| + deepspeed --master_port 59730 --module safe_rlhf.finetune --train_datasets inverse-json::/home/hansirui_1st/jiayi/resist/imdb_data/train/pos/1000/train.json --model_name_or_path /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T --max_length 512 --trust_remote_code True --epochs 1 --per_device_train_batch_size 1 --per_device_eval_batch_size 4 --gradient_accumulation_steps 8 --gradient_checkpointing --learning_rate 1e-5 --lr_warmup_ratio 0 --weight_decay 0.0 --lr_scheduler_type constant --weight_decay 0.0 --seed 42 --output_dir /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-3T/tinyllama-3T-s3-Q1-1000 --log_type wandb --log_run_name imdb-tinyllama-3T-s3-Q1-1000 --log_project Inverse_Alignment_IMDb --zero_stage 3 --offload none --bf16 True --tf32 True --save_16bit |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| nvcc warning : incompatible redefinition for option |
| [rank0]:[W527 19:55:02.022575148 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 0] using 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, or call init_process_group() with a device_id. |
| [rank4]:[W527 19:55:02.029510503 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 4] using GPU 4 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, or call init_process_group() with a device_id. |
| [rank6]:[W527 19:55:02.045668779 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 6] using GPU 6 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, or call init_process_group() with a device_id. |
| [rank3]:[W527 19:55:02.045879149 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 3] using GPU 3 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, or call init_process_group() with a device_id. |
| [rank2]:[W527 19:55:02.046063147 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 2] using GPU 2 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, or call init_process_group() with a device_id. |
| [rank1]:[W527 19:55:02.046199068 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 1] using 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, or call init_process_group() with a device_id. |
| [rank5]:[W527 19:55:02.046305547 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 5] using GPU 5 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, or call init_process_group() with a device_id. |
| [rank7]:[W527 19:55:02.094626666 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 7] using GPU 7 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, or call init_process_group() with a device_id. |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/config.json |
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| Model config LlamaConfig { |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 5632, |
| "max_position_embeddings": 2048, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 32, |
| "num_hidden_layers": 22, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "float32", |
| "transformers_version": "4.52.1", |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/model.safetensors |
| Will use torch_dtype=torch.float32 as defined in model |
| Will use torch_dtype=torch.float32 as defined in model |
| Will use torch_dtype=torch.float32 as defined in model |
| Will use torch_dtype=torch.float32 as defined in model |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Will use torch_dtype=torch.float32 as defined in model |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Will use torch_dtype=torch.float32 as defined in model |
| Will use torch_dtype=torch.float32 as defined in model |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Will use torch_dtype=torch.float32 as defined in model |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Instantiating LlamaForCausalLM model under default dtype torch.float32. |
