[03/05 16:17:17 libai]: Rank of current process: 0. World size: 8
[03/05 16:17:17 libai]: Command line arguments: Namespace(config_file='configs/bert_large_pretrain.py', eval_only=False, fast_dev_run=False, opts=['model.cfg.hidden_dropout_prob=0.1', 'model.cfg.attention_probs_dropout_prob=0.1', 'model.cfg.bias_dropout_fusion=true', 'model.cfg.hidden_layers=24', 'model.cfg.hidden_size=1024', 'model.cfg.num_attention_heads=16', 'model.cfg.intermediate_size=4096', 'model.cfg.ffn_hidden_size=4096', 'model.cfg.head_size=64', 'graph.enabled=true', 'train.dist.pipeline_num_layers=24', 'train.train_micro_batch_size=32', 'train.global_batch_size=512', 'train.dist.tensor_parallel_size=1', 'train.dist.pipeline_parallel_size=4', 'train.amp.enabled=true', 'train.activation_checkpoint.enabled=true', 'train.num_accumulation_steps=8', 'train.evaluation.enabled=false', 'train.train_iter=220', 'train.train_epoch=0', 'train.log_period=100', 'train.zero_optimization.enabled=true', 'train.zero_optimization.stage=2', 'train.load_weight=', 'train.output_dir=test_logs/oneflow-28/NVIDIA_GeForce_RTX_3080_Ti/7d07caf/LibAI_bert_large_pretrain_graph_nl24_nah16_hs1024_FP16_actrue_DP2_MP1_PP4_zerotrue_stage2_mbs32_gbs512_acc8_1n8g'], resume=False)
[03/05 16:17:17 libai]: Contents of args.config_file=configs/bert_large_pretrain.py:
from libai.config import LazyCall
from libai.evaluation import PPLEvaluator
from .common.models.bert import pretrain_model as model
from .common.models.graph import graph
from .common.train import train
from .common.optim import optim
from .common.data.bert_dataset import dataloader, tokenization

vocab_file = "./data_test/bert_data/bert-base-chinese-vocab.txt"
data_prefix = "./data_test/bert_data/loss_compara_content_sentence"

tokenization.tokenizer.vocab_file = vocab_file
dataloader.train.dataset[0].data_prefix = data_prefix
dataloader.train.dataset[0].indexed_dataset.data_prefix = data_prefix

# Bert-large model config
model.cfg.num_attention_heads = 16
model.cfg.hidden_size = 768
model.cfg.hidden_layers = 8

train.input_placement_device = "cpu"

train.dist.pipeline_num_layers = model.cfg.hidden_layers

train.train_micro_batch_size = 16

train.amp.enabled = True

for ds in dataloader.train.dataset:
    ds.max_seq_length = model.cfg.max_position_embeddings

train.evaluation.evaluator = LazyCall(PPLEvaluator)()

train.output_dir = "output/bert_output"

