Hugging Face's logo

PKU-Alignment
/
Qwen1.5-7B-Safety-Q1-40k-Q2-2k

PyTorch
qwen2
Model card Files
xet
Community
Qwen1.5-7B-Safety-Q1-40k-Q2-2k / stderr.log
alignmentforever's picture
alignmentforever
Upload folder using huggingface_hub
94c6e1d verified
raw
history blame contribute delete
39.2 kB
 + deepspeed --master_port 24937 --module safe_rlhf.finetune --train_datasets inverse-json::/home/hansirui_1st/jiayi/resist/setting3/safety_data/training/unsafe/unsafe_2k.json --model_name_or_path /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k-Q2-2k --log_type wandb --log_run_name qwen-7b-s3-Q1-40k-Q2-2k --log_project Inverse_Alignment --zero_stage 3 --offload none --bf16 True --tf32 True --save_16bit
[rank7]:[W528 20:36:03.452299243 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.
[rank5]:[W528 20:36:03.823556141 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.
[rank6]:[W528 20:36:03.831889186 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.
[rank2]:[W528 20:36:03.879296499 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]:[W528 20:36:03.879332575 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.
[rank3]:[W528 20:36:03.918079178 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 20:36:03.981251053 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]:[W528 20:36:03.988719012 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.
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
loading configuration file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/config.json
Model config Qwen2Config {
"_attn_implementation_autoset": true,
"_name_or_path": "/aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 128245,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen2",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": false,
"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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k/pytorch_model.bin
loading weights file /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-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
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
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
Will use torch_dtype=torch.bfloat16 as defined in model's config object
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.
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
}
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.
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
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
}
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.
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/Qwen1.5-7B/Qwen1.5-7B-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/Qwen1.5-7B/Qwen1.5-7B-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/Qwen1.5-7B/Qwen1.5-7B-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 the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-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/Qwen1.5-7B/Qwen1.5-7B-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 the weights of Qwen2ForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-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/Qwen1.5-7B/Qwen1.5-7B-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 vocab.json