| Detected DeepSpeed ZeRO-3: activating zero.init() for this model |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2 |
| } |
|
|
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| loading file tokenizer.model |
| loading file tokenizer.model |
| loading file tokenizer.model |
| loading file tokenizer.model |
| loading file tokenizer.model |
| loading file tokenizer.json |
| loading file tokenizer.json |
| loading file tokenizer.model |
| loading file tokenizer.json |
| loading file tokenizer.json |
| loading file tokenizer.json |
| loading file added_tokens.json |
| loading file added_tokens.json |
| loading file added_tokens.json |
| loading file added_tokens.json |
| loading file special_tokens_map.json |
| loading file special_tokens_map.json |
| loading file special_tokens_map.json |
| loading file added_tokens.json |
| loading file special_tokens_map.json |
| loading file tokenizer.json |
| loading file tokenizer_config.json |
| loading file tokenizer_config.json |
| loading file tokenizer_config.json |
| loading file tokenizer_config.json |
| loading file chat_template.jinja |
| loading file chat_template.jinja |
| loading file added_tokens.json |
| loading file special_tokens_map.json |
| loading file chat_template.jinja |
| loading file chat_template.jinja |
| loading file special_tokens_map.json |
| loading file tokenizer_config.json |
| loading file tokenizer.model |
| loading file tokenizer_config.json |
| loading file chat_template.jinja |
| loading file chat_template.jinja |
| loading file tokenizer.json |
| loading file added_tokens.json |
| loading file special_tokens_map.json |
| loading file tokenizer_config.json |
| loading file chat_template.jinja |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| All model checkpoint weights were used when initializing LlamaForCausalLM. |
|
|
| All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T. |
| If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training. |
| loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-1431k-3T/generation_config.json |
| Generate config GenerationConfig { |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "max_length": 2048, |
| "pad_token_id": 0 |
| } |
|
|
| loading file tokenizer.model |
| loading file tokenizer.json |
| loading file added_tokens.json |
| loading file special_tokens_map.json |
| loading file tokenizer_config.json |
| loading file chat_template.jinja |
| You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new embeddings will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings |
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
|
|
|
|
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
| Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root... |
| Detected CUDA files, patching ldflags |
| Emitting ninja build file /home/hansirui_1st/.cache/torch_extensions/py311_cu124/fused_adam/build.ninja... |
| /aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/torch/utils/cpp_extension.py:2059: UserWarning: TORCH_CUDA_ARCH_LIST is not set, all archs for visible cards are included for compilation. |
| If this is not desired, please set os.environ[ |
| warnings.warn( |
| Building extension module fused_adam... |
| Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) |
| Loading extension module fused_adam... |
| Loading extension module fused_adam...Loading extension module fused_adam...Loading extension module fused_adam...Loading extension module fused_adam... |
|
|
|
|
|
|
| Loading extension module fused_adam... |
| Loading extension module fused_adam... |
| Loading extension module fused_adam... |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