[03/05 16:17:18 libai]: Full config saved to test_logs/oneflow-28/NVIDIA_GeForce_RTX_3080_Ti/7d07caf/LibAI_bert_large_pretrain_graph_nl24_nah16_hs1024_FP16_actrue_DP2_MP1_PP4_zerotrue_stage2_mbs32_gbs512_acc8_1n8g/config.yaml
[03/05 16:17:18 lb.engine.default]: > compiling dataset index builder ...
make: Entering directory '/ssd/home/ouyangyu/libai_week_test/libai/libai/data/data_utils'
make: Nothing to be done for 'default'.
make: Leaving directory '/ssd/home/ouyangyu/libai_week_test/libai/libai/data/data_utils'
[03/05 16:17:18 lb.engine.default]: >>> done with dataset index builder. Compilation time: 0.054 seconds
[03/05 16:17:18 lb.engine.default]: >>> done with compiling. Compilation time: 0.055 seconds
[03/05 16:17:18 lb.engine.default]: Prepare training, validating, testing set
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: building dataset index ...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: warming up index mmap file...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: reading sizes...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: reading pointers...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: reading document index...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: warming up data mmap file...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: creating numpy buffer of mmap...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: creating memory view of numpy buffer...
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: Finished creating indexed dataset in 0.076941 seconds
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: indexed dataset stats:
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: number of documents: 50000
[03/05 16:17:18 lb.data.data_utils.indexed_dataset]: number of sentences: 1249934
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > WARNING: could not find index map file ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_112640mns_509msl_0.10ssp_1234s.npy, building the indices on rank 0 ...
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > building samples index mapping for bert ...
    using uint32 for data mapping...
    using:
     number of documents:            47450
     sentences range:                [0, 1188464)
     total number of sentences:      1188464
     number of epochs:               2147483646
     maximum number of samples:      112640
     maximum sequence length:        509
     short sequence probability:     0.1
     short sequence ration (1/prob): 10
     seed:                           1234
    reached 112640 samples after 1 epochs ...
   number of empty documents: 0
   number of documents with one sentence: 711
   number of documents with long sentences: 2092
   will create mapping for 113036 samples
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > done building samples index maping
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > saved the index mapping in ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_112640mns_509msl_0.10ssp_1234s.npy
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > elapsed time to build and save samples mapping (seconds): 0.014842
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > loading indexed mapping from ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_112640mns_509msl_0.10ssp_1234s.npy
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     loaded indexed file in 0.003 seconds
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     total number of samples: 113036
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > WARNING: could not find index map file ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_64mns_509msl_0.10ssp_1234s.npy, building the indices on rank 0 ...
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > building samples index mapping for bert ...
    using uint32 for data mapping...
    using:
     number of documents:            2500
     sentences range:                [1188464, 1248643)
     total number of sentences:      60179
     number of epochs:               2147483646
     maximum number of samples:      64
     maximum sequence length:        509
     short sequence probability:     0.1
     short sequence ration (1/prob): 10
     seed:                           1234
    reached 64 samples after 1 epochs ...
   number of empty documents: 0
   number of documents with one sentence: 51
   number of documents with long sentences: 89
   will create mapping for 5884 samples
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > done building samples index maping
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > saved the index mapping in ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_64mns_509msl_0.10ssp_1234s.npy
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > elapsed time to build and save samples mapping (seconds): 0.001019
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > loading indexed mapping from ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_64mns_509msl_0.10ssp_1234s.npy
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     loaded indexed file in 0.000 seconds
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     total number of samples: 5884
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:  > loading indexed mapping from ./data_test/bert_data/loss_compara_content_sentence_bert_indexmap_64mns_509msl_0.10ssp_1234s.npy
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     loaded indexed file in 0.000 seconds
[03/05 16:17:18 lb.data.data_utils.dataset_utils]:     total number of samples: 5884
[03/05 16:17:27 lb.engine.default]: Auto-scaling the config to train.train_iter=220, train.warmup_iter=0
[03/05 16:17:27 libai]: > Start building model...
[03/05 16:17:29 lb.engine.default]: Model:
BertForPreTraining(
  (bert): BertModel(
    (embeddings): BertEmbeddings(
      (vocab_embeddings): VocabEmbedding(num_embeddings=21248, embedding_dim=1024)
      (position_embeddings): Embedding(num_embeddings=512, embedding_dim=1024)
      (tokentype_embeddings): Embedding(num_embeddings=2, embedding_dim=1024)
      (embedding_dropout): Dropout(p=0.1, inplace=False)
    )
    (extended_attn_mask): BertExtendedAttnMask()
    (encoders): ModuleList(
      (0): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (1): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (2): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (3): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (4): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (5): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (6): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (7): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (8): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (9): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (10): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (11): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (12): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (13): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (14): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (15): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (16): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (17): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (18): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (19): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (20): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (21): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (22): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
      (23): TransformerLayer(
        (drop_path): Identity()
        (input_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (self_attention): MultiheadAttention(
          hidden_size=1024, num_heads=16, is_cross_attention=False
