to empty. [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Training Plan] There are 1 training stages [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Stage Training Stage] start from step 1 [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: [Start training] datetime: 2024-07-02 16:18:13.916300 | mbs: 2 | grad_accum: 128 | global_batch_size: 1024 | sequence_length: 4096 | train_steps: 20 | start_iteration_step: 0 | consumed_train_samples: 0 [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Resuming training from stage Training Stage, it has trained for 0 samples and has 19 remaining train steps [default0]:07/02/2024 16:18:13 [INFO|DP=0|PP=0|TP=0|ip-26-0-163-134]: Memory usage: 2318.82MiB. Peak allocated 2318.82MiB. Peak reserved: 2338.00MiB [default7]:07/02/2024 16:18:14 [WARNING|DP=3|PP=0|TP=1|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default2]:Repo card metadata block was not found. Setting CardData to empty. [default7]:Repo card metadata block was not found. Setting CardData to empty. [default7]:07/02/2024 16:18:14 [WARNING|DP=3|PP=1|TP=1|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default2]:07/02/2024 16:18:14 [WARNING|DP=1|PP=1|TP=0|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default4]:Repo card metadata block was not found. Setting CardData to empty. [default4]:07/02/2024 16:18:14 [WARNING|DP=2|PP=1|TP=0|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default0]:Repo card metadata block was not found. Setting CardData to empty. [default6]:Repo card metadata block was not found. Setting CardData to empty. [default0]:07/02/2024 16:18:14 [WARNING|DP=0|PP=1|TP=0|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default4]:Repo card metadata block was not found. Setting CardData to empty. [default4]:07/02/2024 16:18:14 [WARNING|DP=2|PP=0|TP=0|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default5]:07/02/2024 16:18:14 [WARNING|DP=2|PP=0|TP=1|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default3]:07/02/2024 16:18:14 [WARNING|DP=1|PP=0|TP=1|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default6]:Repo card metadata block was not found. Setting CardData to empty. [default7]:Repo card metadata block was not found. Setting CardData to empty. [default1]:07/02/2024 16:18:14 [WARNING|DP=0|PP=1|TP=1|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default3]:07/02/2024 16:18:14 [WARNING|DP=1|PP=1|TP=1|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default6]:07/02/2024 16:18:14 [WARNING|DP=3|PP=1|TP=0|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default1]:Repo card metadata block was not found. Setting CardData to empty. [default1]:07/02/2024 16:18:14 [WARNING|DP=0|PP=0|TP=1|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default2]:07/02/2024 16:18:14 [WARNING|DP=1|PP=0|TP=0|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default3]:Repo card metadata block was not found. Setting CardData to empty. [default1]:Repo card metadata block was not found. Setting CardData to empty. [default5]:07/02/2024 16:18:14 [WARNING|DP=2|PP=1|TP=1|ip-26-0-163-147]: Repo card metadata block was not found. Setting CardData to empty. [default5]:Repo card metadata block was not found. Setting CardData to empty. [default5]:Repo card metadata block was not found. Setting CardData to empty. [default3]:Repo card metadata block was not found. Setting CardData to empty. [default6]:07/02/2024 16:18:14 [WARNING|DP=3|PP=0|TP=0|ip-26-0-163-134]: Repo card metadata block was not found. Setting CardData to empty. [default2]:Repo card metadata block was not found. Setting CardData to empty. [default5]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default5]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default4]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default4]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default0]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default0]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default1]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default1]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default7]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default2]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default7]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default2]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default6]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default6]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default3]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default3]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default4]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default4]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default5]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default5]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default7]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default7]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default7]:[rank7]: OSError: [Errno 122] Disk quota exceeded [default7]: [default7]:[rank7]: During handling of the above exception, another exception occurred: [default7]: [default7]:[rank7]: Traceback (most recent call last): [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default7]:[rank7]: trainer.train(dataloader) [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default7]:[rank7]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default7]:[rank7]: outputs = self.pipeline_engine.train_batch_iter( [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 295, in train_batch_iter [default7]:[rank7]: self.backward(context=context, state=state, grad_accumulator=grad_accumulator) [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 86, in backward [default7]:[rank7]: grad_accumulator.backward(sum(activations)) [default7]:[rank7]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/optim/gradient_accumulator.py", line 205, in backward [default7]:[rank7]: result = loss.backward() [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/_tensor.py", line 525, in backward [default7]:[rank7]: torch.autograd.backward( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/__init__.py", line 267, in backward [default7]:[rank7]: _engine_run_backward( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py", line 744, in _engine_run_backward [default7]:[rank7]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/function.py", line 301, in apply [default7]:[rank7]: return user_fn(self, *args) [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 821, in backward [default7]:[rank7]: dx, dw, db, dresidual_in, dx1, dw1, db1 = _layer_norm_bwd( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 643, in _layer_norm_bwd [default7]:[rank7]: _layer_norm_bwd_kernel[grid]( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 167, in [default7]:[rank7]: return lambda *args, **kwargs: self.run(grid=grid, warmup=False, *args, **kwargs) [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in run [default7]:[rank7]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in [default7]:[rank7]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 122, in _bench [default7]:[rank7]: return do_bench(kernel_call, warmup=self.warmup, rep=self.rep, quantiles=(0.5, 0.2, 0.8)) [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/testing.py", line 102, in do_bench [default7]:[rank7]: fn() [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 110, in kernel_call [default7]:[rank7]: self.fn.run( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default7]:[rank7]: return self.fn.run(*args, **kwargs) [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default7]:[rank7]: return self.fn.run(*args, **kwargs) [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default7]:[rank7]: return self.fn.run(*args, **kwargs) [default7]:[rank7]: [Previous line repeated 2 more times] [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 416, in run [default7]:[rank7]: self.cache[device][key] = compile( [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/compiler/compiler.py", line 194, in compile [default7]:[rank7]: metadata_group[f"{src.name}.{ext}"] = fn_cache_manager.put(next_module, f"{src.name}.{ext}") [default7]:[rank7]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/cache.py", line 123, in put [default7]:[rank7]: with open(temp_path, mode) as f: [default7]:[rank7]: OSError: [Errno 122] Disk quota exceeded [default6]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default6]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default6]:[rank6]: OSError: [Errno 122] Disk quota exceeded [default6]: [default6]:[rank6]: During handling of the above exception, another exception occurred: [default6]: [default6]:[rank6]: Traceback (most recent call last): [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default6]:[rank6]: trainer.train(dataloader) [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default6]:[rank6]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default6]:[rank6]: outputs = self.pipeline_engine.train_batch_iter( [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 295, in train_batch_iter [default6]:[rank6]: self.backward(context=context, state=state, grad_accumulator=grad_accumulator) [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 86, in backward [default6]:[rank6]: grad_accumulator.backward(sum(activations)) [default6]:[rank6]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/optim/gradient_accumulator.py", line 205, in backward [default6]:[rank6]: result = loss.backward() [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/_tensor.py", line 525, in backward [default6]:[rank6]: torch.autograd.backward( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/__init__.py", line 267, in backward [default6]:[rank6]: _engine_run_backward( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py", line 744, in _engine_run_backward [default6]:[rank6]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/function.py", line 301, in apply [default6]:[rank6]: return user_fn(self, *args) [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 821, in backward [default6]:[rank6]: dx, dw, db, dresidual_in, dx1, dw1, db1 = _layer_norm_bwd( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 643, in _layer_norm_bwd [default6]:[rank6]: _layer_norm_bwd_kernel[grid]( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 167, in [default6]:[rank6]: return lambda *args, **kwargs: self.run(grid=grid, warmup=False, *args, **kwargs) [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in run [default6]:[rank6]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in [default6]:[rank6]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 122, in _bench [default6]:[rank6]: return do_bench(kernel_call, warmup=self.warmup, rep=self.rep, quantiles=(0.5, 0.2, 0.8)) [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/testing.py", line 102, in do_bench [default6]:[rank6]: fn() [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 110, in kernel_call [default6]:[rank6]: self.fn.run( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default6]:[rank6]: return self.fn.run(*args, **kwargs) [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default6]:[rank6]: return self.fn.run(*args, **kwargs) [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default6]:[rank6]: return self.fn.run(*args, **kwargs) [default6]:[rank6]: [Previous line repeated 2 more times] [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 416, in run [default6]:[rank6]: self.cache[device][key] = compile( [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/compiler/compiler.py", line 194, in compile [default6]:[rank6]: metadata_group[f"{src.name}.{ext}"] = fn_cache_manager.put(next_module, f"{src.name}.