| wandb: Currently logged in as: xtom to https://api.wandb.ai. Use `wandb login --relogin` to force relogin |
| wandb: Tracking run with wandb version 0.19.11 |
| wandb: Run data is saved locally in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-3T/tinyllama-3T-s3-Q1-1000/wandb/run-20250527_195528-z1m1qlsj |
| wandb: Run `wandb offline` to turn off syncing. |
| wandb: Syncing run imdb-tinyllama-3T-s3-Q1-1000 |
| wandb: ⭐️ View project at https://wandb.ai/xtom/Inverse_Alignment_IMDb |
| wandb: 🚀 View run at https://wandb.ai/xtom/Inverse_Alignment_IMDb/runs/z1m1qlsj |
|
Training 1/1 epoch: 0%| | 0/125 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`. |
|
Training 1/1 epoch (loss 2.8450): 0%| | 0/125 [00:09<?, ?it/s]
Training 1/1 epoch (loss 2.8450): 1%| | 1/125 [00:09<19:40, 9.52s/it]
Training 1/1 epoch (loss 2.6666): 1%| | 1/125 [00:11<19:40, 9.52s/it]
Training 1/1 epoch (loss 2.6666): 2%|▏ | 2/125 [00:11<10:42, 5.22s/it]
Training 1/1 epoch (loss 2.7568): 2%|▏ | 2/125 [00:13<10:42, 5.22s/it]
Training 1/1 epoch (loss 2.7568): 2%|▏ | 3/125 [00:13<07:16, 3.57s/it]
Training 1/1 epoch (loss 2.7999): 2%|▏ | 3/125 [00:15<07:16, 3.57s/it]
Training 1/1 epoch (loss 2.7999): 3%|▎ | 4/125 [00:15<05:48, 2.88s/it]
Training 1/1 epoch (loss 2.6586): 3%|▎ | 4/125 [00:15<05:48, 2.88s/it]
Training 1/1 epoch (loss 2.6586): 4%|▍ | 5/125 [00:15<04:08, 2.07s/it]
Training 1/1 epoch (loss 2.8452): 4%|▍ | 5/125 [00:17<04:08, 2.07s/it]
Training 1/1 epoch (loss 2.8452): 5%|▍ | 6/125 [00:17<04:01, 2.03s/it]
Training 1/1 epoch (loss 2.6166): 5%|▍ | 6/125 [00:18<04:01, 2.03s/it]
Training 1/1 epoch (loss 2.6166): 6%|▌ | 7/125 [00:18<03:21, 1.70s/it]
Training 1/1 epoch (loss 2.8243): 6%|▌ | 7/125 [00:20<03:21, 1.70s/it]
Training 1/1 epoch (loss 2.8243): 6%|▋ | 8/125 [00:20<03:16, 1.68s/it]
Training 1/1 epoch (loss 3.0019): 6%|▋ | 8/125 [00:21<03:16, 1.68s/it]
Training 1/1 epoch (loss 3.0019): 7%|▋ | 9/125 [00:21<03:03, 1.58s/it]
Training 1/1 epoch (loss 2.8236): 7%|▋ | 9/125 [00:22<03:03, 1.58s/it]
Training 1/1 epoch (loss 2.8236): 8%|▊ | 10/125 [00:22<02:25, 1.27s/it]
Training 1/1 epoch (loss 2.6169): 8%|▊ | 10/125 [00:23<02:25, 1.27s/it]
Training 1/1 epoch (loss 2.6169): 9%|▉ | 11/125 [00:23<02:30, 1.32s/it]
Training 1/1 epoch (loss 2.6499): 9%|▉ | 11/125 [00:25<02:30, 1.32s/it]
Training 1/1 epoch (loss 2.6499): 10%|▉ | 12/125 [00:25<02:42, 1.44s/it]
Training 1/1 epoch (loss 2.7414): 10%|▉ | 12/125 [00:25<02:42, 1.44s/it]
Training 1/1 epoch (loss 2.7414): 10%|█ | 13/125 [00:25<02:09, 1.15s/it]
Training 1/1 epoch (loss 2.6963): 10%|█ | 13/125 [00:28<02:09, 1.15s/it]
Training 1/1 epoch (loss 2.6963): 11%|█ | 14/125 [00:28<02:51, 1.54s/it]
Training 1/1 epoch (loss 2.5591): 11%|█ | 14/125 [00:29<02:51, 1.54s/it]
Training 1/1 epoch (loss 2.5591): 12%|█▏ | 15/125 [00:29<02:47, 1.53s/it]
Training 1/1 epoch (loss 2.8737): 12%|█▏ | 15/125 [00:30<02:47, 1.53s/it]
Training 1/1 epoch (loss 2.8737): 13%|█▎ | 16/125 [00:30<02:29, 1.37s/it]
Training 1/1 epoch (loss 2.7383): 13%|█▎ | 16/125 [00:32<02:29, 1.37s/it]
Training 1/1 epoch (loss 2.7383): 14%|█▎ | 17/125 [00:32<02:33, 1.42s/it]
Training 1/1 epoch (loss 2.6812): 14%|█▎ | 17/125 [00:33<02:33, 1.42s/it]
Training 1/1 epoch (loss 2.6812): 14%|█▍ | 18/125 [00:33<02:31, 1.41s/it]
Training 1/1 epoch (loss 2.6876): 14%|█▍ | 18/125 [00:35<02:31, 1.41s/it]
Training 1/1 epoch (loss 2.6876): 15%|█▌ | 19/125 [00:35<02:38, 1.49s/it]
Training 1/1 epoch (loss 2.7643): 15%|█▌ | 19/125 [00:36<02:38, 1.49s/it]
Training 1/1 epoch (loss 2.7643): 16%|█▌ | 20/125 [00:36<02:28, 1.42s/it]
Training 1/1 epoch (loss 2.6707): 16%|█▌ | 20/125 [00:37<02:28, 1.42s/it]
Training 1/1 epoch (loss 2.6707): 17%|█▋ | 21/125 [00:37<01:57, 1.13s/it]