          (dropout): Dropout(p=0.1, inplace=False)
          (query_key_value): Linear1D(in_features=1024, out_features=3072, bias=True, parallel=col)
          (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=row)
        )
        (post_attention_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
        (mlp): MLP(
          bias_gelu_fusion=True, bias_dropout_fusion=True, dropout=0.1
          (dense_h_to_4h): Linear1D(in_features=1024, out_features=4096, bias=True, parallel=col)
          (dense_4h_to_h): Linear1D(in_features=4096, out_features=1024, bias=True, parallel=row)
        )
      )
    )
    (final_layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
    (pooler): BertPooler(
      (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=col)
      (activation_func): Tanh()
    )
  )
  (cls_head): BertPreTrainingHeads(
    (predictions): BertLMPredictionHead(
      (dense): Linear1D(in_features=1024, out_features=1024, bias=True, parallel=data)
      (activation_func): GELU()
      (layernorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
    )
    (seq_relationship): Linear1D(in_features=1024, out_features=2, bias=True, parallel=data)
    (lm_logits): LMLogits()
    (loss_func): BertLoss(
      (lm_loss): ParallelCrossEntropyLoss()
    )
  )
)
[03/05 16:17:30 libai]: >>> done with building model. Building time: 2.211 seconds
WARNING [03/05 16:17:30 lb.scheduler.lr_scheduler]: warmup iters equals to zero, return CosineLR
[03/05 16:17:30 lb.engine.trainer]: Starting training from iteration 0
[03/05 16:17:30 lb.models.utils.graph_base]: Start compiling the train graph which may take some time. Please wait for a moment ...
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:50.333, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.334, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:50.337, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:50.338, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:50.340, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.339, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.340, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.342, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.343, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.344, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.344, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
2023/03/05 16:22:50.345, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:50.346, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.347, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
2023/03/05 16:22:50.348, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.349, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.350, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.351, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.352, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.353, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.354, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.354, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.357, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.358, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
2023/03/05 16:22:50.360, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.361, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.362, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.362, NVIDIA GeForce RTX 3080 Ti, 515.65.01timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.363, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.365, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.366, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.372, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.374, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.375, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 4 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.375, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 93 %, 48 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:50.376, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 4 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.378, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.379, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.380, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 4 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.382, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.382, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 48 %, 12288 MiB, 5419 MiB, 6634 MiB
2023/03/05 16:22:50.383, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.384, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.386, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.387, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB2023/03/05 16:22:50.388, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.388, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.389, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 4 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.391, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 7 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.392, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:50.396, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.401, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 2 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.405, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 7 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.413, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 4 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:50.423, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 3 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:50.435, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 7 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/05 16:22:53.349, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 84 %, 43 %, 12288 MiB, 4836 MiB, 7217 MiB
2023/03/05 16:22:53.350, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 85 %, 45 %, 12288 MiB, 5419 MiB, 6634 MiB
2023/03/05 16:22:53.351, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:53.352, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 20 %, 6 %, 12288 MiB, 5775 MiB, 6278 MiB
2023/03/05 16:22:53.353, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:53.353, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 5793 MiB, 6260 MiB
2023/03/05 16:22:53.354, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 1 %, 1 %, 12288 MiB, 8743 MiB, 3310 MiB
2023/03/05 16:22:53.355, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 0 %, 0 %, 12288 MiB, 8743 MiB, 3310 MiB
[03/05 16:22:56 lb.utils.events]:  eta: 0:06:03  iteration: 99/220  consumed_samples: 51200  total_loss: 7.944  lm_loss: 7.242  sop_loss: 0.696  time: 3.0181 s/iter  data_time: 0.0111 s/iter total_throughput: 169.64 samples/s lr: 5.82e-05  
[03/05 16:27:59 lb.utils.events]:  eta: 0:01:00  iteration: 199/220  consumed_samples: 102400  total_loss: 7.902  lm_loss: 7.207  sop_loss: 0.6948  time: 3.0246 s/iter  data_time: 0.0114 s/iter total_throughput: 169.28 samples/s lr: 3.21e-06  
[03/05 16:29:00 lb.utils.events]:  eta: 0:00:00  iteration: 219/220  consumed_samples: 112640  total_loss: 7.896  lm_loss: 7.201  sop_loss: 0.6945  time: 3.0247 s/iter  data_time: 0.0121 s/iter total_throughput: 169.27 samples/s lr: 1.01e-06  
[03/05 16:29:00 lb.engine.hooks]: Overall training speed: 218 iterations in 0:10:59 (3.0247 s / it)
[03/05 16:29:00 lb.engine.hooks]: Total training time: 0:10:59 (0:00:00 on hooks)
*****************************************
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. 
*****************************************
oneflow-version(git_commit)=0.9.1.dev20230304+cu117
oneflow-commit(git_commit)=7d07caf
oneflow-libai(git_commit)=50a973dc5de635b8613ad7666c073c763e238850