{ext}") [default6]:[rank6]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/cache.py", line 123, in put [default6]:[rank6]: with open(temp_path, mode) as f: [default6]:[rank6]: OSError: [Errno 122] Disk quota exceeded [default1]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default1]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default1]:[rank1]: OSError: [Errno 122] Disk quota exceeded [default1]: [default1]:[rank1]: During handling of the above exception, another exception occurred: [default1]: [default1]:[rank1]: Traceback (most recent call last): [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default1]:[rank1]: trainer.train(dataloader) [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default1]:[rank1]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default1]:[rank1]: outputs = self.pipeline_engine.train_batch_iter( [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 295, in train_batch_iter [default1]:[rank1]: self.backward(context=context, state=state, grad_accumulator=grad_accumulator) [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 86, in backward [default1]:[rank1]: grad_accumulator.backward(sum(activations)) [default1]:[rank1]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/optim/gradient_accumulator.py", line 205, in backward [default1]:[rank1]: result = loss.backward() [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/_tensor.py", line 525, in backward [default1]:[rank1]: torch.autograd.backward( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/__init__.py", line 267, in backward [default1]:[rank1]: _engine_run_backward( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py", line 744, in _engine_run_backward [default1]:[rank1]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/function.py", line 301, in apply [default1]:[rank1]: return user_fn(self, *args) [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 821, in backward [default1]:[rank1]: dx, dw, db, dresidual_in, dx1, dw1, db1 = _layer_norm_bwd( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 643, in _layer_norm_bwd [default1]:[rank1]: _layer_norm_bwd_kernel[grid]( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 167, in [default1]:[rank1]: return lambda *args, **kwargs: self.run(grid=grid, warmup=False, *args, **kwargs) [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in run [default1]:[rank1]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in [default1]:[rank1]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 122, in _bench [default1]:[rank1]: return do_bench(kernel_call, warmup=self.warmup, rep=self.rep, quantiles=(0.5, 0.2, 0.8)) [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/testing.py", line 102, in do_bench [default1]:[rank1]: fn() [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 110, in kernel_call [default1]:[rank1]: self.fn.run( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default1]:[rank1]: return self.fn.run(*args, **kwargs) [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default1]:[rank1]: return self.fn.run(*args, **kwargs) [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default1]:[rank1]: return self.fn.run(*args, **kwargs) [default1]:[rank1]: [Previous line repeated 2 more times] [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 416, in run [default1]:[rank1]: self.cache[device][key] = compile( [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/compiler/compiler.py", line 194, in compile [default1]:[rank1]: metadata_group[f"{src.name}.{ext}"] = fn_cache_manager.put(next_module, f"{src.name}.{ext}") [default1]:[rank1]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/cache.py", line 123, in put [default1]:[rank1]: with open(temp_path, mode) as f: [default1]:[rank1]: OSError: [Errno 122] Disk quota exceeded [default0]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default0]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default0]:[rank0]: OSError: [Errno 122] Disk quota exceeded [default0]: [default0]:[rank0]: During handling of the above exception, another exception occurred: [default0]: [default0]:[rank0]: Traceback (most recent call last): [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default0]:[rank0]: trainer.train(dataloader) [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default0]:[rank0]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default0]:[rank0]: outputs = self.pipeline_engine.train_batch_iter( [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 295, in train_batch_iter [default0]:[rank0]: self.backward(context=context, state=state, grad_accumulator=grad_accumulator) [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 86, in backward [default0]:[rank0]: grad_accumulator.backward(sum(activations)) [default0]:[rank0]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/optim/gradient_accumulator.py", line 205, in backward [default0]:[rank0]: result = loss.backward() [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/_tensor.py", line 525, in backward [default0]:[rank0]: torch.autograd.backward( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/__init__.py", line 267, in backward [default0]:[rank0]: _engine_run_backward( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py", line 744, in _engine_run_backward [default0]:[rank0]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/function.py", line 301, in apply [default0]:[rank0]: return user_fn(self, *args) [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 821, in backward [default0]:[rank0]: dx, dw, db, dresidual_in, dx1, dw1, db1 = _layer_norm_bwd( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/flash_attn/ops/triton/layer_norm.py", line 643, in _layer_norm_bwd [default0]:[rank0]: _layer_norm_bwd_kernel[grid]( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 167, in [default0]:[rank0]: return lambda *args, **kwargs: self.run(grid=grid, warmup=False, *args, **kwargs) [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in run [default0]:[rank0]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 143, in [default0]:[rank0]: timings = {config: self._bench(*args, config=config, **kwargs) for config in pruned_configs} [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 122, in _bench [default0]:[rank0]: return do_bench(kernel_call, warmup=self.warmup, rep=self.rep, quantiles=(0.5, 0.2, 0.8)) [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/testing.py", line 102, in do_bench [default0]:[rank0]: fn() [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 110, in kernel_call [default0]:[rank0]: self.fn.run( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default0]:[rank0]: return self.fn.run(*args, **kwargs) [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default0]:[rank0]: return self.fn.run(*args, **kwargs) [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/autotuner.py", line 305, in run [default0]:[rank0]: return self.fn.run(*args, **kwargs) [default0]:[rank0]: [Previous line repeated 2 more times] [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/jit.py", line 416, in run [default0]:[rank0]: self.cache[device][key] = compile( [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/compiler/compiler.py", line 194, in compile [default0]:[rank0]: metadata_group[f"{src.name}.