Training 1/1 epoch (loss 2.7734): 17%|█▋ | 21/125 [00:39<01:57, 1.13s/it]
Training 1/1 epoch (loss 2.7734): 18%|█▊ | 22/125 [00:39<02:21, 1.37s/it]
Training 1/1 epoch (loss 2.3809): 18%|█▊ | 22/125 [00:40<02:21, 1.37s/it]
Training 1/1 epoch (loss 2.3809): 18%|█▊ | 23/125 [00:40<02:17, 1.35s/it]
Training 1/1 epoch (loss 2.6845): 18%|█▊ | 23/125 [00:41<02:17, 1.35s/it]
Training 1/1 epoch (loss 2.6845): 19%|█▉ | 24/125 [00:41<02:02, 1.22s/it]
Training 1/1 epoch (loss 2.3017): 19%|█▉ | 24/125 [00:43<02:02, 1.22s/it]
Training 1/1 epoch (loss 2.3017): 20%|██ | 25/125 [00:43<02:14, 1.34s/it]
Training 1/1 epoch (loss 2.6367): 20%|██ | 25/125 [00:45<02:14, 1.34s/it]
Training 1/1 epoch (loss 2.6367): 21%|██ | 26/125 [00:45<02:31, 1.53s/it]
Training 1/1 epoch (loss 2.5659): 21%|██ | 26/125 [00:45<02:31, 1.53s/it]
Training 1/1 epoch (loss 2.5659): 22%|██▏ | 27/125 [00:45<02:07, 1.30s/it]
Training 1/1 epoch (loss 2.3844): 22%|██▏ | 27/125 [00:47<02:07, 1.30s/it]
Training 1/1 epoch (loss 2.3844): 22%|██▏ | 28/125 [00:47<02:08, 1.33s/it]
Training 1/1 epoch (loss 2.7085): 22%|██▏ | 28/125 [00:48<02:08, 1.33s/it]
Training 1/1 epoch (loss 2.7085): 23%|██▎ | 29/125 [00:48<02:17, 1.43s/it]
Training 1/1 epoch (loss 2.9094): 23%|██▎ | 29/125 [00:49<02:17, 1.43s/it]
Training 1/1 epoch (loss 2.9094): 24%|██▍ | 30/125 [00:49<01:58, 1.25s/it]
Training 1/1 epoch (loss 2.5890): 24%|██▍ | 30/125 [00:51<01:58, 1.25s/it]
Training 1/1 epoch (loss 2.5890): 25%|██▍ | 31/125 [00:51<02:17, 1.46s/it]
Training 1/1 epoch (loss 2.6660): 25%|██▍ | 31/125 [00:52<02:17, 1.46s/it]
Training 1/1 epoch (loss 2.6660): 26%|██▌ | 32/125 [00:52<02:00, 1.29s/it]
Training 1/1 epoch (loss 2.5878): 26%|██▌ | 32/125 [00:54<02:00, 1.29s/it]
Training 1/1 epoch (loss 2.5878): 26%|██▋ | 33/125 [00:54<02:26, 1.59s/it]
Training 1/1 epoch (loss 2.6732): 26%|██▋ | 33/125 [00:56<02:26, 1.59s/it]
Training 1/1 epoch (loss 2.6732): 27%|██▋ | 34/125 [00:56<02:32, 1.67s/it]
Training 1/1 epoch (loss 2.7195): 27%|██▋ | 34/125 [00:57<02:32, 1.67s/it]
Training 1/1 epoch (loss 2.7195): 28%|██▊ | 35/125 [00:57<02:00, 1.34s/it]
Training 1/1 epoch (loss 2.5712): 28%|██▊ | 35/125 [00:58<02:00, 1.34s/it]
Training 1/1 epoch (loss 2.5712): 29%|██▉ | 36/125 [00:58<01:57, 1.32s/it]
Training 1/1 epoch (loss 2.6352): 29%|██▉ | 36/125 [00:59<01:57, 1.32s/it]
Training 1/1 epoch (loss 2.6352): 30%|██▉ | 37/125 [00:59<01:59, 1.36s/it]
Training 1/1 epoch (loss 2.6017): 30%|██▉ | 37/125 [01:00<01:59, 1.36s/it]
Training 1/1 epoch (loss 2.6017): 30%|███ | 38/125 [01:00<01:34, 1.09s/it]
Training 1/1 epoch (loss 2.7384): 30%|███ | 38/125 [01:02<01:34, 1.09s/it]
Training 1/1 epoch (loss 2.7384): 31%|███ | 39/125 [01:02<01:56, 1.35s/it]
Training 1/1 epoch (loss 2.5369): 31%|███ | 39/125 [01:03<01:56, 1.35s/it]
Training 1/1 epoch (loss 2.5369): 32%|███▏ | 40/125 [01:03<01:59, 1.41s/it]
Training 1/1 epoch (loss 2.6261): 32%|███▏ | 40/125 [01:04<01:59, 1.41s/it]
Training 1/1 epoch (loss 2.6261): 33%|███▎ | 41/125 [01:04<01:41, 1.21s/it]
Training 1/1 epoch (loss 2.6124): 33%|███▎ | 41/125 [01:06<01:41, 1.21s/it]
Training 1/1 epoch (loss 2.6124): 34%|███▎ | 42/125 [01:06<01:58, 1.42s/it]
Training 1/1 epoch (loss 2.6091): 34%|███▎ | 42/125 [01:07<01:58, 1.42s/it]
Training 1/1 epoch (loss 2.6091): 34%|███▍ | 43/125 [01:07<01:49, 1.34s/it]
Training 1/1 epoch (loss 2.6810): 34%|███▍ | 43/125 [01:08<01:49, 1.34s/it]
Training 1/1 epoch (loss 2.6810): 35%|███▌ | 44/125 [01:08<01:44, 1.29s/it]
Training 1/1 epoch (loss 2.4745): 35%|███▌ | 44/125 [01:10<01:44, 1.29s/it]
Training 1/1 epoch (loss 2.4745): 36%|███▌ | 45/125 [01:10<02:00, 1.50s/it]
Training 1/1 epoch (loss 2.5688): 36%|███▌ | 45/125 [01:11<02:00, 1.50s/it]
Training 1/1 epoch (loss 2.5688): 37%|███▋ | 46/125 [01:11<01:46, 1.35s/it]