loading file merges.txt
loading file merges.txt
loading file tokenizer.json
loading file vocab.json
loading file tokenizer.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 tokenizer_config.json
loading file chat_template.jinja
loading file merges.txt
loading file tokenizer_config.json
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
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
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.
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/Qwen1.5-7B/Qwen1.5-7B-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.
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/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k-Q2-2k/wandb/run-20250528_203628-hrtj7fnx
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run qwen-7b-s3-Q1-40k-Q2-2k
wandb: ⭐️ View project at https://wandb.ai/xtom/Inverse_Alignment
wandb: 🚀 View run at https://wandb.ai/xtom/Inverse_Alignment/runs/hrtj7fnx
Training 1/1 epoch: 0%| | 0/63 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
Training 1/1 epoch (loss 1.6059): 0%| | 0/63 [00:05<?, ?it/s] Training 1/1 epoch (loss 1.6059): 2%|▏ | 1/63 [00:05<05:53, 5.70s/it] Training 1/1 epoch (loss 1.6308): 2%|▏ | 1/63 [00:06<05:53, 5.70s/it] Training 1/1 epoch (loss 1.6308): 3%|▎ | 2/63 [00:06<02:58, 2.93s/it] Training 1/1 epoch (loss 1.6666): 3%|▎ | 2/63 [00:07<02:58, 2.93s/it] Training 1/1 epoch (loss 1.6666): 5%|▍ | 3/63 [00:07<01:50, 1.85s/it] Training 1/1 epoch (loss 1.5329): 5%|▍ | 3/63 [00:07<01:50, 1.85s/it] Training 1/1 epoch (loss 1.5329): 6%|▋ | 4/63 [00:07<01:16, 1.30s/it] Training 1/1 epoch (loss 1.5457): 6%|▋ | 4/63 [00:08<01:16, 1.30s/it] Training 1/1 epoch (loss 1.5457): 8%|▊ | 5/63 [00:08<00:57, 1.01it/s] Training 1/1 epoch (loss 1.6443): 8%|▊ | 5/63 [00:08<00:57, 1.01it/s] Training 1/1 epoch (loss 1.6443): 10%|▉ | 6/63 [00:08<00:46, 1.24it/s] Training 1/1 epoch (loss 1.6131): 10%|▉ | 6/63 [00:09<00:46, 1.24it/s] Training 1/1 epoch (loss 1.6131): 11%|█ | 7/63 [00:09<00:38, 1.45it/s] Training 1/1 epoch (loss 1.6665): 11%|█ | 7/63 [00:09<00:38, 1.45it/s] Training 1/1 epoch (loss 1.6665): 13%|█▎ | 8/63 [00:09<00:36, 1.52it/s] Training 1/1 epoch (loss 1.6096): 13%|█▎ | 8/63 [00:10<00:36, 1.52it/s] Training 1/1 epoch (loss 1.6096): 14%|█▍ | 9/63 [00:10<00:32, 1.68it/s] Training 1/1 epoch (loss 1.5994): 14%|█▍ | 9/63 [00:10<00:32, 1.68it/s] Training 1/1 epoch (loss 1.5994): 16%|█▌ | 10/63 [00:10<00:29, 1.82it/s] Training 1/1 epoch (loss 1.6340): 16%|█▌ | 10/63 [00:11<00:29, 1.82it/s] Training 1/1 epoch (loss 1.6340): 17%|█▋ | 11/63 [00:11<00:27, 1.91it/s] Training 1/1 epoch (loss 1.6695): 17%|█▋ | 11/63 [00:11<00:27, 1.91it/s] Training 1/1 epoch (loss 1.6695): 19%|█▉ | 12/63 [00:11<00:25, 1.98it/s] Training 1/1 