{ext}"] = fn_cache_manager.put(next_module, f"{src.name}.{ext}") [default0]:[rank0]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/triton/runtime/cache.py", line 123, in put [default0]:[rank0]: with open(temp_path, mode) as f: [default0]:[rank0]: OSError: [Errno 122] Disk quota exceeded [default3]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default3]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default2]:/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::allreduce_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.) [default2]: return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass [default1]:[rank9]: Traceback (most recent call last): [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default1]:[rank9]: trainer.train(dataloader) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default1]:[rank9]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default1]:[rank9]: outputs = self.pipeline_engine.train_batch_iter( [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 278, in train_batch_iter [default1]:[rank9]: output = self.forward(context=context, state=state, micro_batch=micro_batch, model=model) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 44, in forward [default1]:[rank9]: output = model(**micro_batch) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default1]:[rank9]: return self._call_impl(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default1]:[rank9]: return forward_call(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 891, in forward [default1]:[rank9]: sharded_logits = self.model( [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default1]:[rank9]: return self._call_impl(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default1]:[rank9]: return forward_call(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 764, in forward [default1]:[rank9]: return self.forward_with_hidden_states(input_ids=input_ids, input_mask=input_mask)[0] [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 780, in forward_with_hidden_states [default1]:[rank9]: hidden_encoder_states = encoder_block(**hidden_encoder_states) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default1]:[rank9]: return self._call_impl(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default1]:[rank9]: return forward_call(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/block.py", line 126, in forward [default1]:[rank9]: new_kwargs[name] = recv_from_pipeline_state_buffer( [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/functional.py", line 117, in recv_from_pipeline_state_buffer [default1]:[rank9]: pipeline_state.run_communication() [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 150, in run_communication [default1]:[rank9]: recv_activation_tensor = recv_activation() [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 31, in __call__ [default1]:[rank9]: return self.p2p.recv_tensors(num_tensors=1, from_rank=self.from_rank)[0] [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 353, in recv_tensors [default1]:[rank9]: buffers, futures = self.irecv_tensors(num_tensors=num_tensors, from_rank=from_rank, tag=tag) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 326, in irecv_tensors [default1]:[rank9]: meta = self._recv_meta(from_rank=from_rank, tag=tag) [default1]:[rank9]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 269, in _recv_meta [default1]:[rank9]: dist.recv( [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/c10d_logger.py", line 75, in wrapper [default1]:[rank9]: return func(*args, **kwargs) [default1]:[rank9]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/distributed_c10d.py", line 1932, in recv [default1]:[rank9]: pg.recv([tensor], group_src_rank, tag).wait() [default1]:[rank9]: torch.distributed.DistBackendError: NCCL communicator was aborted on rank 1. [default6]:[rank14]: Traceback (most recent call last): [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default6]:[rank14]: trainer.train(dataloader) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default6]:[rank14]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default6]:[rank14]: outputs = self.pipeline_engine.train_batch_iter( [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 278, in train_batch_iter [default6]:[rank14]: output = self.forward(context=context, state=state, micro_batch=micro_batch, model=model) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 44, in forward [default6]:[rank14]: output = model(**micro_batch) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default6]:[rank14]: return self._call_impl(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default6]:[rank14]: return forward_call(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 891, in forward [default6]:[rank14]: sharded_logits = self.model( [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default6]:[rank14]: return self._call_impl(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default6]:[rank14]: return forward_call(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 764, in forward [default6]:[rank14]: return self.forward_with_hidden_states(input_ids=input_ids, input_mask=input_mask)[0] [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 780, in forward_with_hidden_states [default6]:[rank14]: hidden_encoder_states = encoder_block(**hidden_encoder_states) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default6]:[rank14]: return self._call_impl(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default6]:[rank14]: return forward_call(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/block.py", line 126, in forward [default6]:[rank14]: new_kwargs[name] = recv_from_pipeline_state_buffer( [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/functional.py", line 117, in recv_from_pipeline_state_buffer [default6]:[rank14]: pipeline_state.run_communication() [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 150, in run_communication [default6]:[rank14]: recv_activation_tensor = recv_activation() [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 31, in __call__ [default6]:[rank14]: return self.p2p.recv_tensors(num_tensors=1, from_rank=self.from_rank)[0] [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 353, in recv_tensors [default6]:[rank14]: buffers, futures = self.irecv_tensors(num_tensors=num_tensors, from_rank=from_rank, tag=tag) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 326, in irecv_tensors [default6]:[rank14]: meta = self._recv_meta(from_rank=from_rank, tag=tag) [default6]:[rank14]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 269, in _recv_meta [default6]:[rank14]: dist.recv( [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/c10d_logger.py", line 75, in wrapper [default6]:[rank14]: return func(*args, **kwargs) [default6]:[rank14]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/distributed_c10d.py", line 1932, in recv [default6]:[rank14]: pg.recv([tensor], group_src_rank, tag).wait() [default6]:[rank14]: torch.distributed.DistBackendError: NCCL communicator was aborted on rank 1. [default0]:[rank8]: Traceback (most recent call last): [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default0]:[rank8]: trainer.train(dataloader) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default0]:[rank8]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default0]:[rank8]: outputs = self.pipeline_engine.train_batch_iter( [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 278, in train_batch_iter [default0]:[rank8]: output = self.forward(context=context, state=state, micro_batch=micro_batch, model=model) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 44, in forward [default0]:[rank8]: output = model(**micro_batch) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default0]:[rank8]: return self._call_impl(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default0]:[rank8]: return forward_call(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 891, in forward [default0]:[rank8]: sharded_logits = self.model( [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default0]:[rank8]: return self._call_impl(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default0]:[rank8]: return forward_call(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 764, in forward [default0]:[rank8]: return self.forward_with_hidden_states(input_ids=input_ids, input_mask=input_mask)[0] [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 780, in forward_with_hidden_states [default0]:[rank8]: hidden_encoder_states = encoder_block(**hidden_encoder_states) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default0]:[rank8]: return self._call_impl(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default0]:[rank8]: return forward_call(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/block.py", line 126, in forward [default0]:[rank8]: new_kwargs[name] = recv_from_pipeline_state_buffer( [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/functional.py", line 117, in recv_from_pipeline_state_buffer [default0]:[rank8]: pipeline_state.run_communication() [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 150, in run_communication [default0]:[rank8]: recv_activation_tensor = recv_activation() [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 31, in __call__ [default0]:[rank8]: return self.p2p.recv_tensors(num_tensors=1, from_rank=self.from_rank)[0] [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 353, in recv_tensors [default0]:[rank8]: buffers, futures = self.irecv_tensors(num_tensors=num_tensors, from_rank=from_rank, tag=tag) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 326, in irecv_tensors [default0]:[rank8]: meta = self._recv_meta(from_rank=from_rank, tag=tag) [default0]:[rank8]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 269, in _recv_meta [default0]:[rank8]: dist.recv( [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/c10d_logger.py", line 75, in wrapper [default0]:[rank8]: return func(*args, **kwargs) [default0]:[rank8]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/distributed_c10d.py", line 1932, in recv [default0]:[rank8]: pg.recv([tensor], group_src_rank, tag).wait() [default0]:[rank8]: torch.distributed.DistBackendError: NCCL communicator was aborted on rank 1. [default7]:[rank15]: Traceback (most recent call last): [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py", line 237, in [default7]:[rank15]: trainer.train(dataloader) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 429, in train [default7]:[rank15]: outputs, loss_avg = self.training_step(dataloader=self.current_dataloader) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/trainer.py", line 462, in training_step [default7]:[rank15]: outputs = self.pipeline_engine.train_batch_iter( [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 278, in train_batch_iter [default7]:[rank15]: output = self.forward(context=context, state=state, micro_batch=micro_batch, model=model) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/engine.py", line 44, in forward [default7]:[rank15]: output = model(**micro_batch) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default7]:[rank15]: return self._call_impl(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default7]:[rank15]: return forward_call(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 891, in forward [default7]:[rank15]: sharded_logits = self.model( [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default7]:[rank15]: return self._call_impl(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default7]:[rank15]: return forward_call(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 764, in forward [default7]:[rank15]: return self.forward_with_hidden_states(input_ids=input_ids, input_mask=input_mask)[0] [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/models/llama.py", line 780, in forward_with_hidden_states [default7]:[rank15]: hidden_encoder_states = encoder_block(**hidden_encoder_states) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl [default7]:[rank15]: return self._call_impl(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl [default7]:[rank15]: return forward_call(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/block.py", line 126, in forward [default7]:[rank15]: new_kwargs[name] = recv_from_pipeline_state_buffer( [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/functional.py", line 117, in recv_from_pipeline_state_buffer [default7]:[rank15]: pipeline_state.run_communication() [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 150, in run_communication [default7]:[rank15]: recv_activation_tensor = recv_activation() [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/state.py", line 