Training 1/1 epoch (loss 2.7272): 37%|███▋ | 46/125 [01:13<01:46, 1.35s/it]
Training 1/1 epoch (loss 2.7272): 38%|███▊ | 47/125 [01:13<01:45, 1.35s/it]
Training 1/1 epoch (loss 2.7701): 38%|███▊ | 47/125 [01:14<01:45, 1.35s/it]
Training 1/1 epoch (loss 2.7701): 38%|███▊ | 48/125 [01:14<01:51, 1.45s/it]
Training 1/1 epoch (loss 2.6407): 38%|███▊ | 48/125 [01:15<01:51, 1.45s/it]
Training 1/1 epoch (loss 2.6407): 39%|███▉ | 49/125 [01:15<01:33, 1.24s/it]
Training 1/1 epoch (loss 2.6324): 39%|███▉ | 49/125 [01:16<01:33, 1.24s/it]
Training 1/1 epoch (loss 2.6324): 40%|████ | 50/125 [01:16<01:29, 1.19s/it]
Training 1/1 epoch (loss 2.4201): 40%|████ | 50/125 [01:17<01:29, 1.19s/it]
Training 1/1 epoch (loss 2.4201): 41%|████ | 51/125 [01:17<01:29, 1.21s/it]
Training 1/1 epoch (loss 2.6686): 41%|████ | 51/125 [01:18<01:29, 1.21s/it]
Training 1/1 epoch (loss 2.6686): 42%|████▏ | 52/125 [01:18<01:14, 1.03s/it]
Training 1/1 epoch (loss 2.5764): 42%|████▏ | 52/125 [01:19<01:14, 1.03s/it]
Training 1/1 epoch (loss 2.5764): 42%|████▏ | 53/125 [01:19<01:14, 1.04s/it]
Training 1/1 epoch (loss 2.6735): 42%|████▏ | 53/125 [01:21<01:14, 1.04s/it]
Training 1/1 epoch (loss 2.6735): 43%|████▎ | 54/125 [01:21<01:39, 1.41s/it]
Training 1/1 epoch (loss 2.5607): 43%|████▎ | 54/125 [01:22<01:39, 1.41s/it]
Training 1/1 epoch (loss 2.5607): 44%|████▍ | 55/125 [01:22<01:18, 1.12s/it]
Training 1/1 epoch (loss 2.5851): 44%|████▍ | 55/125 [01:24<01:18, 1.12s/it]
Training 1/1 epoch (loss 2.5851): 45%|████▍ | 56/125 [01:24<01:45, 1.53s/it]
Training 1/1 epoch (loss 2.6456): 45%|████▍ | 56/125 [01:26<01:45, 1.53s/it]
Training 1/1 epoch (loss 2.6456): 46%|████▌ | 57/125 [01:26<01:53, 1.67s/it]
Training 1/1 epoch (loss 2.6544): 46%|████▌ | 57/125 [01:28<01:53, 1.67s/it]
Training 1/1 epoch (loss 2.6544): 46%|████▋ | 58/125 [01:28<01:46, 1.59s/it]
Training 1/1 epoch (loss 2.5826): 46%|████▋ | 58/125 [01:30<01:46, 1.59s/it]
Training 1/1 epoch (loss 2.5826): 47%|████▋ | 59/125 [01:30<02:01, 1.85s/it]
Training 1/1 epoch (loss 2.7206): 47%|████▋ | 59/125 [01:32<02:01, 1.85s/it]
Training 1/1 epoch (loss 2.7206): 48%|████▊ | 60/125 [01:32<01:49, 1.69s/it]
Training 1/1 epoch (loss 2.4390): 48%|████▊ | 60/125 [01:34<01:49, 1.69s/it]
Training 1/1 epoch (loss 2.4390): 49%|████▉ | 61/125 [01:34<02:00, 1.88s/it]
Training 1/1 epoch (loss 2.7160): 49%|████▉ | 61/125 [01:36<02:00, 1.88s/it]
Training 1/1 epoch (loss 2.7160): 50%|████▉ | 62/125 [01:36<02:00, 1.92s/it]
Training 1/1 epoch (loss 2.6458): 50%|████▉ | 62/125 [01:36<02:00, 1.92s/it]
Training 1/1 epoch (loss 2.6458): 50%|█████ | 63/125 [01:36<01:30, 1.47s/it]
Training 1/1 epoch (loss 2.6238): 50%|█████ | 63/125 [01:39<01:30, 1.47s/it]
Training 1/1 epoch (loss 2.6238): 51%|█████ | 64/125 [01:39<01:45, 1.72s/it]
Training 1/1 epoch (loss 2.6915): 51%|█████ | 64/125 [01:40<01:45, 1.72s/it]
Training 1/1 epoch (loss 2.6915): 52%|█████▏ | 65/125 [01:40<01:45, 1.77s/it]
Training 1/1 epoch (loss 2.5598): 52%|█████▏ | 65/125 [01:41<01:45, 1.77s/it]
Training 1/1 epoch (loss 2.5598): 53%|█████▎ | 66/125 [01:41<01:28, 1.49s/it]
Training 1/1 epoch (loss 2.6082): 53%|█████▎ | 66/125 [01:43<01:28, 1.49s/it]
Training 1/1 epoch (loss 2.6082): 54%|█████▎ | 67/125 [01:43<01:24, 1.45s/it]
Training 1/1 epoch (loss 2.6914): 54%|█████▎ | 67/125 [01:44<01:24, 1.45s/it]
Training 1/1 epoch (loss 2.6914): 54%|█████▍ | 68/125 [01:44<01:13, 1.29s/it]
Training 1/1 epoch (loss 2.7666): 54%|█████▍ | 68/125 [01:45<01:13, 1.29s/it]
Training 1/1 epoch (loss 2.7666): 55%|█████▌ | 69/125 [01:45<01:22, 1.47s/it]
Training 1/1 epoch (loss 2.4771): 55%|█████▌ | 69/125 [01:47<01:22, 1.47s/it]
Training 1/1 epoch (loss 2.4771): 56%|█████▌ | 70/125 [01:47<01:25, 1.56s/it]
Training 1/1 epoch (loss 2.6558): 56%|█████▌ | 70/125 [01:48<01:25, 1.56s/it]