epoch (loss 1.5639): 19%|█▉ | 12/63 [00:11<00:25, 1.98it/s] Training 1/1 epoch (loss 1.5639): 21%|██ | 13/63 [00:11<00:24, 2.05it/s] Training 1/1 epoch (loss 1.5918): 21%|██ | 13/63 [00:12<00:24, 2.05it/s] Training 1/1 epoch (loss 1.5918): 22%|██▏ | 14/63 [00:12<00:23, 2.10it/s] Training 1/1 epoch (loss 1.6043): 22%|██▏ | 14/63 [00:12<00:23, 2.10it/s] Training 1/1 epoch (loss 1.6043): 24%|██▍ | 15/63 [00:12<00:22, 2.14it/s] Training 1/1 epoch (loss 1.6275): 24%|██▍ | 15/63 [00:13<00:22, 2.14it/s] Training 1/1 epoch (loss 1.6275): 25%|██▌ | 16/63 [00:13<00:22, 2.09it/s] Training 1/1 epoch (loss 1.7586): 25%|██▌ | 16/63 [00:13<00:22, 2.09it/s] Training 1/1 epoch (loss 1.7586): 27%|██▋ | 17/63 [00:13<00:21, 2.14it/s] Training 1/1 epoch (loss 1.6914): 27%|██▋ | 17/63 [00:14<00:21, 2.14it/s] Training 1/1 epoch (loss 1.6914): 29%|██▊ | 18/63 [00:14<00:20, 2.16it/s] Training 1/1 epoch (loss 1.7466): 29%|██▊ | 18/63 [00:14<00:20, 2.16it/s] Training 1/1 epoch (loss 1.7466): 30%|███ | 19/63 [00:14<00:19, 2.20it/s] Training 1/1 epoch (loss 1.7239): 30%|███ | 19/63 [00:15<00:19, 2.20it/s] Training 1/1 epoch (loss 1.7239): 32%|███▏ | 20/63 [00:15<00:19, 2.22it/s] Training 1/1 epoch (loss 1.6923): 32%|███▏ | 20/63 [00:15<00:19, 2.22it/s] Training 1/1 epoch (loss 1.6923): 33%|███▎ | 21/63 [00:15<00:18, 2.22it/s] Training 1/1 epoch (loss 1.6782): 33%|███▎ | 21/63 [00:15<00:18, 2.22it/s] Training 1/1 epoch (loss 1.6782): 35%|███▍ | 22/63 [00:15<00:18, 2.22it/s] Training 1/1 epoch (loss 1.6388): 35%|███▍ | 22/63 [00:16<00:18, 2.22it/s] Training 1/1 epoch (loss 1.6388): 37%|███▋ | 23/63 [00:16<00:17, 2.23it/s] Training 1/1 epoch (loss 1.6102): 37%|███▋ | 23/63 [00:16<00:17, 2.23it/s] Training 1/1 epoch (loss 1.6102): 38%|███▊ | 24/63 [00:16<00:18, 2.15it/s] Training 1/1 epoch (loss 1.5856): 38%|███▊ | 24/63 [00:17<00:18, 2.15it/s] Training 1/1 epoch (loss 1.5856): 40%|███▉ | 25/63 [00:17<00:17, 2.19it/s] Training 1/1 epoch (loss 1.5694): 40%|███▉ | 25/63 [00:17<00:17, 2.19it/s] Training 1/1 epoch (loss 1.5694): 41%|████▏ | 26/63 [00:17<00:16, 2.21it/s] Training 1/1 epoch (loss 1.6984): 41%|████▏ | 26/63 [00:18<00:16, 2.21it/s] Training 1/1 epoch (loss 1.6984): 43%|████▎ | 27/63 [00:18<00:16, 2.21it/s] Training 1/1 epoch (loss 1.6798): 43%|████▎ | 27/63 [00:18<00:16, 2.21it/s] Training 1/1 epoch (loss 1.6798): 44%|████▍ | 28/63 [00:18<00:16, 2.15it/s] Training 1/1 epoch (loss 1.6722): 44%|████▍ | 28/63 [00:19<00:16, 2.15it/s] Training 1/1 epoch (loss 1.6722): 46%|████▌ | 29/63 [00:19<00:15, 2.18it/s] Training 1/1 epoch (loss 1.5383): 46%|████▌ | 29/63 [00:19<00:15, 2.18it/s] Training 1/1 epoch (loss 1.5383): 48%|████▊ | 30/63 [00:19<00:15, 2.18it/s] Training 1/1 epoch (loss 1.6506): 48%|████▊ | 30/63 [00:20<00:15, 2.18it/s] Training 1/1 epoch (loss 1.6506): 49%|████▉ | 31/63 [00:20<00:14, 2.20it/s] Training 1/1 epoch (loss 1.7083): 49%|████▉ | 31/63 [00:20<00:14, 2.20it/s] Training 1/1 epoch (loss 1.7083): 51%|█████ | 32/63 [00:20<00:14, 2.12it/s] Training 1/1 epoch (loss 