31, in __call__ [default7]:[rank15]: return self.p2p.recv_tensors(num_tensors=1, from_rank=self.from_rank)[0] [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 353, in recv_tensors [default7]:[rank15]: buffers, futures = self.irecv_tensors(num_tensors=num_tensors, from_rank=from_rank, tag=tag) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 326, in irecv_tensors [default7]:[rank15]: meta = self._recv_meta(from_rank=from_rank, tag=tag) [default7]:[rank15]: File "/fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/src/nanotron/parallel/pipeline_parallel/p2p.py", line 269, in _recv_meta [default7]:[rank15]: dist.recv( [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/c10d_logger.py", line 75, in wrapper [default7]:[rank15]: return func(*args, **kwargs) [default7]:[rank15]: File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/distributed_c10d.py", line 1932, in recv [default7]:[rank15]: pg.recv([tensor], group_src_rank, tag).wait() [default7]:[rank15]: torch.distributed.DistBackendError: NCCL communicator was aborted on rank 1. [default1]:[rank9]:[E ProcessGroupNCCL.cpp:577] [Rank 1] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [default1]:[rank9]:[E ProcessGroupNCCL.cpp:583] [Rank 1] To avoid data inconsistency, we are taking the entire process down. [default1]:[rank9]:[E ProcessGroupNCCL.cpp:1414] [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default1]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default1]:Last error: [default1]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<50190> [default1]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default1]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7fc01ef21897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default1]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7fc0201fa5f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7fc0201fa83c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7fc0201ffa60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7fc020200dcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #5: + 0xd3e95 (0x7fc06bc99e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default1]:frame #6: + 0x8609 (0x7fc070ce0609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default1]:frame #7: clone + 0x43 (0x7fc070aab353 in /lib/x86_64-linux-gnu/libc.so.6) [default1]: [default1]:terminate called after throwing an instance of 'c10::DistBackendError' [default1]: what(): [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default1]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default1]:Last error: [default1]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<50190> [default1]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default1]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7fc01ef21897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default1]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7fc0201fa5f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7fc0201fa83c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7fc0201ffa60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7fc020200dcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #5: + 0xd3e95 (0x7fc06bc99e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default1]:frame #6: + 0x8609 (0x7fc070ce0609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default1]:frame #7: clone + 0x43 (0x7fc070aab353 in /lib/x86_64-linux-gnu/libc.so.6) [default1]: [default1]:Exception raised from ncclCommWatchdog at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1418 (most recent call first): [default1]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7fc01ef21897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default1]:frame #1: + 0xe32119 (0x7fc01fe84119 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default1]:frame #2: + 0xd3e95 (0x7fc06bc99e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default1]:frame #3: + 0x8609 (0x7fc070ce0609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default1]:frame #4: clone + 0x43 (0x7fc070aab353 in /lib/x86_64-linux-gnu/libc.so.6) [default1]: [default0]:[rank8]:[E ProcessGroupNCCL.cpp:577] [Rank 1] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [default0]:[rank8]:[E ProcessGroupNCCL.cpp:583] [Rank 1] To avoid data inconsistency, we are taking the entire process down. [default0]:[rank8]:[E ProcessGroupNCCL.cpp:1414] [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default0]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default0]:Last error: [default0]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<33166> [default0]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default0]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f49d716b897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default0]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7f49d84445f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7f49d844483c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7f49d8449a60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7f49d844adcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #5: + 0xd3e95 (0x7f4a23ee3e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default0]:frame #6: + 0x8609 (0x7f4a28f2a609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default0]:frame #7: clone + 0x43 (0x7f4a28cf5353 in /lib/x86_64-linux-gnu/libc.so.6) [default0]: [default0]:terminate called after throwing an instance of 'c10::DistBackendError' [default0]: what(): [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default0]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default0]:Last error: [default0]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<33166> [default0]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default0]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f49d716b897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default0]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7f49d84445f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7f49d844483c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7f49d8449a60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7f49d844adcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #5: + 0xd3e95 (0x7f4a23ee3e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default0]:frame #6: + 0x8609 (0x7f4a28f2a609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default0]:frame #7: clone + 0x43 (0x7f4a28cf5353 in /lib/x86_64-linux-gnu/libc.so.6) [default0]: [default0]:Exception raised from ncclCommWatchdog at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1418 (most recent call first): [default0]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f49d716b897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default0]:frame #1: + 0xe32119 (0x7f49d80ce119 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default0]:frame #2: + 0xd3e95 (0x7f4a23ee3e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default0]:frame #3: + 0x8609 (0x7f4a28f2a609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default0]:frame #4: clone + 0x43 (0x7f4a28cf5353 in /lib/x86_64-linux-gnu/libc.so.6) [default0]: [default6]:[rank14]:[E ProcessGroupNCCL.cpp:577] [Rank 1] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [default6]:[rank14]:[E ProcessGroupNCCL.cpp:583] [Rank 1] To avoid data inconsistency, we are taking the entire process down. [default6]:[rank14]:[E ProcessGroupNCCL.cpp:1414] [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default6]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default6]:Last error: [default6]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<53328> [default6]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default6]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f51dbab3897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default6]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7f51dcd8c5f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7f51dcd8c83c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7f51dcd91a60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7f51dcd92dcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #5: + 0xd3e95 (0x7f522882be95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default6]:frame #6: + 0x8609 (0x7f522d872609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default6]:frame #7: clone + 0x43 (0x7f522d63d353 in /lib/x86_64-linux-gnu/libc.so.6) [default6]: [default6]:terminate called after throwing an instance of 'c10::DistBackendError' [default6]: what(): [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default6]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default6]:Last error: [default6]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<53328> [default6]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default6]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f51dbab3897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default6]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7f51dcd8c5f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7f51dcd8c83c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7f51dcd91a60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7f51dcd92dcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #5: + 0xd3e95 (0x7f522882be95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default6]:frame #6: + 0x8609 (0x7f522d872609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default6]:frame #7: clone + 0x43 (0x7f522d63d353 in /lib/x86_64-linux-gnu/libc.so.6) [default6]: [default6]:Exception raised from ncclCommWatchdog at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1418 (most recent call first): [default6]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7f51dbab3897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default6]:frame #1: + 0xe32119 (0x7f51dca16119 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default6]:frame #2: + 0xd3e95 (0x7f522882be95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default6]:frame #3: + 0x8609 (0x7f522d872609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default6]:frame #4: clone + 0x43 (0x7f522d63d353 in /lib/x86_64-linux-gnu/libc.so.6) [default6]: [default7]:[rank15]:[E ProcessGroupNCCL.cpp:577] [Rank 1] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [default7]:[rank15]:[E ProcessGroupNCCL.cpp:583] [Rank 1] To avoid data inconsistency, we are taking the entire process down. [default7]:[rank15]:[E ProcessGroupNCCL.cpp:1414] [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default7]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default7]:Last error: [default7]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<54772> [default7]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default7]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7ff170e30897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default7]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7ff1721095f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7ff17210983c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7ff17210ea60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7ff17210fdcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #5: + 0xd3e95 (0x7ff1bdba8e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default7]:frame #6: + 0x8609 (0x7ff1c2bef609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default7]:frame #7: clone + 0x43 (0x7ff1c29ba353 in /lib/x86_64-linux-gnu/libc.so.6) [default7]: [default7]:terminate called after throwing an instance of 'c10::DistBackendError' [default7]: what(): [PG 4 Rank 1] Process group watchdog thread terminated with exception: NCCL error: remote process exited or there was a network error, NCCL version 2.20.5 [default7]:ncclRemoteError: A call failed possibly due to a network error or a remote process exiting prematurely. [default7]:Last error: [default7]:socketProgress: Connection closed by remote peer ip-26-0-163-134.ec2.internal<54772> [default7]:Exception raised from checkForNCCLErrorsInternal at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1723 (most recent call first): [default7]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7ff170e30897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default7]:frame #1: c10d::ProcessGroupNCCL::checkForNCCLErrorsInternal(std::shared_ptr&) + 0x220 (0x7ff1721095f0 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #2: c10d::ProcessGroupNCCL::WorkNCCL::checkAndSetException() + 0x7c (0x7ff17210983c in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #3: c10d::ProcessGroupNCCL::watchdogHandler() + 0x180 (0x7ff17210ea60 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #4: c10d::ProcessGroupNCCL::ncclCommWatchdog() + 