+ TRAINER=pretrain + SETUP=for-newvocab ++ cd pya0 ++ pwd ++ git rev-parse HEAD + CODE_VER='/home/w32zhong/projects/rrg-jimmylin/w32zhong/pya0 8f207c0036a9f81f91e26f7ecedcfa84025ae680' + COMMAND='/var/spool/slurmd/job25031358/slurm_script pretrain for-newvocab' + EPOCHS=40 + TEST_CYCLE=100 + case $TRAINER-${SETUP} in + DEV_BSIZE=8 + SAVE_FOLD=10 + DATA_VER=arjmPWtGwzKrkmR + START_POINT=bert-base-uncased + TOK_CKPOINT=bert-tokenizer + SHARDS_LIST=shards-for-newvocab.txt + TEST_FILE=test.txt + EXTRA_DAT=mse-aops-2021-vocab.pkl + EXTRA_ARG= + DATA_DIR=data.arjmPWtGwzKrkmR + set -e + '[' '!' -e data.arjmPWtGwzKrkmR ']' + set +e ++ cat /var/spool/slurmd/job25031358/slurm_script ++ grep -Po '(?<=SBATCH --nodes=)[0-9]+' + N_NODE=4 ++ cat /var/spool/slurmd/job25031358/slurm_script ++ grep -Po '(?<=SBATCH --gres=gpu:)[0-9]+' + N_GPUS=2 + export NCCL_BLOCKING_WAIT=1 + NCCL_BLOCKING_WAIT=1 + export SLURM_ACCOUNT=def-jimmylin + SLURM_ACCOUNT=def-jimmylin + export SBATCH_ACCOUNT=def-jimmylin + SBATCH_ACCOUNT=def-jimmylin + export SALLOC_ACCOUNT=def-jimmylin + SALLOC_ACCOUNT=def-jimmylin + which srun /opt/software/slurm/bin/srun ++ hostname + srun --unbuffered python ./pya0/utils/transformer.py pretrain data.arjmPWtGwzKrkmR/bert-base-uncased data.arjmPWtGwzKrkmR/bert-tokenizer data.arjmPWtGwzKrkmR/mse-aops-2021-vocab.pkl --test_file data.arjmPWtGwzKrkmR/test.txt --test_cycle 100 --shards_list data.arjmPWtGwzKrkmR/shards-for-newvocab.txt --cluster tcp://blg4302.int.ets1.calculquebec.ca:8912 --batch_size 64 --save_fold 10 --epochs 40 Loading model data.arjmPWtGwzKrkmR/bert-base-uncased... { "_name_or_path": "data.arjmPWtGwzKrkmR/bert-base-uncased", "add_cross_attention": false, "architectures": [ "BertForPreTraining" ], "attention_probs_dropout_prob": 0.1, "bad_words_ids": null, "bos_token_id": null, "chunk_size_feed_forward": 0, "decoder_start_token_id": null, "diversity_penalty": 0.0, "do_sample": false, "early_stopping": false, "encoder_no_repeat_ngram_size": 0, "eos_token_id": null, "finetuning_task": null, "forced_bos_token_id": null, "forced_eos_token_id": null, "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 768, "id2label": { "0": "LABEL_0", "1": "LABEL_1" }, "initializer_range": 0.02, "intermediate_size": 3072, "is_decoder": false, "is_encoder_decoder": false, "label2id": { "LABEL_0": 0, "LABEL_1": 1 }, "layer_norm_eps": 1e-12, "length_penalty": 1.0, "max_length": 20, "max_position_embeddings": 512, "min_length": 0, "model_type": "bert", "no_repeat_ngram_size": 0, "num_attention_heads": 12, "num_beam_groups": 1, "num_beams": 1, "num_hidden_layers": 12, "num_return_sequences": 1, "output_attentions": false, "output_hidden_states": false, "output_scores": false, "pad_token_id": 0, "position_embedding_type": "absolute", "prefix": null, "problem_type": null, "pruned_heads": {}, "remove_invalid_values": false, "repetition_penalty": 1.0, "return_dict": true, "return_dict_in_generate": false, "sep_token_id": null, "task_specific_params": null, "temperature": 1.0, "tie_encoder_decoder": false, "tie_word_embeddings": true, "tokenizer_class": null, "top_k": 50, "top_p": 1.0, "torch_dtype": null, "torchscript": false, "transformers_version": "4.9.2", "type_vocab_size": 2, "use_bfloat16": false, "use_cache": true, "vocab_size": 30522 } Before loading new vocabulary: 30522 After loading new vocabulary: 31061 Resize model embedding and save new tokenizer ... Invoke training ... [caller] pretrain [node#3 rank#6] Training on device cuda:0 [node#3 rank#6] 2 x Tesla V100-SXM2-16GB: 0% [node#3 rank#7] Training on device cuda:1 [node#3 rank#7] 2 x Tesla V100-SXM2-16GB: 5% [node#3 rank#6] Initialized process group ... [node#3 rank#7] Initialized process group ... Loading model data.arjmPWtGwzKrkmR/bert-base-uncased... Loading model data.arjmPWtGwzKrkmR/bert-base-uncased... Loading model data.arjmPWtGwzKrkmR/bert-base-uncased... { "_name_or_path": "data.arjmPWtGwzKrkmR/bert-base-uncased", "add_cross_attention": false, "architectures": [ "BertForPreTraining" ], "attention_probs_dropout_prob": 0.1, "bad_words_ids": null, "bos_token_id": null, "chunk_size_feed_forward": 0, "decoder_start_token_id": null, "diversity_penalty": 0.0, "do_sample": false, "early_stopping": false, "encoder_no_repeat_ngram_size": 0, "eos_token_id": null, "finetuning_task": null, "forced_bos_token_id": null, "forced_eos_token_id": null, "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 