[03/10 15:11:28 libai]: Rank of current process: 0. World size: 8
[03/10 15:11:28 libai]: Command line arguments: Namespace(config_file='configs/gpt2_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=8', 'train.global_batch_size=256', 'train.dist.tensor_parallel_size=2', 'train.dist.pipeline_parallel_size=1', '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/1ea2bb7/LibAI_gpt2_pretrain_graph_nl24_nah16_hs1024_FP16_actrue_DP4_MP2_PP1_zerotrue_stage2_mbs8_gbs256_acc8_1n8g'], resume=False)
[03/10 15:11:28 libai]: Contents of args.config_file=configs/gpt2_pretrain.py:
from libai.config import LazyCall
from libai.evaluation import PPLEvaluator
from .common.models.gpt import pretrain_model as model
from .common.train import train
from .common.optim import optim
from .common.data.gpt_dataset import dataloader, tokenization

from .common.models.graph import graph

vocab_file = "./data_test/gpt_data/gpt2-vocab.json"
merge_files = "./data_test/gpt_data/gpt2-merges.txt"
data_prefix = "./data_test/gpt_data/loss_compara_content_sentence"

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

# GPT-2 model config
model.cfg.embedding_dropout_prob = 0.1
model.cfg.attention_dropout_prob = 0.1
model.cfg.num_attention_heads = 16
model.cfg.hidden_size = 384
model.cfg.ffn_hidden_size = 1536
model.cfg.hidden_layers = 6
model.cfg.max_seq_length = 1024

train.input_placement_device = "cpu"

train.dist.pipeline_num_layers = model.cfg.hidden_layers

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

optim.lr = 1.5e-4

train.train_micro_batch_size = 4
train.amp.enabled = True

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

train.output_dir = "./output/gpt2_output"

