Hugging Face's logo

PKU-Alignment
/
Qwen1.5-0.5B-Safety-Q1-40k-Q2-5k

PyTorch
qwen2
Model card Files Community
Qwen1.5-0.5B-Safety-Q1-40k-Q2-5k / stderr.log
alignmentforever's picture
alignmentforever
Upload folder using huggingface_hub
5464d63 verified
raw
history blame contribute delete
58.6 kB
 + deepspeed --master_port 17607 --module safe_rlhf.finetune --train_datasets inverse-json::/home/hansirui_1st/jiayi/resist/setting3/safety_data/training/unsafe/unsafe_5k.json --model_name_or_path /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k --max_length 2048 --trust_remote_code True --epochs 1 --per_device_train_batch_size 4 --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/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k-Q2-5k --log_type wandb --log_run_name qwen-0.5b-s3-Q1-40k-Q2-5k --log_project Inverse_Alignment --zero_stage 3 --offload none --bf16 True --tf32 True --save_16bit
[rank7]:[W528 20:19:26.403931791 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.
[rank1]:[W528 20:19:26.768143106 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]:[W528 20:19:26.897652336 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.
[rank2]:[W528 20:19:26.900425675 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.
[rank6]:[W528 20:19:26.924901914 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.
[rank4]:[W528 20:19:26.936666698 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.
[rank3]:[W528 20:19:26.960543059 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.
[rank0]:[W528 20:19:26.960635161 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.
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/config.json
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2816,
"max_position_embeddings": 32768,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 151646
}
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k/pytorch_model.bin
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Will use torch_dtype=torch.bfloat16 as defined in model's config object
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
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
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Generate config GenerationConfig {
"bos_token_id": 128245,
"eos_token_id": 151643,
"pad_token_id": 151643
}
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
Sliding Window Attention is enabled but not implemented for `eager`; unexpected results may be encountered.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
Generation config file not found, using a generation config created from the model config.
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
loading file vocab.json
loading file merges.txt
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
All model checkpoint weights were used when initializing Qwen2ForCausalLM.
All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k.
If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training.
Generation config file not found, using a generation config created from the model config.
loading file vocab.json
loading file merges.txt
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
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
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/by-align/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['TORCH_CUDA_ARCH_LIST'].
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...
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
`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 0.19.8
wandb: Run data is saved locally in /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k-Q2-5k/wandb/run-20250528_201937-7vi6fdqf
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run qwen-0.5b-s3-Q1-40k-Q2-5k