Training 1/1 epoch (loss 2.6558): 57%|█████▋ | 71/125 [01:48<01:11, 1.32s/it]
Training 1/1 epoch (loss 2.5620): 57%|█████▋ | 71/125 [01:51<01:11, 1.32s/it]
Training 1/1 epoch (loss 2.5620): 58%|█████▊ | 72/125 [01:51<01:29, 1.69s/it]
Training 1/1 epoch (loss 2.5650): 58%|█████▊ | 72/125 [01:53<01:29, 1.69s/it]
Training 1/1 epoch (loss 2.5650): 58%|█████▊ | 73/125 [01:53<01:35, 1.84s/it]
Training 1/1 epoch (loss 2.4194): 58%|█████▊ | 73/125 [01:54<01:35, 1.84s/it]
Training 1/1 epoch (loss 2.4194): 59%|█████▉ | 74/125 [01:54<01:26, 1.69s/it]
Training 1/1 epoch (loss 2.6248): 59%|█████▉ | 74/125 [01:57<01:26, 1.69s/it]
Training 1/1 epoch (loss 2.6248): 60%|██████ | 75/125 [01:57<01:35, 1.92s/it]
Training 1/1 epoch (loss 2.4859): 60%|██████ | 75/125 [01:58<01:35, 1.92s/it]
Training 1/1 epoch (loss 2.4859): 61%|██████ | 76/125 [01:58<01:29, 1.82s/it]
Training 1/1 epoch (loss 2.5208): 61%|██████ | 76/125 [01:59<01:29, 1.82s/it]
Training 1/1 epoch (loss 2.5208): 62%|██████▏ | 77/125 [01:59<01:17, 1.61s/it]
Training 1/1 epoch (loss 2.5306): 62%|██████▏ | 77/125 [02:01<01:17, 1.61s/it]
Training 1/1 epoch (loss 2.5306): 62%|██████▏ | 78/125 [02:01<01:22, 1.77s/it]
Training 1/1 epoch (loss 2.6660): 62%|██████▏ | 78/125 [02:02<01:22, 1.77s/it]
Training 1/1 epoch (loss 2.6660): 63%|██████▎ | 79/125 [02:02<01:03, 1.39s/it]
Training 1/1 epoch (loss 2.5823): 63%|██████▎ | 79/125 [02:04<01:03, 1.39s/it]
Training 1/1 epoch (loss 2.5823): 64%|██████▍ | 80/125 [02:04<01:08, 1.52s/it]
Training 1/1 epoch (loss 2.7474): 64%|██████▍ | 80/125 [02:05<01:08, 1.52s/it]
Training 1/1 epoch (loss 2.7474): 65%|██████▍ | 81/125 [02:05<01:09, 1.59s/it]
Training 1/1 epoch (loss 2.4930): 65%|██████▍ | 81/125 [02:07<01:09, 1.59s/it]
Training 1/1 epoch (loss 2.4930): 66%|██████▌ | 82/125 [02:07<01:05, 1.53s/it]
Training 1/1 epoch (loss 2.7988): 66%|██████▌ | 82/125 [02:08<01:05, 1.53s/it]
Training 1/1 epoch (loss 2.7988): 66%|██████▋ | 83/125 [02:08<00:59, 1.43s/it]
Training 1/1 epoch (loss 2.6061): 66%|██████▋ | 83/125 [02:09<00:59, 1.43s/it]
Training 1/1 epoch (loss 2.6061): 67%|██████▋ | 84/125 [02:09<00:51, 1.25s/it]
Training 1/1 epoch (loss 2.7646): 67%|██████▋ | 84/125 [02:11<00:51, 1.25s/it]
Training 1/1 epoch (loss 2.7646): 68%|██████▊ | 85/125 [02:11<00:54, 1.37s/it]
Training 1/1 epoch (loss 2.6303): 68%|██████▊ | 85/125 [02:12<00:54, 1.37s/it]
Training 1/1 epoch (loss 2.6303): 69%|██████▉ | 86/125 [02:12<00:53, 1.38s/it]
Training 1/1 epoch (loss 2.6916): 69%|██████▉ | 86/125 [02:12<00:53, 1.38s/it]
Training 1/1 epoch (loss 2.6916): 70%|██████▉ | 87/125 [02:12<00:42, 1.11s/it]
Training 1/1 epoch (loss 2.5079): 70%|██████▉ | 87/125 [02:14<00:42, 1.11s/it]
Training 1/1 epoch (loss 2.5079): 70%|███████ | 88/125 [02:14<00:50, 1.36s/it]
Training 1/1 epoch (loss 2.5251): 70%|███████ | 88/125 [02:16<00:50, 1.36s/it]
Training 1/1 epoch (loss 2.5251): 71%|███████ | 89/125 [02:16<00:53, 1.50s/it]
Training 1/1 epoch (loss 2.5878): 71%|███████ | 89/125 [02:17<00:53, 1.50s/it]
Training 1/1 epoch (loss 2.5878): 72%|███████▏ | 90/125 [02:17<00:49, 1.43s/it]
Training 1/1 epoch (loss 2.8245): 72%|███████▏ | 90/125 [02:20<00:49, 1.43s/it]
Training 1/1 epoch (loss 2.8245): 73%|███████▎ | 91/125 [02:20<00:59, 1.74s/it]
Training 1/1 epoch (loss 2.3536): 73%|███████▎ | 91/125 [02:22<00:59, 1.74s/it]
Training 1/1 epoch (loss 2.3536): 74%|███████▎ | 92/125 [02:22<00:57, 1.73s/it]
Training 1/1 epoch (loss 2.5098): 74%|███████▎ | 92/125 [02:22<00:57, 1.73s/it]
Training 1/1 epoch (loss 2.5098): 74%|███████▍ | 93/125 [02:22<00:46, 1.45s/it]
Training 1/1 epoch (loss 2.4870): 74%|███████▍ | 93/125 [02:25<00:46, 1.45s/it]
Training 1/1 epoch (loss 2.4870): 75%|███████▌ | 94/125 [02:25<00:51, 1.68s/it]