[03/10 15:11:28 libai]: Full config saved to test_logs/oneflow-28/NVIDIA_GeForce_RTX_3080_Ti/1ea2bb7/LibAI_gpt2_pretrain_graph_nl24_nah16_hs1024_FP16_actrue_DP4_MP2_PP1_zerotrue_stage2_mbs8_gbs256_acc8_1n8g/config.yaml
[03/10 15:11:28 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'
[03/10 15:11:29 lb.engine.default]: >>> done with dataset index builder. Compilation time: 0.053 seconds
[03/10 15:11:29 lb.engine.default]: >>> done with compiling. Compilation time: 0.055 seconds
[03/10 15:11:29 lb.engine.default]: Prepare training, validating, testing set
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: building dataset index ...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: warming up index mmap file...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: reading sizes...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: reading pointers...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: reading document index...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: warming up data mmap file...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: creating numpy buffer of mmap...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: creating memory view of numpy buffer...
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: Finished creating indexed dataset in 0.072962 seconds
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: indexed dataset stats:
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: number of documents: 50000
[03/10 15:11:29 lb.data.data_utils.indexed_dataset]: number of sentences: 1249934
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading doc-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_56320ns_1024sl_1234s_doc_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading sample-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_56320ns_1024sl_1234s_sample_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading shuffle-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_56320ns_1024sl_1234s_shuffle_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     loaded indexed file in 0.004 seconds
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of samples: 57333
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of epochs: 1
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading doc-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_doc_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading sample-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_sample_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading shuffle-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_shuffle_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     loaded indexed file in 0.001 seconds
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of samples: 57333
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of epochs: 1
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading doc-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_doc_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading sample-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_sample_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:  > loading shuffle-idx mapping from ./data_test/gpt_data/loss_compara_content_sentence_gpt-2_indexmap_128ns_1024sl_1234s_shuffle_idx.npy
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     loaded indexed file in 0.001 seconds
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of samples: 57333
[03/10 15:11:29 lb.data.datasets.gpt_dataset]:     total number of epochs: 1
[03/10 15:11:38 lb.engine.default]: Auto-scaling the config to train.train_iter=220, train.warmup_iter=0
[03/10 15:11:38 libai]: > Start building model...
[03/10 15:11:40 lb.engine.default]: Model:
GPTForPreTraining(
  (GPT_model): GPTModel(
    (embeddings): GPTEmbedding(
      (token_embeddings): VocabEmbedding(num_embeddings=50432, embedding_dim=1024)
      (position_embeddings): Embedding(num_embeddings=1024, embedding_dim=1024)
      (dropout): Dropout(p=0.1, inplace=False)
    )
    (transformer): Transformer(
      (layers): 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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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
            (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)
          )
        )
      )
      (layernorm_f): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
    )
    (lm_head): LMLogits()
  )
  (loss_func): GPTLoss(
    (lm_loss): ParallelCrossEntropyLoss()
  )
)
[03/10 15:11:40 libai]: >>> done with building model. Building time: 1.995 seconds
WARNING [03/10 15:11:40 lb.scheduler.lr_scheduler]: warmup iters equals to zero, return CosineLR
[03/10 15:11:40 lb.engine.trainer]: Starting training from iteration 0
W20230310 15:11:40.838986 3119500 eager_local_op_interpreter.cpp:256] Casting a local tensor to a global tensor with Broadcast sbp will modify the data of input! If you want to keep the input local tensor unchanged, please set the arg copy to True.
[03/10 15:11:40 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]
2023/03/10 15:24:22.148, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.150, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
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]
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/10 15:24:22.152, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 25 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.153, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.154, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.155, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/10 15:24:22.156, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 22 %, 12288 MiB, 6841 MiB, 5212 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/10 15:24:22.157, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.159, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.161, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.162, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 24 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.161, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/10 15:24:22.164, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 25 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.163, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.166, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 4 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.166, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 4 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.168, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.169, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.170, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 22 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.169, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 23 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.172, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.180, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 22 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.181, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 22 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.183, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.185, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 4 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.186, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.187, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 99 %, 15 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.188, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 4 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.191, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.192, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.194, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.196, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 7 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.197, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.198, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 4 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.198, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 7 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.201, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.203, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.206, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.208, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.208, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 7 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.209, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.212, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.214, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.216, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.218, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.219, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:22.220, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 20 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.223, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.225, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.227, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.230, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 10 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.231, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.237, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.237, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 94 %, 9 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.239, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:22.243, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 96 %, 6 %, 12288 MiB, 6843 MiB, 5210 MiB
timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB]
2023/03/10 15:24:29.789, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 33 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:29.790, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 26 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:29.791, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 78 %, 13 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:29.792, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 33 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:29.793, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 70 %, 10 %, 12288 MiB, 6841 MiB, 5212 MiB
2023/03/10 15:24:29.794, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 100 %, 33 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:29.794, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 95 %, 17 %, 12288 MiB, 6843 MiB, 5210 MiB
2023/03/10 15:24:29.795, NVIDIA GeForce RTX 3080 Ti, 515.65.01, 75 %, 11 %, 12288 MiB, 6843 MiB, 5210 MiB
[03/10 15:24:37 lb.utils.events]:  eta: 0:14:36  iteration: 99/220  consumed_samples: 25600  total_loss: 7.252  time: 7.3364 s/iter  data_time: 0.0061 s/iter total_throughput: 34.89 samples/s lr: 8.74e-05  
[03/10 15:36:55 lb.utils.events]:  eta: 0:02:27  iteration: 199/220  consumed_samples: 51200  total_loss: 6.944  time: 7.3598 s/iter  data_time: 0.0066 s/iter total_throughput: 34.78 samples/s lr: 4.81e-06  
[03/10 15:39:24 lb.utils.events]:  eta: 0:00:00  iteration: 219/220  consumed_samples: 56320  total_loss: 6.696  time: 7.3665 s/iter  data_time: 0.0065 s/iter total_throughput: 34.75 samples/s lr: 1.51e-06  
[03/10 15:39:24 lb.engine.hooks]: Overall training speed: 218 iterations in 0:26:45 (7.3665 s / it)
[03/10 15:39:24 lb.engine.hooks]: Total training time: 0:26:45 (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.dev20230309+cu117
oneflow-commit(git_commit)=1ea2bb7
oneflow-libai(git_commit)=50a973dc5de635b8613ad7666c073c763e238850