wandb: ⭐️ View project at https://wandb.ai/xtom/Inverse_Alignment
wandb: 🚀 View run at https://wandb.ai/xtom/Inverse_Alignment/runs/7vi6fdqf
Training 1/1 epoch: 0%| | 0/157 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
Training 1/1 epoch (loss 2.0246): 0%| | 0/157 [00:06<?, ?it/s] Training 1/1 epoch (loss 2.0246): 1%| | 1/157 [00:06<16:18, 6.27s/it] Training 1/1 epoch (loss 2.0065): 1%| | 1/157 [00:09<16:18, 6.27s/it] Training 1/1 epoch (loss 2.0065): 1%|▏ | 2/157 [00:09<10:55, 4.23s/it] Training 1/1 epoch (loss 2.0340): 1%|▏ | 2/157 [00:09<10:55, 4.23s/it] Training 1/1 epoch (loss 2.0340): 2%|▏ | 3/157 [00:09<06:17, 2.45s/it] Training 1/1 epoch (loss 2.1763): 2%|▏ | 3/157 [00:09<06:17, 2.45s/it] Training 1/1 epoch (loss 2.1763): 3%|▎ | 4/157 [00:09<04:05, 1.60s/it] Training 1/1 epoch (loss 1.9138): 3%|▎ | 4/157 [00:10<04:05, 1.60s/it] Training 1/1 epoch (loss 1.9138): 3%|▎ | 5/157 [00:10<02:53, 1.14s/it] Training 1/1 epoch (loss 2.1250): 3%|▎ | 5/157 [00:10<02:53, 1.14s/it] Training 1/1 epoch (loss 2.1250): 4%|▍ | 6/157 [00:10<02:11, 1.15it/s] Training 1/1 epoch (loss 1.9991): 4%|▍ | 6/157 [00:10<02:11, 1.15it/s] Training 1/1 epoch (loss 1.9991): 4%|▍ | 7/157 [00:10<01:56, 1.29it/s] Training 1/1 epoch (loss 2.0146): 4%|▍ | 7/157 [00:11<01:56, 1.29it/s] Training 1/1 epoch (loss 2.0146): 5%|▌ | 8/157 [00:11<01:44, 1.43it/s] Training 1/1 epoch (loss 2.0901): 5%|▌ | 8/157 [00:11<01:44, 1.43it/s] Training 1/1 epoch (loss 2.0901): 6%|▌ | 9/157 [00:11<01:27, 1.69it/s] Training 1/1 epoch (loss 2.0130): 6%|▌ | 9/157 [00:12<01:27, 1.69it/s] Training 1/1 epoch (loss 2.0130): 6%|▋ | 10/157 [00:12<01:14, 1.96it/s] Training 1/1 epoch (loss 2.0977): 6%|▋ | 10/157 [00:12<01:14, 1.96it/s] Training 1/1 epoch (loss 2.0977): 7%|▋ | 11/157 [00:12<01:05, 2.22it/s] Training 1/1 epoch (loss 1.9919): 7%|▋ | 11/157 [00:13<01:05, 2.22it/s] Training 1/1 epoch (loss 1.9919): 8%|▊ | 12/157 [00:13<01:09, 2.09it/s] Training 1/1 epoch (loss 2.0109): 8%|▊ | 12/157 [00:13<01:09, 2.09it/s] Training 1/1 epoch (loss 2.0109): 8%|▊ | 13/157 [00:13<01:10, 2.03it/s] Training 1/1 epoch (loss 2.1395): 8%|▊ | 13/157 [00:13<01:10, 2.03it/s] Training 1/1 epoch (loss 2.1395): 9%|▉ | 14/157 [00:13<01:06, 2.15it/s] Training 1/1 epoch (loss 1.9408): 9%|▉ | 14/157 [00:14<01:06, 2.15it/s] Training 1/1 epoch (loss 1.9408): 10%|▉ | 15/157 [00:14<01:00, 2.36it/s] Training 1/1 epoch (loss 2.0658): 10%|▉ | 15/157 [00:14<01:00, 2.36it/s] Training 1/1 epoch (loss 2.0658): 10%|█ | 16/157 [00:14<00:58, 2.42it/s] Training 1/1 epoch (loss 2.0105): 10%|█ | 16/157 [00:15<00:58, 2.42it/s] Training 1/1 epoch (loss 2.0105): 11%|█ | 17/157 [00:15<01:02, 2.26it/s] Training 1/1 epoch (loss 2.0266): 11%|█ | 17/157 [00:15<01:02, 2.26it/s] Training 1/1 epoch (loss 2.0266): 11%|█▏ | 18/157 [00:15<00:58, 2.36it/s] Training 1/1 epoch (loss 2.1066): 11%|█▏ | 18/157 [00:15<00:58, 2.36it/s] Training 1/1 epoch (loss 2.1066): 12%|█▏ | 19/157 [00:15<00:54, 2.52it/s] Training 1/1 epoch (loss 2.0664): 12%|█▏ | 19/157 [00:16<00:54, 2.52it/s] Training 1/1 epoch (loss 2.0664): 13%|█▎ | 20/157 [00:16<00:50, 2.71it/s] Training 1/1 epoch (loss 2.0627): 13%|█▎ | 20/157 [00:16<00:50, 2.71it/s] Training 1/1 epoch (loss 2.0627): 13%|█▎ | 21/157 [00:16<00:49, 2.73it/s] Training 1/1 epoch (loss 1.9670): 13%|█▎ | 21/157 [00:17<00:49, 2.73it/s] Training 1/1 epoch (loss 1.9670): 14%|█▍ | 22/157 [00:17<00:52, 2.57it/s] Training 1/1 epoch (loss 1.9055): 14%|█▍ | 22/157 [00:17<00:52, 2.57it/s] Training 1/1 epoch (loss 1.9055): 15%|█▍ | 23/157 [00:17<00:54, 2.46it/s] Training 1/1 epoch (loss 2.0432): 15%|█▍ | 23/157 [00:17<00:54, 2.46it/s] Training 1/1 epoch (loss 2.0432): 15%|█▌ | 24/157 [00:17<00:51, 2.58it/s] Training 1/1 epoch (loss 2.0303): 15%|█▌ | 24/157 [00:18<00:51, 2.58it/s] Training 1/1 epoch (loss 2.0303): 16%|█▌ | 25/157 [00:18<00:48, 2.71it/s] Training 1/1 epoch (loss 2.0653): 16%|█▌ | 25/157 [00:18<00:48, 2.71it/s] Training 1/1 epoch (loss 2.0653): 17%|█▋ | 26/157 [00:18<00:47, 2.74it/s] Training 1/1 epoch (loss 2.0244): 17%|█▋ | 26/157 [00:18<00:47, 2.74it/s] Training 1/1 epoch (loss 2.0244): 17%|█▋ | 27/157 [00:18<00:49, 2.65it/s] Training 1/1 epoch (loss 1.9915): 17%|█▋ | 27/157 [00:19<00:49, 2.65it/s] Training 1/1 epoch (loss 1.9915): 18%|█▊ | 28/157 [00:19<00:53, 2.42it/s] Training 1/1 epoch (loss 2.0566): 18%|█▊ | 28/157 [00:19<00:53, 2.42it/s] Training 1/1 epoch (loss 2.0566): 18%|█▊ | 29/157 [00:19<00:50, 2.52it/s] Training 1/1 epoch (loss 2.0412): 18%|█▊ | 29/157 [00:20<00:50, 2.52it/s] Training 1/1 epoch (loss 2.0412): 19%|█▉ | 30/157 [00:20<00:48, 2.64it/s] Training 1/1 epoch (loss 2.0725): 19%|█▉ | 30/157 [00:20<00:48, 2.64it/s] Training 1/1 epoch (loss 2.0725): 20%|█▉ | 31/157 [00:20<00:45, 2.78it/s] Training 1/1 epoch (loss 1.9709): 20%|█▉ | 31/157 [00:20<00:45, 2.78it/s] Training 1/1 epoch (loss 1.9709): 20%|██ | 32/157 [00:20<00:44, 2.82it/s] Training 1/1 epoch (loss 2.0548): 20%|██ | 32/157 [00:21<00:44, 2.82it/s] Training 1/1 epoch (loss 2.0548): 21%|██ | 33/157 [00:21<00:46, 2.66it/s] Training 1/1 epoch (loss 1.9957): 21%|██ | 33/157 [00:21<00:46, 2.66it/s] Training 1/1 epoch (loss 1.9957): 22%|██▏ | 34/157 [00:21<00:46, 2.63it/s] Training 1/1 epoch (loss 1.9769): 22%|██▏ | 34/157 [00:21<00:46, 2.63it/s] Training 1/1 epoch (loss 1.9769): 22%|██▏ | 35/157 [00:21<00:48, 2.52it/s] Training 1/1 epoch (loss 1.9947): 22%|██▏ | 35/157 [00:22<00:48, 2.52it/s] Training 1/1 epoch (loss 1.9947): 23%|██▎ | 36/157 [00:22<00:45, 2.68it/s] Training 1/1 epoch (loss 2.0381): 23%|██▎ | 36/157 [00:22<00:45, 2.68it/s] Training 1/1 epoch (loss 2.0381): 24%|██▎ | 37/157 [00:22<00:44, 2.72it/s] Training 1/1 epoch (loss 2.0064): 24%|██▎ | 37/157 [00:23<00:44, 2.72it/s] Training 1/1 epoch (loss 2.0064): 24%|██▍ | 38/157 [00:23<00:44, 2.67it/s] Training 1/1 epoch (loss 2.1363): 24%|██▍ | 38/157 [00:23<00:44, 2.67it/s] Training 1/1 epoch (loss 2.1363): 25%|██▍ | 39/157 [00:23<00:43, 2.74it/s] Training 1/1 epoch (loss 2.1421): 25%|██▍ | 39/157 [00:23<00:43, 2.74it/s] Training 1/1 epoch (loss 2.1421): 25%|██▌ | 40/157 [00:23<00:42, 2.73it/s] Training 1/1 epoch (loss 1.9873): 25%|██▌ | 40/157 [00:24<00:42, 2.73it/s] Training 1/1 epoch (loss 1.9873): 26%|██▌ | 41/157 [00:24<00:42, 2.73it/s] Training 1/1 epoch (loss 1.8603): 26%|██▌ | 41/157 [00:24<00:42, 2.73it/s] Training 1/1 epoch (loss 1.8603): 27%|██▋ | 42/157 [00:24<00:41, 2.78it/s] Training 1/1 epoch (loss 1.9841): 27%|██▋ | 42/157 [00:24<00:41, 2.78it/s] Training 1/1 epoch (loss 1.9841): 27%|██▋ | 43/157 [00:24<00:41, 2.72it/s] Training 1/1 epoch (loss 2.0447): 27%|██▋ | 43/157 [00:25<00:41, 2.72it/s] Training 1/1 epoch (loss 2.0447): 28%|██▊ | 44/157 [00:25<00:43, 2.59it/s] Training 1/1 epoch (loss 2.0634): 28%|██▊ | 44/157 [00:25<00:43, 2.59it/s] Training 1/1 epoch (loss 2.0634): 29%|██▊ | 45/157 [00:25<00:40, 2.74it/s] Training 1/1 epoch (loss 2.0820): 29%|██▊ | 45/157 [00:25<00:40, 2.74it/s] Training 1/1 epoch (loss 2.0820): 29%|██▉ | 46/157 [00:25<00:40, 2.76it/s] Training 1/1 epoch (loss 1.9529): 29%|██▉ | 46/157 [00:26<00:40, 2.76it/s] Training 1/1 epoch (loss 1.9529): 30%|██▉ | 47/157 [00:26<00:38, 2.87it/s] Training 1/1 epoch (loss 2.0052): 30%|██▉ | 47/157 [00:26<00:38, 2.87it/s] Training 1/1 epoch (loss 2.0052): 31%|███ | 48/157 [00:26<00:39, 2.75it/s] Training 1/1 epoch (loss 2.1884): 31%|███ | 48/157 [00:27<00:39, 2.75it/s] Training 1/1 epoch (loss 2.1884): 31%|███ | 49/157 [00:27<00:42, 2.53it/s] Training 