768, "id2label": { "0": "LABEL_0", "1": "LABEL_1" }, "initializer_range": 0.02, "intermediate_size": 3072, "is_decoder": false, "is_encoder_decoder": false, "label2id": { "LABEL_0": 0, "LABEL_1": 1 }, "layer_norm_eps": 1e-12, "length_penalty": 1.0, "max_length": 20, "max_position_embeddings": 512, "min_length": 0, "model_type": "bert", "no_repeat_ngram_size": 0, "num_attention_heads": 12, "num_beam_groups": 1, "num_beams": 1, "num_hidden_layers": 12, "num_return_sequences": 1, "output_attentions": false, "output_hidden_states": false, "output_scores": false, "pad_token_id": 0, "position_embedding_type": "absolute", "prefix": null, "problem_type": null, "pruned_heads": {}, "remove_invalid_values": false, "repetition_penalty": 1.0, "return_dict": true, "return_dict_in_generate": false, "sep_token_id": null, "task_specific_params": null, "temperature": 1.0, "tie_encoder_decoder": false, "tie_word_embeddings": true, "tokenizer_class": null, "top_k": 50, "top_p": 1.0, "torch_dtype": null, "torchscript": false, "transformers_version": "4.9.2", "type_vocab_size": 2, "use_bfloat16": false, "use_cache": true, "vocab_size": 30522 } { "_name_or_path": "data.arjmPWtGwzKrkmR/bert-base-uncased", "add_cross_attention": false, "architectures": [ "BertForPreTraining" ], "attention_probs_dropout_prob": 0.1, "bad_words_ids": null, "bos_token_id": null, "chunk_size_feed_forward": 0, "decoder_start_token_id": null, "diversity_penalty": 0.0, "do_sample": false, "early_stopping": false, "encoder_no_repeat_ngram_size": 0, "eos_token_id": null, "finetuning_task": null, "forced_bos_token_id": null, "forced_eos_token_id": null, "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 768, "id2label": { "0": "LABEL_0", "1": "LABEL_1" }, "initializer_range": 0.02, "intermediate_size": 3072, "is_decoder": false, "is_encoder_decoder": false, "label2id": { "LABEL_0": 0, "LABEL_1": 1 }, "layer_norm_eps": 1e-12, "length_penalty": 1.0, "max_length": 20, "max_position_embeddings": 512, "min_length": 0, "model_type": "bert", "no_repeat_ngram_size": 0, "num_attention_heads": 12, "num_beam_groups": 1, "num_beams": 1, "num_hidden_layers": 12, "num_return_sequences": 1, "output_attentions": false, "output_hidden_states": false, "output_scores": false, "pad_token_id": 0, "position_embedding_type": "absolute", "prefix": null, "problem_type": null, "pruned_heads": {}, "remove_invalid_values": false, "repetition_penalty": 1.0, "return_dict": true, "return_dict_in_generate": false, "sep_token_id": null, "task_specific_params": null, "temperature": 1.0, "tie_encoder_decoder": false, "tie_word_embeddings": true, "tokenizer_class": null, "top_k": 50, "top_p": 1.0, "torch_dtype": null, "torchscript": false, "transformers_version": "4.9.2", "type_vocab_size": 2, "use_bfloat16": false, "use_cache": true, "vocab_size": 30522 } { "_name_or_path": "data.arjmPWtGwzKrkmR/bert-base-uncased", "add_cross_attention": false, "architectures": [ "BertForPreTraining" ], "attention_probs_dropout_prob": 0.1, "bad_words_ids": null, "bos_token_id": null, "chunk_size_feed_forward": 0, "decoder_start_token_id": null, "diversity_penalty": 0.0, "do_sample": false, "early_stopping": false, "encoder_no_repeat_ngram_size": 0, "eos_token_id": null, "finetuning_task": null, "forced_bos_token_id": null, "forced_eos_token_id": null, "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 768, "id2label": { "0": "LABEL_0", "1": "LABEL_1" }, "initializer_range": 0.02, "intermediate_size": 3072, "is_decoder": false, "is_encoder_decoder": false, "label2id": { "LABEL_0": 0, "LABEL_1": 1 }, "layer_norm_eps": 1e-12, "length_penalty": 1.0, "max_length": 20, "max_position_embeddings": 512, "min_length": 0, "model_type": "bert", "no_repeat_ngram_size": 0, "num_attention_heads": 12, "num_beam_groups": 1, "num_beams": 1, "num_hidden_layers": 12, "num_return_sequences": 1, "output_attentions": false, "output_hidden_states": false, "output_scores": false, "pad_token_id": 0, "position_embedding_type": "absolute", "prefix": null, "problem_type": null, "pruned_heads": {}, "remove_invalid_values": false, "repetition_penalty": 1.0, "return_dict": true, "return_dict_in_generate": false, "sep_token_id": null, "task_specific_params": null, "temperature": 1.0, "tie_encoder_decoder": false, "tie_word_embeddings": true, "tokenizer_class": null, "top_k": 50, "top_p": 1.0, "torch_dtype": null, "torchscript": false, "transformers_version": "4.9.2", "type_vocab_size": 