0x10c (0x7ff17210fdcc in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #5: + 0xd3e95 (0x7ff1bdba8e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default7]:frame #6: + 0x8609 (0x7ff1c2bef609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default7]:frame #7: clone + 0x43 (0x7ff1c29ba353 in /lib/x86_64-linux-gnu/libc.so.6) [default7]: [default7]:Exception raised from ncclCommWatchdog at ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1418 (most recent call first): [default7]:frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x57 (0x7ff170e30897 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libc10.so) [default7]:frame #1: + 0xe32119 (0x7ff171d93119 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/lib/libtorch_cuda.so) [default7]:frame #2: + 0xd3e95 (0x7ff1bdba8e95 in /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/../lib/libstdc++.so.6) [default7]:frame #3: + 0x8609 (0x7ff1c2bef609 in /lib/x86_64-linux-gnu/libpthread.so.0) [default7]:frame #4: clone + 0x43 (0x7ff1c29ba353 in /lib/x86_64-linux-gnu/libc.so.6) [default7]: W0702 16:18:31.870000 140033536743232 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 1452817 closing signal SIGTERM W0702 16:18:31.874000 140033536743232 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 1452818 closing signal SIGTERM W0702 16:18:31.878000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134237 closing signal SIGTERM W0702 16:18:31.879000 140033536743232 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 1452819 closing signal SIGTERM W0702 16:18:31.883000 140033536743232 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 1452820 closing signal SIGTERM W0702 16:18:31.879000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134239 closing signal SIGTERM W0702 16:18:31.880000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134240 closing signal SIGTERM W0702 16:18:31.880000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134241 closing signal SIGTERM W0702 16:18:31.889000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134242 closing signal SIGTERM W0702 16:18:31.897000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:851] Sending process 134244 closing signal SIGTERM E0702 16:18:33.414000 140694550624064 torch/distributed/elastic/multiprocessing/api.py:826] failed (exitcode: -6) local_rank: 1 (pid: 134238) of binary: /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/python3.10 Traceback (most recent call last): File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/torchrun", line 8, in sys.exit(main()) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 347, in wrapper return f(*args, **kwargs) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/run.py", line 879, in main run(args) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/run.py", line 870, in run elastic_launch( File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 132, in __call__ return launch_agent(self._config, self._entrypoint, list(args)) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 263, in launch_agent raise ChildFailedError( torch.distributed.elastic.multiprocessing.errors.ChildFailedError: ============================================================ /fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py FAILED ------------------------------------------------------------ Failures: [1]: time : 2024-07-02_16:18:31 host : ip-26-0-163-147.ec2.internal rank : 14 (local_rank: 6) exitcode : -6 (pid: 134243) error_file: traceback : Signal 6 (SIGABRT) received by PID 134243 ------------------------------------------------------------ Root Cause (first observed failure): [0]: time : 2024-07-02_16:18:31 host : ip-26-0-163-147.ec2.internal rank : 9 (local_rank: 1) exitcode : -6 (pid: 134238) error_file: traceback : Signal 6 (SIGABRT) received by PID 134238 ============================================================ E0702 16:18:33.821000 140033536743232 torch/distributed/elastic/multiprocessing/api.py:826] failed (exitcode: 1) local_rank: 0 (pid: 1452815) of binary: /fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/python3.10 srun: error: ip-26-0-163-147: task 1: Exited with exit code 1 Traceback (most recent call last): File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/bin/torchrun", line 8, in sys.exit(main()) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 347, in wrapper return f(*args, **kwargs) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/run.py", line 879, in main run(args) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/run.py", line 870, in run elastic_launch( File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 132, in __call__ return launch_agent(self._config, self._entrypoint, list(args)) File "/fsx/ferdinandmom/miniforge3/envs/env-bench-cluster/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 263, in launch_agent raise ChildFailedError( torch.distributed.elastic.multiprocessing.errors.ChildFailedError: ============================================================ /fsx/ferdinandmom/ferdinand-hf/bench_cluster/nanotron/run_train.py FAILED ------------------------------------------------------------ Failures: [1]: time : 2024-07-02_16:18:31 host : ip-26-0-163-134.ec2.internal rank : 1 (local_rank: 1) exitcode : 1 (pid: 1452816) error_file: traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html [2]: time : 2024-07-02_16:18:31 host : ip-26-0-163-134.ec2.internal rank : 6 (local_rank: 6) exitcode : 1 (pid: 1452821) error_file: traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html [3]: time : 2024-07-02_16:18:31 host : ip-26-0-163-134.ec2.internal rank : 7 (local_rank: 7) exitcode : 1 (pid: 1452822) error_file: traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html ------------------------------------------------------------ Root Cause (first observed failure): [0]: time : 2024-07-02_16:18:31 host : ip-26-0-163-134.ec2.internal rank : 0 (local_rank: 0) exitcode : 1 (pid: 1452815) error_file: traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html ============================================================ srun: error: ip-26-0-163-134: task 0: Exited with exit code 1 Consider using `hf_transfer` for faster uploads. This solution comes with some limitations. See https://huggingface.co/docs/huggingface_hub/hf_transfer for more details.