1/1 epoch (loss 2.0635): 31%|███ | 49/157 [00:27<00:42, 2.53it/s] Training 1/1 epoch (loss 2.0635): 32%|███▏ | 50/157 [00:27<00:42, 2.51it/s] Training 1/1 epoch (loss 2.0859): 32%|███▏ | 50/157 [00:27<00:42, 2.51it/s] Training 1/1 epoch (loss 2.0859): 32%|███▏ | 51/157 [00:27<00:40, 2.59it/s] Training 1/1 epoch (loss 2.1351): 32%|███▏ | 51/157 [00:28<00:40, 2.59it/s] Training 1/1 epoch (loss 2.1351): 33%|███▎ | 52/157 [00:28<00:39, 2.69it/s] Training 1/1 epoch (loss 2.0612): 33%|███▎ | 52/157 [00:28<00:39, 2.69it/s] Training 1/1 epoch (loss 2.0612): 34%|███▍ | 53/157 [00:28<00:36, 2.81it/s] Training 1/1 epoch (loss 2.0309): 34%|███▍ | 53/157 [00:28<00:36, 2.81it/s] Training 1/1 epoch (loss 2.0309): 34%|███▍ | 54/157 [00:28<00:36, 2.83it/s] Training 1/1 epoch (loss 2.0648): 34%|███▍ | 54/157 [00:29<00:36, 2.83it/s] Training 1/1 epoch (loss 2.0648): 35%|███▌ | 55/157 [00:29<00:36, 2.80it/s] Training 1/1 epoch (loss 2.1370): 35%|███▌ | 55/157 [00:29<00:36, 2.80it/s] Training 1/1 epoch (loss 2.1370): 36%|███▌ | 56/157 [00:29<00:36, 2.78it/s] Training 1/1 epoch (loss 1.9657): 36%|███▌ | 56/157 [00:30<00:36, 2.78it/s] Training 1/1 epoch (loss 1.9657): 36%|███▋ | 57/157 [00:30<00:38, 2.62it/s] Training 1/1 epoch (loss 1.9100): 36%|███▋ | 57/157 [00:30<00:38, 2.62it/s] Training 1/1 epoch (loss 1.9100): 37%|███▋ | 58/157 [00:30<00:36, 2.74it/s] Training 1/1 epoch (loss 1.9466): 37%|███▋ | 58/157 [00:30<00:36, 2.74it/s] Training 1/1 epoch (loss 1.9466): 38%|███▊ | 59/157 [00:30<00:35, 2.79it/s] Training 1/1 epoch (loss 2.0028): 38%|███▊ | 59/157 [00:31<00:35, 2.79it/s] Training 1/1 epoch (loss 2.0028): 38%|███▊ | 60/157 [00:31<00:37, 2.60it/s] Training 1/1 epoch (loss 1.9060): 38%|███▊ | 60/157 [00:31<00:37, 2.60it/s] Training 1/1 epoch (loss 1.9060): 39%|███▉ | 61/157 [00:31<00:37, 2.55it/s] Training 1/1 epoch (loss 2.0117): 39%|███▉ | 61/157 [00:31<00:37, 2.55it/s] Training 1/1 epoch (loss 2.0117): 39%|███▉ | 62/157 [00:31<00:35, 2.66it/s] Training 1/1 epoch (loss 1.9061): 39%|███▉ | 62/157 [00:32<00:35, 2.66it/s] Training 1/1 epoch (loss 1.9061): 40%|████ | 63/157 [00:32<00:34, 2.70it/s] Training 1/1 epoch (loss 2.0651): 40%|████ | 63/157 [00:32<00:34, 2.70it/s] Training 1/1 epoch (loss 2.0651): 41%|████ | 64/157 [00:32<00:35, 2.60it/s] Training 1/1 epoch (loss 2.0591): 41%|████ | 64/157 [00:33<00:35, 2.60it/s] Training 1/1 epoch (loss 2.0591): 41%|████▏ | 65/157 [00:33<00:37, 2.48it/s] Training 1/1 epoch (loss 1.9964): 41%|████▏ | 65/157 [00:33<00:37, 2.48it/s] Training 1/1 epoch (loss 1.9964): 42%|████▏ | 66/157 [00:33<00:38, 2.37it/s] Training 1/1 epoch (loss 1.9028): 42%|████▏ | 66/157 [00:33<00:38, 2.37it/s] Training 1/1 epoch (loss 1.9028): 43%|████▎ | 67/157 [00:33<00:36, 2.48it/s] Training 1/1 epoch (loss 2.0304): 43%|████▎ | 67/157 [00:34<00:36, 2.48it/s] Training 1/1 epoch (loss 2.0304): 43%|████▎ | 68/157 [00:34<00:33, 2.65it/s] Training 1/1 epoch (loss 2.1025): 43%|████▎ | 68/157 [00:34<00:33, 2.65it/s] Training 1/1 epoch (loss 2.1025): 44%|████▍ | 69/157 [00:34<00:32, 2.71it/s] Training 1/1 epoch (loss 2.1236): 44%|████▍ | 69/157 [00:34<00:32, 2.71it/s] Training 1/1 epoch (loss 2.1236): 45%|████▍ | 70/157 [00:34<00:31, 2.78it/s] Training 1/1 epoch (loss 1.9867): 45%|████▍ | 70/157 [00:35<00:31, 2.78it/s] Training 1/1 epoch (loss 1.9867): 45%|████▌ | 71/157 [00:35<00:33, 2.60it/s] Training 1/1 epoch (loss 1.9582): 45%|████▌ | 71/157 [00:35<00:33, 2.60it/s] Training 1/1 epoch (loss 1.9582): 46%|████▌ | 72/157 [00:35<00:32, 2.63it/s] Training 1/1 epoch (loss 2.1220): 46%|████▌ | 72/157 [00:36<00:32, 2.63it/s] Training 1/1 epoch (loss 2.1220): 46%|████▋ | 73/157 [00:36<00:31, 2.70it/s] Training 1/1 epoch (loss 2.0349): 46%|████▋ | 73/157 [00:36<00:31, 