Training 1/1 epoch (loss 2.4971): 75%|███████▌ | 94/125 [02:26<00:51, 1.68s/it]
Training 1/1 epoch (loss 2.4971): 76%|███████▌ | 95/125 [02:26<00:47, 1.58s/it]
Training 1/1 epoch (loss 2.6515): 76%|███████▌ | 95/125 [02:28<00:47, 1.58s/it]
Training 1/1 epoch (loss 2.6515): 77%|███████▋ | 96/125 [02:28<00:48, 1.68s/it]
Training 1/1 epoch (loss 2.5271): 77%|███████▋ | 96/125 [02:30<00:48, 1.68s/it]
Training 1/1 epoch (loss 2.5271): 78%|███████▊ | 97/125 [02:30<00:51, 1.83s/it]
Training 1/1 epoch (loss 2.7909): 78%|███████▊ | 97/125 [02:31<00:51, 1.83s/it]
Training 1/1 epoch (loss 2.7909): 78%|███████▊ | 98/125 [02:31<00:41, 1.54s/it]
Training 1/1 epoch (loss 2.6217): 78%|███████▊ | 98/125 [02:33<00:41, 1.54s/it]
Training 1/1 epoch (loss 2.6217): 79%|███████▉ | 99/125 [02:33<00:43, 1.67s/it]
Training 1/1 epoch (loss 2.5330): 79%|███████▉ | 99/125 [02:34<00:43, 1.67s/it]
Training 1/1 epoch (loss 2.5330): 80%|████████ | 100/125 [02:34<00:36, 1.44s/it]
Training 1/1 epoch (loss 2.6377): 80%|████████ | 100/125 [02:35<00:36, 1.44s/it]
Training 1/1 epoch (loss 2.6377): 81%|████████ | 101/125 [02:35<00:32, 1.34s/it]
Training 1/1 epoch (loss 2.4057): 81%|████████ | 101/125 [02:37<00:32, 1.34s/it]
Training 1/1 epoch (loss 2.4057): 82%|████████▏ | 102/125 [02:37<00:33, 1.45s/it]
Training 1/1 epoch (loss 2.5328): 82%|████████▏ | 102/125 [02:37<00:33, 1.45s/it]
Training 1/1 epoch (loss 2.5328): 82%|████████▏ | 103/125 [02:37<00:26, 1.18s/it]
Training 1/1 epoch (loss 2.5419): 82%|████████▏ | 103/125 [02:39<00:26, 1.18s/it]
Training 1/1 epoch (loss 2.5419): 83%|████████▎ | 104/125 [02:39<00:27, 1.30s/it]
Training 1/1 epoch (loss 2.6396): 83%|████████▎ | 104/125 [02:40<00:27, 1.30s/it]
Training 1/1 epoch (loss 2.6396): 84%|████████▍ | 105/125 [02:40<00:26, 1.33s/it]
Training 1/1 epoch (loss 2.5299): 84%|████████▍ | 105/125 [02:42<00:26, 1.33s/it]
Training 1/1 epoch (loss 2.5299): 85%|████████▍ | 106/125 [02:42<00:25, 1.36s/it]
Training 1/1 epoch (loss 2.7221): 85%|████████▍ | 106/125 [02:44<00:25, 1.36s/it]
Training 1/1 epoch (loss 2.7221): 86%|████████▌ | 107/125 [02:44<00:29, 1.66s/it]
Training 1/1 epoch (loss 2.6305): 86%|████████▌ | 107/125 [02:45<00:29, 1.66s/it]
Training 1/1 epoch (loss 2.6305): 86%|████████▋ | 108/125 [02:45<00:26, 1.53s/it]
Training 1/1 epoch (loss 2.4943): 86%|████████▋ | 108/125 [02:47<00:26, 1.53s/it]
Training 1/1 epoch (loss 2.4943): 87%|████████▋ | 109/125 [02:47<00:24, 1.53s/it]
Training 1/1 epoch (loss 2.5377): 87%|████████▋ | 109/125 [02:49<00:24, 1.53s/it]
Training 1/1 epoch (loss 2.5377): 88%|████████▊ | 110/125 [02:49<00:24, 1.62s/it]
Training 1/1 epoch (loss 2.5779): 88%|████████▊ | 110/125 [02:49<00:24, 1.62s/it]
Training 1/1 epoch (loss 2.5779): 89%|████████▉ | 111/125 [02:49<00:17, 1.28s/it]
Training 1/1 epoch (loss 2.4752): 89%|████████▉ | 111/125 [02:51<00:17, 1.28s/it]
Training 1/1 epoch (loss 2.4752): 90%|████████▉ | 112/125 [02:51<00:18, 1.40s/it]
Training 1/1 epoch (loss 2.4622): 90%|████████▉ | 112/125 [02:52<00:18, 1.40s/it]
Training 1/1 epoch (loss 2.4622): 90%|█████████ | 113/125 [02:52<00:17, 1.50s/it]
Training 1/1 epoch (loss 2.4666): 90%|█████████ | 113/125 [02:53<00:17, 1.50s/it]
Training 1/1 epoch (loss 2.4666): 91%|█████████ | 114/125 [02:53<00:13, 1.24s/it]
Training 1/1 epoch (loss 2.7229): 91%|█████████ | 114/125 [02:56<00:13, 1.24s/it]
Training 1/1 epoch (loss 2.7229): 92%|█████████▏| 115/125 [02:56<00:16, 1.60s/it]
Training 1/1 epoch (loss 2.6925): 92%|█████████▏| 115/125 [02:57<00:16, 1.60s/it]
Training 1/1 epoch (loss 2.6925): 93%|█████████▎| 116/125 [02:57<00:14, 1.63s/it]
Training 1/1 epoch (loss 2.6347): 93%|█████████▎| 116/125 [02:58<00:14, 1.63s/it]
Training 1/1 epoch (loss 2.6347): 94%|█████████▎| 117/125 [02:58<00:12, 1.50s/it]
