[2024-01-26 12:54:39,523] torch.distributed.run: [WARNING] master_addr is only used for static rdzv_backend and when rdzv_endpoint is not specified. 01/26/2024 12:54:44 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1distributed training: True, 16-bits training: False 01/26/2024 12:54:44 - INFO - __main__ - Training/evaluation parameters Seq2SeqTrainingArguments( _n_gpu=1, adafactor=False, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, auto_find_batch_size=False, bf16=False, bf16_full_eval=False, data_seed=None, dataloader_drop_last=False, dataloader_num_workers=0, dataloader_persistent_workers=False, dataloader_pin_memory=True, ddp_backend=None, ddp_broadcast_buffers=None, ddp_bucket_cap_mb=None, ddp_find_unused_parameters=False, ddp_timeout=1800, debug=[], deepspeed=None, disable_tqdm=False, dispatch_batches=None, do_eval=False, do_predict=False, do_train=False, eval_accumulation_steps=None, eval_delay=0, eval_steps=None, evaluation_strategy=no, fp16=False, fp16_backend=auto, fp16_full_eval=False, fp16_opt_level=O1, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_grad_ckpt': False}, fsdp_min_num_params=0, fsdp_transformer_layer_cls_to_wrap=None, full_determinism=False, generation_config=None, generation_max_length=None, generation_num_beams=None, gradient_accumulation_steps=32, gradient_checkpointing=False, gradient_checkpointing_kwargs=None, greater_is_better=None, group_by_length=False, half_precision_backend=auto, hub_always_push=False, hub_model_id=None, hub_private_repo=False, hub_strategy=every_save, hub_token=, ignore_data_skip=False, include_inputs_for_metrics=False, include_num_input_tokens_seen=False, include_tokens_per_second=False, jit_mode_eval=False, label_names=None, label_smoothing_factor=0.0, learning_rate=0.02, length_column_name=length, load_best_model_at_end=False, local_rank=0, log_level=passive, log_level_replica=warning, log_on_each_node=True, logging_dir=output/privacy_detection_pt-20240126-125436-128-2e-2/runs/Jan26_12-54-44_ubuntu1804, logging_first_step=False, logging_nan_inf_filter=True, logging_steps=1.0, logging_strategy=steps, lr_scheduler_kwargs={}, lr_scheduler_type=linear, max_grad_norm=1.0, max_steps=100, metric_for_best_model=None, mp_parameters=, neftune_noise_alpha=None, no_cuda=False, num_train_epochs=3.0, optim=adamw_torch, optim_args=None, output_dir=output/privacy_detection_pt-20240126-125436-128-2e-2, overwrite_output_dir=False, past_index=-1, per_device_eval_batch_size=8, per_device_train_batch_size=1, predict_with_generate=False, prediction_loss_only=False, push_to_hub=False, push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=, ray_scope=last, remove_unused_columns=True, report_to=[], resume_from_checkpoint=True, run_name=output/privacy_detection_pt-20240126-125436-128-2e-2, save_on_each_node=False, save_only_model=False, save_safetensors=False, save_steps=500, save_strategy=steps, save_total_limit=None, seed=42, skip_memory_metrics=True, sortish_sampler=False, split_batches=False, tf32=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, torchdynamo=None, tpu_metrics_debug=False, tpu_num_cores=None, use_cpu=False, use_ipex=False, use_legacy_prediction_loop=False, use_mps_device=False, warmup_ratio=0.0, warmup_steps=0, weight_decay=0.0, ) [INFO|configuration_utils.py:729] 2024-01-26 12:54:45,398 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--THUDM--chatglm3-6b/snapshots/37f2196f481f8989ea443be625d05f97043652ea/config.json [INFO|configuration_utils.py:729] 2024-01-26 12:54:45,957 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--THUDM--chatglm3-6b/snapshots/37f2196f481f8989ea443be625d05f97043652ea/config.json [INFO|configuration_utils.py:792] 2024-01-26 12:54:45,960 >> Model config ChatGLMConfig { "_name_or_path": "THUDM/chatglm3-6b", "add_bias_linear": false, "add_qkv_bias": true, "apply_query_key_layer_scaling": true, "apply_residual_connection_post_layernorm": false, "architectures": [ "ChatGLMModel" ], "attention_dropout": 0.0, "attention_softmax_in_fp32": true, "auto_map": { "AutoConfig": "THUDM/chatglm3-6b--configuration_chatglm.ChatGLMConfig", "AutoModel": "THUDM/chatglm3-6b--modeling_chatglm.ChatGLMForConditionalGeneration", "AutoModelForCausalLM": "THUDM/chatglm3-6b--modeling_chatglm.ChatGLMForConditionalGeneration", "AutoModelForSeq2SeqLM": "THUDM/chatglm3-6b--modeling_chatglm.ChatGLMForConditionalGeneration", "AutoModelForSequenceClassification": "THUDM/chatglm3-6b--modeling_chatglm.ChatGLMForSequenceClassification" }, "bias_dropout_fusion": true, "classifier_dropout": null, "eos_token_id": 2, "ffn_hidden_size": 13696, "fp32_residual_connection": false, "hidden_dropout": 0.0, "hidden_size": 4096, "kv_channels": 128, "layernorm_epsilon": 1e-05, "model_type": "chatglm", "multi_query_attention": true, "multi_query_group_num": 2, "num_attention_heads": 32, "num_layers": 28, "original_rope": true, "pad_token_id": 0, "padded_vocab_size": 65024, "post_layer_norm": true, "pre_seq_len": null, "prefix_projection": false, "quantization_bit": 0, "rmsnorm": true, "seq_length": 8192, "tie_word_embeddings": false, "torch_dtype": "float16", "transformers_version": "4.37.1", "use_cache": true, "vocab_size": 65024 } [INFO|tokenization_utils_base.py:2027] 2024-01-26 12:54:46,519 >> loading file tokenizer.model from cache at /root/.cache/huggingface/hub/models--THUDM--chatglm3-6b/snapshots/37f2196f481f8989ea443be625d05f97043652ea/tokenizer.model [INFO|tokenization_utils_base.py:2027] 2024-01-26 12:54:46,519 >> loading file added_tokens.json from cache at None [INFO|tokenization_utils_base.py:2027] 2024-01-26 12:54:46,519 >> loading file special_tokens_map.json from cache at None [INFO|tokenization_utils_base.py:2027] 2024-01-26 12:54:46,519 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--THUDM--chatglm3-6b/snapshots/37f2196f481f8989ea443be625d05f97043652ea/tokenizer_config.json [INFO|tokenization_utils_base.py:2027] 2024-01-26 12:54:46,519 >> loading file tokenizer.json from cache at None [INFO|modeling_utils.py:3478] 2024-01-26 12:54:47,170 >> loading weights file model.safetensors from cache at /root/.cache/huggingface/hub/models--THUDM--chatglm3-6b/snapshots/37f2196f481f8989ea443be625d05f97043652ea/model.safetensors.index.json [INFO|configuration_utils.py:826] 2024-01-26 12:54:47,177 >> Generate config GenerationConfig { "eos_token_id": 2, "pad_token_id": 0, "use_cache": false } Loading checkpoint shards: 0%| | 0/7 [00:00> All model checkpoint weights were used when initializing ChatGLMForConditionalGeneration. [WARNING|modeling_utils.py:4354] 2024-01-26 12:55:07,173 >> Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at THUDM/chatglm3-6b and are newly initialized: ['transformer.prefix_encoder.embedding.weight'] You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference. [INFO|modeling_utils.py:3897] 2024-01-26 12:55:07,458 >> Generation config file not found, using a generation config created from the model config. Sanity Check >>>>>>>>>>>>> '[gMASK]': 64790 -> -100 'sop': 64792 -> -100 '': 30910 -> -100 '请': 55073 -> -100 '找出': 40369 -> -100 '下面': 33182 -> -100 '文本': 36704 -> -100 '中的': 31697 -> -100 'position': 6523 -> -100 '：': 31211 -> -100 '艺术': 31835 -> -100 '是': 54532 -> -100 '相同的': 38815 -> -100 '，': 31123 -> -100 '音乐': 32000 -> -100 '美术': 33020 -> -100 '体育': 32214 -> -100 '三': 54645 -> -100 '样': 54741 -> -100 '都是': 31700 -> -100 '艺术': 31835 -> -100 '。