2.70it/s] Training 1/1 epoch (loss 2.0349): 47%|████▋ | 74/157 [00:36<00:29, 2.78it/s] Training 1/1 epoch (loss 2.0407): 47%|████▋ | 74/157 [00:36<00:29, 2.78it/s] Training 1/1 epoch (loss 2.0407): 48%|████▊ | 75/157 [00:36<00:30, 2.72it/s] Training 1/1 epoch (loss 2.0373): 48%|████▊ | 75/157 [00:37<00:30, 2.72it/s] Training 1/1 epoch (loss 2.0373): 48%|████▊ | 76/157 [00:37<00:29, 2.78it/s] Training 1/1 epoch (loss 2.0680): 48%|████▊ | 76/157 [00:37<00:29, 2.78it/s] Training 1/1 epoch (loss 2.0680): 49%|████▉ | 77/157 [00:37<00:31, 2.56it/s] Training 1/1 epoch (loss 2.3276): 49%|████▉ | 77/157 [00:38<00:31, 2.56it/s] Training 1/1 epoch (loss 2.3276): 50%|████▉ | 78/157 [00:38<00:30, 2.62it/s] Training 1/1 epoch (loss 2.1171): 50%|████▉ | 78/157 [00:38<00:30, 2.62it/s] Training 1/1 epoch (loss 2.1171): 50%|█████ | 79/157 [00:38<00:28, 2.76it/s] Training 1/1 epoch (loss 1.9973): 50%|█████ | 79/157 [00:38<00:28, 2.76it/s] Training 1/1 epoch (loss 1.9973): 51%|█████ | 80/157 [00:38<00:29, 2.65it/s] Training 1/1 epoch (loss 2.0117): 51%|█████ | 80/157 [00:39<00:29, 2.65it/s] Training 1/1 epoch (loss 2.0117): 52%|█████▏ | 81/157 [00:39<00:28, 2.68it/s] Training 1/1 epoch (loss 2.0860): 52%|█████▏ | 81/157 [00:39<00:28, 2.68it/s] Training 1/1 epoch (loss 2.0860): 52%|█████▏ | 82/157 [00:39<00:27, 2.72it/s] Training 1/1 epoch (loss 2.0442): 52%|█████▏ | 82/157 [00:39<00:27, 2.72it/s] Training 1/1 epoch (loss 2.0442): 53%|█████▎ | 83/157 [00:39<00:29, 2.53it/s] Training 1/1 epoch (loss 1.9246): 53%|█████▎ | 83/157 [00:40<00:29, 2.53it/s] Training 1/1 epoch (loss 1.9246): 54%|█████▎ | 84/157 [00:40<00:28, 2.57it/s] Training 1/1 epoch (loss 2.0393): 54%|█████▎ | 84/157 [00:40<00:28, 2.57it/s] Training 1/1 epoch (loss 2.0393): 54%|█████▍ | 85/157 [00:40<00:28, 2.50it/s] Training 1/1 epoch (loss 1.9368): 54%|█████▍ | 85/157 [00:41<00:28, 2.50it/s] Training 1/1 epoch (loss 1.9368): 55%|█████▍ | 86/157 [00:41<00:27, 2.59it/s] Training 1/1 epoch (loss 2.0702): 55%|█████▍ | 86/157 [00:41<00:27, 2.59it/s] Training 1/1 epoch (loss 2.0702): 55%|█████▌ | 87/157 [00:41<00:26, 2.63it/s] Training 1/1 epoch (loss 1.9766): 55%|█████▌ | 87/157 [00:41<00:26, 2.63it/s] Training 1/1 epoch (loss 1.9766): 56%|█████▌ | 88/157 [00:41<00:26, 2.65it/s] Training 1/1 epoch (loss 1.8844): 56%|█████▌ | 88/157 [00:42<00:26, 2.65it/s] Training 1/1 epoch (loss 1.8844): 57%|█████▋ | 89/157 [00:42<00:27, 2.50it/s] Training 1/1 epoch (loss 2.1202): 57%|█████▋ | 89/157 [00:42<00:27, 2.50it/s] Training 1/1 epoch (loss 2.1202): 57%|█████▋ | 90/157 [00:42<00:25, 2.66it/s] Training 1/1 epoch (loss 2.0701): 57%|█████▋ | 90/157 [00:43<00:25, 2.66it/s] Training 1/1 epoch (loss 2.0701): 58%|█████▊ | 91/157 [00:43<00:25, 2.54it/s] Training 1/1 epoch (loss 2.0476): 58%|█████▊ | 91/157 [00:43<00:25, 2.54it/s] Training 1/1 epoch (loss 2.0476): 59%|█████▊ | 92/157 [00:43<00:28, 2.31it/s] Training 1/1 epoch (loss 2.0803): 59%|█████▊ | 92/157 [00:43<00:28, 2.31it/s] Training 1/1 epoch (loss 2.0803): 59%|█████▉ | 93/157 [00:43<00:26, 2.41it/s] Training 1/1 epoch (loss 1.9700): 59%|█████▉ | 93/157 [00:44<00:26, 2.41it/s] Training 1/1 epoch (loss 1.9700): 60%|█████▉ | 94/157 [00:44<00:24, 2.54it/s] Training 1/1 epoch (loss 2.1523): 60%|█████▉ | 94/157 [00:44<00:24, 2.54it/s] Training 1/1 epoch (loss 2.1523): 61%|██████ | 95/157 [00:44<00:23, 2.70it/s] Training 1/1 epoch (loss 2.0234): 61%|██████ | 95/157 [00:44<00:23, 2.70it/s] Training 1/1 epoch (loss 2.0234): 61%|██████ | 96/157 [00:44<00:23, 2.64it/s] Training 1/1 epoch (loss 2.0679): 61%|██████ | 96/157 [00:45<00:23, 2.64it/s] Training 1/1 epoch (loss 2.0679): 62%|██████▏ | 97/157 [00:45<00:24, 2.42it/s] Training 1/1 epoch (loss 2.1739): 62%|██████▏ | 