1.6033): 51%|█████ | 32/63 [00:21<00:14, 2.12it/s] Training 1/1 epoch (loss 1.6033): 52%|█████▏ | 33/63 [00:21<00:14, 2.14it/s] Training 1/1 epoch (loss 1.5984): 52%|█████▏ | 33/63 [00:21<00:14, 2.14it/s] Training 1/1 epoch (loss 1.5984): 54%|█████▍ | 34/63 [00:21<00:13, 2.17it/s] Training 1/1 epoch (loss 1.5998): 54%|█████▍ | 34/63 [00:21<00:13, 2.17it/s] Training 1/1 epoch (loss 1.5998): 56%|█████▌ | 35/63 [00:21<00:12, 2.20it/s] Training 1/1 epoch (loss 1.6650): 56%|█████▌ | 35/63 [00:22<00:12, 2.20it/s] Training 1/1 epoch (loss 1.6650): 57%|█████▋ | 36/63 [00:22<00:12, 2.22it/s] Training 1/1 epoch (loss 1.5882): 57%|█████▋ | 36/63 [00:22<00:12, 2.22it/s] Training 1/1 epoch (loss 1.5882): 59%|█████▊ | 37/63 [00:22<00:11, 2.25it/s] Training 1/1 epoch (loss 1.5717): 59%|█████▊ | 37/63 [00:23<00:11, 2.25it/s] Training 1/1 epoch (loss 1.5717): 60%|██████ | 38/63 [00:23<00:11, 2.23it/s] Training 1/1 epoch (loss 1.6135): 60%|██████ | 38/63 [00:23<00:11, 2.23it/s] Training 1/1 epoch (loss 1.6135): 62%|██████▏ | 39/63 [00:23<00:10, 2.24it/s] Training 1/1 epoch (loss 1.6085): 62%|██████▏ | 39/63 [00:24<00:10, 2.24it/s] Training 1/1 epoch (loss 1.6085): 63%|██████▎ | 40/63 [00:24<00:10, 2.15it/s] Training 1/1 epoch (loss 1.5770): 63%|██████▎ | 40/63 [00:24<00:10, 2.15it/s] Training 1/1 epoch (loss 1.5770): 65%|██████▌ | 41/63 [00:24<00:10, 2.17it/s] Training 1/1 epoch (loss 1.5733): 65%|██████▌ | 41/63 [00:25<00:10, 2.17it/s] Training 1/1 epoch (loss 1.5733): 67%|██████▋ | 42/63 [00:25<00:09, 2.22it/s] Training 1/1 epoch (loss 1.6788): 67%|██████▋ | 42/63 [00:25<00:09, 2.22it/s] Training 1/1 epoch (loss 1.6788): 68%|██████▊ | 43/63 [00:25<00:09, 2.22it/s] Training 1/1 epoch (loss 1.5074): 68%|██████▊ | 43/63 [00:26<00:09, 2.22it/s] Training 1/1 epoch (loss 1.5074): 70%|██████▉ | 44/63 [00:26<00:08, 2.24it/s] Training 1/1 epoch (loss 1.6573): 70%|██████▉ | 44/63 [00:26<00:08, 2.24it/s] Training 1/1 epoch (loss 1.6573): 71%|███████▏ | 45/63 [00:26<00:08, 2.25it/s] Training 1/1 epoch (loss 1.5843): 71%|███████▏ | 45/63 [00:26<00:08, 2.25it/s] Training 1/1 epoch (loss 1.5843): 73%|███████▎ | 46/63 [00:26<00:07, 2.27it/s] Training 1/1 epoch (loss 1.6471): 73%|███████▎ | 46/63 [00:27<00:07, 2.27it/s] Training 1/1 epoch (loss 1.6471): 75%|███████▍ | 47/63 [00:27<00:06, 2.29it/s] Training 1/1 epoch (loss 1.5904): 75%|███████▍ | 47/63 [00:27<00:06, 2.29it/s] Training 1/1 epoch (loss 1.5904): 76%|███████▌ | 48/63 [00:27<00:06, 2.20it/s] Training 1/1 epoch (loss 1.6856): 76%|███████▌ | 48/63 [00:28<00:06, 2.20it/s] Training 1/1 epoch (loss 1.6856): 78%|███████▊ | 49/63 [00:28<00:06, 2.19it/s] Training 1/1 epoch (loss 1.4872): 78%|███████▊ | 49/63 [00:28<00:06, 2.19it/s] Training 1/1 epoch (loss 1.4872): 79%|███████▉ | 50/63 [00:28<00:05, 2.22it/s] Training 1/1 epoch (loss 1.5536): 79%|███████▉ | 50/63 [00:29<00:05, 2.22it/s] Training 1/1 epoch (loss 1.5536): 81%|████████ | 51/63 [00:29<00:05, 2.25it/s] Training 1/1 epoch (loss 1.5252): 81%|████████ | 51/63 [00:29<00:05, 