，': 37843 -> -100 '三': 54645 -> -100 '样': 54741 -> -100 '艺术': 31835 -> -100 '都是': 31700 -> -100 '靠': 55518 -> -100 '感觉': 32044 -> -100 '的': 54530 -> -100 '。': 31155 -> -100 '感觉': 32044 -> -100 '好玩': 42814 -> -100 '起来': 31841 -> -100 '就很': 40030 -> -100 '轻松': 33550 -> -100 '，': 31123 -> -100 '所以': 31672 -> -100 '叫做': 35528 -> -100 '玩': 55409 -> -100 '艺术': 31835 -> -100 '。': 31155 -> -100 '没': 54721 -> -100 '感觉': 32044 -> -100 '找不到': 37779 -> -100 '北': 54760 -> -100 '的': 54530 -> -100 '干脆': 43396 -> -100 '别': 54835 -> -100 '玩': 55409 -> -100 '了': 54537 -> -100 '！': 31404 -> -100 '，': 31123 -> -100 '香港': 31776 -> -100 '电影': 31867 -> -100 '国语': 54385 -> -100 '配音': 40392 -> -100 '名家': 40465 -> -100 '周': 54896 -> -100 '思': 54872 -> -100 '平': 54678 -> -100 '，': 31123 -> -100 '代表作': 43527 -> -100 '有': 54536 -> -100 'TVB': 42671 -> -100 '《': 54611 -> -100 '上海': 31770 -> -100 '滩': 56928 -> -100 '》': 54612 -> -100 '周': 54896 -> -100 '润': 55826 -> -100 '发': 54559 -> -100 '等': 54609 -> -100 '香港': 37944 -> 37944 '电影': 31867 -> 31867 '国语': 54385 -> 54385 '配音': 40392 -> 40392 '名家': 40465 -> 40465 '': 2 -> 2 <<<<<<<<<<<<< Sanity Check 01/26/2024 12:55:08 - WARNING - accelerate.utils.other - Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. [INFO|trainer.py:522] 2024-01-26 12:55:20,019 >> max_steps is given, it will override any value given in num_train_epochs [WARNING|modeling_utils.py:2134] 2024-01-26 12:55:20,020 >> You are using an old version of the checkpointing format that is deprecated (We will also silently ignore `gradient_checkpointing_kwargs` in case you passed it).Please update to the new format on your modeling file. To use the new format, you need to completely remove the definition of the method `_set_gradient_checkpointing` in your model. [INFO|trainer.py:1721] 2024-01-26 12:55:21,544 >> ***** Running training ***** [INFO|trainer.py:1722] 2024-01-26 12:55:21,544 >> Num examples = 2,515 [INFO|trainer.py:1723] 2024-01-26 12:55:21,544 >> Num Epochs = 2 [INFO|trainer.py:1724] 2024-01-26 12:55:21,544 >> Instantaneous batch size per device = 1 [INFO|trainer.py:1727] 2024-01-26 12:55:21,544 >> Total train batch size (w. parallel, distributed & accumulation) = 32 [INFO|trainer.py:1728] 2024-01-26 12:55:21,544 >> Gradient Accumulation steps = 32 [INFO|trainer.py:1729] 2024-01-26 12:55:21,544 >> Total optimization steps = 100 [INFO|trainer.py:1730] 2024-01-26 12:55:21,545 >> Number of trainable parameters = 1,835,008 0%| | 0/100 [00:00> Training completed. Do not forget to share your model on huggingface.co/models =) {'train_runtime': 1218.4689, 'train_samples_per_second': 2.626, 'train_steps_per_second': 0.082, 'train_loss': 0.7395605874061585, 'epoch': 1.27} 100%|██████████| 100/100 [20:18<00:00, 12.19s/it] 100%|██████████| 100/100 [20:18<00:00, 12.18s/it] Saving PrefixEncoder [INFO|configuration_utils.py:473] 2024-01-26 13:15:40,038 >> Configuration saved in output/privacy_detection_pt-20240126-125436-128-2e-2/config.json [INFO|configuration_utils.py:594] 2024-01-26 13:15:40,039 >> Configuration saved in output/privacy_detection_pt-20240126-125436-128-2e-2/generation_config.json [INFO|modeling_utils.py:2495] 2024-01-26 13:15:40,068 >> Model weights saved in output/privacy_detection_pt-20240126-125436-128-2e-2/pytorch_model.bin [INFO|tokenization_utils_base.py:2433] 2024-01-26 13:15:40,069 >> tokenizer config file saved in output/privacy_detection_pt-20240126-125436-128-2e-2/tokenizer_config.json [INFO|tokenization_utils_base.py:2442] 2024-01-26 13:15:40,069 >> Special tokens file saved in output/privacy_detection_pt-20240126-125436-128-2e-2/special_tokens_map.json