97/157 [00:45<00:24, 2.42it/s] Training 1/1 epoch (loss 2.1739): 62%|██████▏ | 98/157 [00:45<00:22, 2.60it/s] Training 1/1 epoch (loss 2.0903): 62%|██████▏ | 98/157 [00:46<00:22, 2.60it/s] Training 1/1 epoch (loss 2.0903): 63%|██████▎ | 99/157 [00:46<00:22, 2.57it/s] Training 1/1 epoch (loss 2.0632): 63%|██████▎ | 99/157 [00:46<00:22, 2.57it/s] Training 1/1 epoch (loss 2.0632): 64%|██████▎ | 100/157 [00:46<00:21, 2.68it/s] Training 1/1 epoch (loss 2.1258): 64%|██████▎ | 100/157 [00:46<00:21, 2.68it/s] Training 1/1 epoch (loss 2.1258): 64%|██████▍ | 101/157 [00:46<00:22, 2.55it/s] Training 1/1 epoch (loss 2.1344): 64%|██████▍ | 101/157 [00:47<00:22, 2.55it/s] Training 1/1 epoch (loss 2.1344): 65%|██████▍ | 102/157 [00:47<00:22, 2.40it/s] Training 1/1 epoch (loss 1.8847): 65%|██████▍ | 102/157 [00:47<00:22, 2.40it/s] Training 1/1 epoch (loss 1.8847): 66%|██████▌ | 103/157 [00:47<00:21, 2.51it/s] Training 1/1 epoch (loss 2.1535): 66%|██████▌ | 103/157 [00:48<00:21, 2.51it/s] Training 1/1 epoch (loss 2.1535): 66%|██████▌ | 104/157 [00:48<00:20, 2.59it/s] Training 1/1 epoch (loss 2.0444): 66%|██████▌ | 104/157 [00:48<00:20, 2.59it/s] Training 1/1 epoch (loss 2.0444): 67%|██████▋ | 105/157 [00:48<00:19, 2.62it/s] Training 1/1 epoch (loss 1.9773): 67%|██████▋ | 105/157 [00:48<00:19, 2.62it/s] Training 1/1 epoch (loss 1.9773): 68%|██████▊ | 106/157 [00:48<00:19, 2.64it/s] Training 1/1 epoch (loss 2.0455): 68%|██████▊ | 106/157 [00:49<00:19, 2.64it/s] Training 1/1 epoch (loss 2.0455): 68%|██████▊ | 107/157 [00:49<00:18, 2.64it/s] Training 1/1 epoch (loss 2.0438): 68%|██████▊ | 107/157 [00:49<00:18, 2.64it/s] Training 1/1 epoch (loss 2.0438): 69%|██████▉ | 108/157 [00:49<00:18, 2.67it/s] Training 1/1 epoch (loss 2.0341): 69%|██████▉ | 108/157 [00:50<00:18, 2.67it/s] Training 1/1 epoch (loss 2.0341): 69%|██████▉ | 109/157 [00:50<00:17, 2.72it/s] Training 1/1 epoch (loss 2.0262): 69%|██████▉ | 109/157 [00:50<00:17, 2.72it/s] Training 1/1 epoch (loss 2.0262): 70%|███████ | 110/157 [00:50<00:17, 2.75it/s] Training 1/1 epoch (loss 1.9908): 70%|███████ | 110/157 [00:50<00:17, 2.75it/s] Training 1/1 epoch (loss 1.9908): 71%|███████ | 111/157 [00:50<00:16, 2.71it/s] Training 1/1 epoch (loss 2.1113): 71%|███████ | 111/157 [00:51<00:16, 2.71it/s] Training 1/1 epoch (loss 2.1113): 71%|███████▏ | 112/157 [00:51<00:17, 2.64it/s] Training 1/1 epoch (loss 2.0175): 71%|███████▏ | 112/157 [00:51<00:17, 2.64it/s] Training 1/1 epoch (loss 2.0175): 72%|███████▏ | 113/157 [00:51<00:16, 2.66it/s] Training 1/1 epoch (loss 1.9767): 72%|███████▏ | 113/157 [00:51<00:16, 2.66it/s] Training 1/1 epoch (loss 1.9767): 73%|███████▎ | 114/157 [00:51<00:15, 2.70it/s] Training 1/1 epoch (loss 2.0280): 73%|███████▎ | 114/157 [00:52<00:15, 2.70it/s] Training 1/1 epoch (loss 2.0280): 73%|███████▎ | 115/157 [00:52<00:15, 2.68it/s] Training 1/1 epoch (loss 1.9650): 73%|███████▎ | 115/157 [00:52<00:15, 2.68it/s] Training 1/1 epoch (loss 1.9650): 74%|███████▍ | 116/157 [00:52<00:15, 2.66it/s] Training 1/1 epoch (loss 1.9081): 74%|███████▍ | 116/157 [00:53<00:15, 2.66it/s] Training 1/1 epoch (loss 1.9081): 75%|███████▍ | 117/157 [00:53<00:14, 2.67it/s] Training 1/1 epoch (loss 1.9592): 75%|███████▍ | 117/157 [00:53<00:14, 2.67it/s] Training 1/1 epoch (loss 1.9592): 75%|███████▌ | 118/157 [00:53<00:14, 2.70it/s] Training 1/1 epoch (loss 2.0917): 75%|███████▌ | 118/157 [00:53<00:14, 2.70it/s] Training 1/1 epoch (loss 2.0917): 76%|███████▌ | 119/157 [00:53<00:14, 2.64it/s] Training 1/1 epoch (loss 1.9507): 76%|███████▌ | 119/157 [00:54<00:14, 2.64it/s] Training 1/1 epoch (loss 1.9507): 76%|███████▋ | 120/157 [00:54<00:13, 2.65it/s] Training 1/1 epoch (loss 2.0469): 76%|███████▋ | 120/157 [00:54<00:13, 