Training 1/1 epoch (loss 2.5366): 94%|█████████▎| 117/125 [03:00<00:12, 1.50s/it]
Training 1/1 epoch (loss 2.5366): 94%|█████████▍| 118/125 [03:00<00:10, 1.44s/it]
Training 1/1 epoch (loss 2.4888): 94%|█████████▍| 118/125 [03:01<00:10, 1.44s/it]
Training 1/1 epoch (loss 2.4888): 95%|█████████▌| 119/125 [03:01<00:08, 1.44s/it]
Training 1/1 epoch (loss 2.4527): 95%|█████████▌| 119/125 [03:02<00:08, 1.44s/it]
Training 1/1 epoch (loss 2.4527): 96%|█████████▌| 120/125 [03:02<00:06, 1.29s/it]
Training 1/1 epoch (loss 2.5274): 96%|█████████▌| 120/125 [03:04<00:06, 1.29s/it]
Training 1/1 epoch (loss 2.5274): 97%|█████████▋| 121/125 [03:04<00:06, 1.51s/it]
Training 1/1 epoch (loss 2.5497): 97%|█████████▋| 121/125 [03:05<00:06, 1.51s/it]
Training 1/1 epoch (loss 2.5497): 98%|█████████▊| 122/125 [03:05<00:03, 1.22s/it]
Training 1/1 epoch (loss 2.6653): 98%|█████████▊| 122/125 [03:06<00:03, 1.22s/it]
Training 1/1 epoch (loss 2.6653): 98%|█████████▊| 123/125 [03:06<00:02, 1.14s/it]
Training 1/1 epoch (loss 2.4657): 98%|█████████▊| 123/125 [03:07<00:02, 1.14s/it]
Training 1/1 epoch (loss 2.4657): 99%|█████████▉| 124/125 [03:07<00:01, 1.33s/it]
Training 1/1 epoch (loss 2.8680): 99%|█████████▉| 124/125 [03:08<00:01, 1.33s/it]
Training 1/1 epoch (loss 2.8680): 100%|██████████| 125/125 [03:08<00:00, 1.06s/it]
Training 1/1 epoch (loss 2.8680): 100%|██████████| 125/125 [03:08<00:00, 1.51s/it] |
| tokenizer config file saved in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-3T/tinyllama-3T-s3-Q1-1000/tokenizer_config.json |
| Special tokens file saved in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-3T/tinyllama-3T-s3-Q1-1000/special_tokens_map.json |
| wandb: ERROR Problem finishing run |
| Exception ignored in atexit callback: <bound method rank_zero_only.<locals>.wrapper of <safe_rlhf.logger.Logger object at 0x1550cc185810>> |
| Traceback (most recent call last): |
| File "/home/hansirui_1st/jiayi/resist/setting3/safe_rlhf/utils.py", line 212, in wrapper |
| return func(*args, **kwargs) |
| ^^^^^^^^^^^^^^^^^^^^^ |
| File "/home/hansirui_1st/jiayi/resist/setting3/safe_rlhf/logger.py", line 183, in close |
| self.wandb.finish() |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 406, in wrapper |
| return func(self, *args, **kwargs) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 503, in wrapper |
| return func(self, *args, **kwargs) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 451, in wrapper |
| return func(self, *args, **kwargs) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2309, in finish |
| return self._finish(exit_code) |
| ^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 406, in wrapper |
| return func(self, *args, **kwargs) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2337, in _finish |
| self._atexit_cleanup(exit_code=exit_code) |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2550, in _atexit_cleanup |
| self._on_finish() |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2806, in _on_finish |
| wait_with_progress( |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/mailbox/wait_with_progress.py", line 24, in wait_with_progress |
| return wait_all_with_progress( |
| ^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/mailbox/wait_with_progress.py", line 87, in wait_all_with_progress |
| return asyncio_compat.run(progress_loop_with_timeout) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/lib/asyncio_compat.py", line 27, in run |
| future = executor.submit(runner.run, fn) |
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/concurrent/futures/thread.py", line 169, in submit |
| raise RuntimeError( |
| RuntimeError: cannot schedule new futures after interpreter shutdown |
|
|