2, "use_bfloat16": false, "use_cache": true, "vocab_size": 30522 } Before loading new vocabulary: 30522 Before loading new vocabulary: 30522 After loading new vocabulary: 31061 Resize model embedding and save new tokenizer ... Before loading new vocabulary: 30522 After loading new vocabulary: 31061 Resize model embedding and save new tokenizer ... After loading new vocabulary: 31061 Resize model embedding and save new tokenizer ... Invoke training ... Invoke training ... Invoke training ... [caller] pretrain [caller] pretrain [caller] pretrain [node#0 rank#0] Training on device cuda:0 [node#2 rank#4] Training on device cuda:0 [node#0 rank#0] 2 x Tesla V100-SXM2-16GB: 0% [node#2 rank#4] 2 x Tesla V100-SXM2-16GB: 0% [node#1 rank#2] Training on device cuda:0 [node#1 rank#2] 2 x Tesla V100-SXM2-16GB: 0% [node#0 rank#1] Training on device cuda:1 [node#2 rank#5] Training on device cuda:1 [node#0 rank#1] 2 x Tesla V100-SXM2-16GB: 2% [node#2 rank#5] 2 x Tesla V100-SXM2-16GB: 2% [node#1 rank#3] Training on device cuda:1 [node#1 rank#3] 2 x Tesla V100-SXM2-16GB: 2% [node#2 rank#4] Initialized process group ... [node#0 rank#0] Initialized process group ... [node#1 rank#2] Initialized process group ... [node#0 rank#1] Initialized process group ... [node#2 rank#5] Initialized process group ... [node#1 rank#3] Initialized process group ... [node#0 rank#0] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 0 using best-guess GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#2 rank#4] Enter Torch DDP. [node#0 rank#1] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 1 using best-guess GPU 1 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [W ProcessGroupNCCL.cpp:1569] Rank 4 using best-guess GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#1 rank#3] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 3 using best-guess GPU 1 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#3 rank#7] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 7 using best-guess GPU 1 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#2 rank#5] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 5 using best-guess GPU 1 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#3 rank#6] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 6 using best-guess GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. [node#1 rank#2] Enter Torch DDP. [W ProcessGroupNCCL.cpp:1569] Rank 2 using best-guess GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device. AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) [node#3 rank#7] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#3 rank#7] Start training at: (0, 0, -1) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) [node#3 rank#6] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#3 rank#6] Start training at: (0, 0, -1) [node#3 rank#7] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) AdamW ( Parameter Group 0 betas: (0.9, 0.999) correct_bias: True eps: 1e-06 lr: 1e-06 weight_decay: 0.01 ) [node#1 rank#2] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#1 rank#2] Start training at: (0, 0, -1) Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) [node#3 rank#6] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... [node#1 rank#3] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#1 rank#3] Start training at: (0, 0, -1) [node#2 rank#4] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#2 rank#4] Start training at: (0, 0, -1) [node#2 rank#5] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#2 rank#5] Start training at: (0, 0, -1) Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) [node#2 rank#4] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... [node#2 rank#5] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) Loading test data: data.arjmPWtGwzKrkmR/test.txt (bsize=8) [node#1 rank#3] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... [node#1 rank#2] Loading shard data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730 ... [node#0 rank#1] Shards: ['data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.6632730', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7074912', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7517094', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.7959276', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8401458', 'data.arjmPWtGwzKrkmR/mse-aops-2021-data.pkl.pairs.8843640'] [node#0 rank#1] Start training at: (0, 0, -1)