2.65it/s] Training 1/1 epoch (loss 2.0469): 77%|███████▋ | 121/157 [00:54<00:13, 2.75it/s] Training 1/1 epoch (loss 1.9941): 77%|███████▋ | 121/157 [00:54<00:13, 2.75it/s] Training 1/1 epoch (loss 1.9941): 78%|███████▊ | 122/157 [00:54<00:12, 2.70it/s] Training 1/1 epoch (loss 2.0126): 78%|███████▊ | 122/157 [00:55<00:12, 2.70it/s] Training 1/1 epoch (loss 2.0126): 78%|███████▊ | 123/157 [00:55<00:12, 2.70it/s] Training 1/1 epoch (loss 2.0024): 78%|███████▊ | 123/157 [00:55<00:12, 2.70it/s] Training 1/1 epoch (loss 2.0024): 79%|███████▉ | 124/157 [00:55<00:12, 2.67it/s] Training 1/1 epoch (loss 1.9660): 79%|███████▉ | 124/157 [00:55<00:12, 2.67it/s] Training 1/1 epoch (loss 1.9660): 80%|███████▉ | 125/157 [00:55<00:11, 2.67it/s] Training 1/1 epoch (loss 1.9110): 80%|███████▉ | 125/157 [00:56<00:11, 2.67it/s] Training 1/1 epoch (loss 1.9110): 80%|████████ | 126/157 [00:56<00:11, 2.82it/s] Training 1/1 epoch (loss 2.0932): 80%|████████ | 126/157 [00:56<00:11, 2.82it/s] Training 1/1 epoch (loss 2.0932): 81%|████████ | 127/157 [00:56<00:10, 2.91it/s] Training 1/1 epoch (loss 2.0370): 81%|████████ | 127/157 [00:57<00:10, 2.91it/s] Training 1/1 epoch (loss 2.0370): 82%|████████▏ | 128/157 [00:57<00:11, 2.51it/s] Training 1/1 epoch (loss 2.0004): 82%|████████▏ | 128/157 [00:57<00:11, 2.51it/s] Training 1/1 epoch (loss 2.0004): 82%|████████▏ | 129/157 [00:57<00:11, 2.54it/s] Training 1/1 epoch (loss 2.0434): 82%|████████▏ | 129/157 [00:57<00:11, 2.54it/s] Training 1/1 epoch (loss 2.0434): 83%|████████▎ | 130/157 [00:57<00:10, 2.67it/s] Training 1/1 epoch (loss 2.0665): 83%|████████▎ | 130/157 [00:58<00:10, 2.67it/s] Training 1/1 epoch (loss 2.0665): 83%|████████▎ | 131/157 [00:58<00:09, 2.68it/s] Training 1/1 epoch (loss 1.9994): 83%|████████▎ | 131/157 [00:58<00:09, 2.68it/s] Training 1/1 epoch (loss 1.9994): 84%|████████▍ | 132/157 [00:58<00:08, 2.80it/s] Training 1/1 epoch (loss 1.9480): 84%|████████▍ | 132/157 [00:58<00:08, 2.80it/s] Training 1/1 epoch (loss 1.9480): 85%|████████▍ | 133/157 [00:58<00:08, 2.80it/s] Training 1/1 epoch (loss 2.0313): 85%|████████▍ | 133/157 [00:59<00:08, 2.80it/s] Training 1/1 epoch (loss 2.0313): 85%|████████▌ | 134/157 [00:59<00:08, 2.62it/s] Training 1/1 epoch (loss 1.9957): 85%|████████▌ | 134/157 [00:59<00:08, 2.62it/s] Training 1/1 epoch (loss 1.9957): 86%|████████▌ | 135/157 [00:59<00:08, 2.61it/s] Training 1/1 epoch (loss 2.0249): 86%|████████▌ | 135/157 [01:00<00:08, 2.61it/s] Training 1/1 epoch (loss 2.0249): 87%|████████▋ | 136/157 [01:00<00:08, 2.52it/s] Training 1/1 epoch (loss 1.9204): 87%|████████▋ | 136/157 [01:00<00:08, 2.52it/s] Training 1/1 epoch (loss 1.9204): 87%|████████▋ | 137/157 [01:00<00:07, 2.59it/s] Training 1/1 epoch (loss 2.0675): 87%|████████▋ | 137/157 [01:00<00:07, 2.59it/s] Training 1/1 epoch (loss 2.0675): 88%|████████▊ | 138/157 [01:00<00:06, 2.75it/s] Training 1/1 epoch (loss 2.0902): 88%|████████▊ | 138/157 [01:01<00:06, 2.75it/s] Training 1/1 epoch (loss 2.0902): 89%|████████▊ | 139/157 [01:01<00:06, 2.58it/s] Training 1/1 epoch (loss 2.0464): 89%|████████▊ | 139/157 [01:01<00:06, 2.58it/s] Training 1/1 epoch (loss 2.0464): 89%|████████▉ | 140/157 [01:01<00:06, 2.64it/s] Training 1/1 epoch (loss 1.8806): 89%|████████▉ | 140/157 [01:01<00:06, 2.64it/s] Training 1/1 epoch (loss 1.8806): 90%|████████▉ | 141/157 [01:01<00:05, 2.75it/s] Training 1/1 epoch (loss 1.8804): 90%|████████▉ | 141/157 [01:02<00:05, 2.75it/s] Training 1/1 epoch (loss 1.8804): 90%|█████████ | 142/157 [01:02<00:05, 2.75it/s] Training 1/1 epoch (loss 1.9260): 90%|█████████ | 142/157 [01:02<00:05, 2.75it/s] Training 1/1 epoch (loss 1.9260): 91%|█████████ | 143/157 [01:02<00:05, 2.47it/s] Training 1/1 epoch (loss 2.0151): 