2.25it/s] Training 1/1 epoch (loss 1.5252): 83%|████████▎ | 52/63 [00:29<00:04, 2.25it/s] Training 1/1 epoch (loss 1.5549): 83%|████████▎ | 52/63 [00:30<00:04, 2.25it/s] Training 1/1 epoch (loss 1.5549): 84%|████████▍ | 53/63 [00:30<00:04, 2.24it/s] Training 1/1 epoch (loss 1.7051): 84%|████████▍ | 53/63 [00:30<00:04, 2.24it/s] Training 1/1 epoch (loss 1.7051): 86%|████████▌ | 54/63 [00:30<00:04, 2.24it/s] Training 1/1 epoch (loss 1.6551): 86%|████████▌ | 54/63 [00:30<00:04, 2.24it/s] Training 1/1 epoch (loss 1.6551): 87%|████████▋ | 55/63 [00:30<00:03, 2.27it/s] Training 1/1 epoch (loss 1.5699): 87%|████████▋ | 55/63 [00:31<00:03, 2.27it/s] Training 1/1 epoch (loss 1.5699): 89%|████████▉ | 56/63 [00:31<00:03, 2.17it/s] Training 1/1 epoch (loss 1.4918): 89%|████████▉ | 56/63 [00:31<00:03, 2.17it/s] Training 1/1 epoch (loss 1.4918): 90%|█████████ | 57/63 [00:31<00:02, 2.21it/s] Training 1/1 epoch (loss 1.5779): 90%|█████████ | 57/63 [00:32<00:02, 2.21it/s] Training 1/1 epoch (loss 1.5779): 92%|█████████▏| 58/63 [00:32<00:02, 2.23it/s] Training 1/1 epoch (loss 1.6452): 92%|█████████▏| 58/63 [00:32<00:02, 2.23it/s] Training 1/1 epoch (loss 1.6452): 94%|█████████▎| 59/63 [00:32<00:01, 2.17it/s] Training 1/1 epoch (loss 1.6185): 94%|█████████▎| 59/63 [00:33<00:01, 2.17it/s] Training 1/1 epoch (loss 1.6185): 95%|█████████▌| 60/63 [00:33<00:01, 2.21it/s] Training 1/1 epoch (loss 1.5172): 95%|█████████▌| 60/63 [00:33<00:01, 2.21it/s] Training 1/1 epoch (loss 1.5172): 97%|█████████▋| 61/63 [00:33<00:00, 2.23it/s] Training 1/1 epoch (loss 1.6084): 97%|█████████▋| 61/63 [00:34<00:00, 2.23it/s] Training 1/1 epoch (loss 1.6084): 98%|█████████▊| 62/63 [00:34<00:00, 2.25it/s] Training 1/1 epoch (loss 1.5964): 98%|█████████▊| 62/63 [00:34<00:00, 2.25it/s] Training 1/1 epoch (loss 1.5964): 100%|██████████| 63/63 [00:34<00:00, 2.24it/s] Training 1/1 epoch (loss 1.5964): 100%|██████████| 63/63 [00:34<00:00, 1.82it/s]
tokenizer config file saved in /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k-Q2-2k/tokenizer_config.json
Special tokens file saved in /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k-Q2-2k/special_tokens_map.json
wandb:
wandb:
wandb: Run history:
wandb: train/epoch ▁▁▁▂▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▆▇▇████
wandb: train/loss ▄▅▆▂▃▄▅▆▃▄█▆█▇▅▄▃▆▆▆▇▄▄▃▄▆▂▅▄▅▁▃▂▃▇▃▅▄▂▄
wandb: train/lr ▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: train/step ▁▁▂▂▂▂▂▃▃▃▃▃▃▃▄▄▄▄▄▄▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇▇███
wandb:
wandb: Run summary:
wandb: train/epoch 1
wandb: train/loss 1.59638
wandb: train/lr 1e-05
wandb: train/step 63
wandb:
wandb: 🚀 View run qwen-7b-s3-Q1-40k-Q2-2k at: https://wandb.ai/xtom/Inverse_Alignment/runs/hrtj7fnx
wandb: ⭐️ View project at: https://wandb.ai/xtom/Inverse_Alignment
wandb: Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
wandb: Find logs at: /aifs4su/hansirui_1st/boyuan/resist/setting3-safety/Qwen1.5-7B/Qwen1.5-7B-s3-Q1-40k-Q2-2k/wandb/run-20250528_203628-hrtj7fnx/logs