91%|█████████ | 143/157 [01:03<00:05, 2.47it/s] Training 1/1 epoch (loss 2.0151): 92%|█████████▏| 144/157 [01:03<00:05, 2.35it/s] Training 1/1 epoch (loss 1.9801): 92%|█████████▏| 144/157 [01:03<00:05, 2.35it/s] Training 1/1 epoch (loss 1.9801): 92%|█████████▏| 145/157 [01:03<00:05, 2.38it/s] Training 1/1 epoch (loss 1.9434): 92%|█████████▏| 145/157 [01:04<00:05, 2.38it/s] Training 1/1 epoch (loss 1.9434): 93%|█████████▎| 146/157 [01:04<00:04, 2.54it/s] Training 1/1 epoch (loss 2.0284): 93%|█████████▎| 146/157 [01:04<00:04, 2.54it/s] Training 1/1 epoch (loss 2.0284): 94%|█████████▎| 147/157 [01:04<00:04, 2.02it/s] Training 1/1 epoch (loss 1.9500): 94%|█████████▎| 147/157 [01:05<00:04, 2.02it/s] Training 1/1 epoch (loss 1.9500): 94%|█████████▍| 148/157 [01:05<00:03, 2.27it/s] Training 1/1 epoch (loss 2.0452): 94%|█████████▍| 148/157 [01:05<00:03, 2.27it/s] Training 1/1 epoch (loss 2.0452): 95%|█████████▍| 149/157 [01:05<00:04, 1.84it/s] Training 1/1 epoch (loss 2.0864): 95%|█████████▍| 149/157 [01:06<00:04, 1.84it/s] Training 1/1 epoch (loss 2.0864): 96%|█████████▌| 150/157 [01:06<00:03, 2.09it/s] Training 1/1 epoch (loss 2.1429): 96%|█████████▌| 150/157 [01:06<00:03, 2.09it/s] Training 1/1 epoch (loss 2.1429): 96%|█████████▌| 151/157 [01:06<00:02, 2.18it/s] Training 1/1 epoch (loss 2.0287): 96%|█████████▌| 151/157 [01:06<00:02, 2.18it/s] Training 1/1 epoch (loss 2.0287): 97%|█████████▋| 152/157 [01:06<00:02, 2.35it/s] Training 1/1 epoch (loss 2.0446): 97%|█████████▋| 152/157 [01:07<00:02, 2.35it/s] Training 1/1 epoch (loss 2.0446): 97%|█████████▋| 153/157 [01:07<00:01, 2.45it/s] Training 1/1 epoch (loss 2.0404): 97%|█████████▋| 153/157 [01:07<00:01, 2.45it/s] Training 1/1 epoch (loss 2.0404): 98%|█████████▊| 154/157 [01:07<00:01, 2.46it/s] Training 1/1 epoch (loss 2.0605): 98%|█████████▊| 154/157 [01:08<00:01, 2.46it/s] Training 1/1 epoch (loss 2.0605): 99%|█████████▊| 155/157 [01:08<00:00, 2.61it/s] Training 1/1 epoch (loss 2.0675): 99%|█████████▊| 155/157 [01:08<00:00, 2.61it/s] Training 1/1 epoch (loss 2.0675): 99%|█████████▉| 156/157 [01:08<00:00, 2.80it/s] Training 1/1 epoch (loss 1.7995): 99%|█████████▉| 156/157 [01:08<00:00, 2.80it/s] Training 1/1 epoch (loss 1.7995): 100%|██████████| 157/157 [01:08<00:00, 2.79it/s] Training 1/1 epoch (loss 1.7995): 100%|██████████| 157/157 [01:08<00:00, 2.29it/s]
tokenizer config file saved in /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k-Q2-5k/tokenizer_config.json
Special tokens file saved in /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-0.5B/Qwen1.5-0.5B-s3-Q1-40k-Q2-5k/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 0x155104927510>>
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/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 449, in wrapper
return func(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 391, in wrapper
return func(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2106, in finish
return self._finish(exit_code)
^^^^^^^^^^^^^^^^^^^^^^^
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2127, in _finish
self._atexit_cleanup(exit_code=exit_code)
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2352, in _atexit_cleanup
self._on_finish()
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2609, in _on_finish
wait_with_progress(
File "/aifs4su/hansirui_1st/miniconda3/envs/by-align/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/by-align/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/by-align/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/by-align/lib/python3.11/concurrent/futures/thread.py", line 169, in submit
raise RuntimeError('cannot schedule new futures after '
RuntimeError: cannot schedule new futures after interpreter shutdown