LLM-Tuning
LLM-Tuning copied to clipboard
beyondguo
Tuning LLMs with no tears💦, sharing LLM-tools with love❤️.
Using pad_token, but it is not set yet. Loading base model for ppo training... 加载base 加载lora 加载ppo WARNING:root:A model is loaded from '/root/autodl-tmp/LLM/weights/sft_lora', and no v_head weight is found. This...
In the ```predict.py``` file, the following line shows an error: ```from my_templates import generate_prompt``` where I can find this ```my_templates``` module? How would I get it installed?
File "baichuan_lora_tuning.py", line 142, in main() File "baichuan_lora_tuning.py", line 134, in main trainer.train() File "/data01/conda_pkg/envs/baichuan/lib/python3.8/site-packages/transformers/trainer.py", line 1664, in train return inner_training_loop( File "/data01/conda_pkg/envs/baichuan/lib/python3.8/site-packages/transformers/trainer.py", line 1940, in _inner_training_loop tr_loss_step = self.training_step(model,...
你好， 我现在在训练PPO的时候出现了CUDA out of memory的问题，我是用了8个A100 GPUs，每一个GPU有80GB显存。下面是我运行的命令。我是用的stanford-alpaca提供的代码用8个A100全参微调训练了llama2-13B的sft model，reward模型是用的LLM-tuning项目提供的训练reward的代码基于llama2-13B训练的。现在就是在运行下面的ppo的时候出现了爆显存的问题，请问有什么办法可以降低显存吗？谢谢 ``` CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python rl_training.py \ --base_model_name /data/qbao775/Explanation-Generation/llama-2/llama-2-13B \ --merged_sft_model_path /data/qbao775/Explanation-Generation/llama_2_13B_merged_all_generator_avg_3_lenexp_10 \ --sft_model_lora_path /data/qbao775/Explanation-Generation/llama_2_13B_merged_all_generator_avg_3_lenexp_10 \ --reward_model_lora_path ../weights/llama-2-13B_beyond_reward_chinese_5000_peft_last_checkpoint \ --adafactor False \ --save_freq 10 \...
请问怎么改动代码，使用低精度模型，显卡不支持 Log如下 '\nlen(dataset)=1\n' 'loading init model...' Failed to load cpm_kernels:No module named 'cpm_kernels' Traceback (most recent call last): File "LLM-Tuning/chatglm2_lora_tuning.py", line 172, in main() File "LLM-Tuning/chatglm2_lora_tuning.py", line 98, in main...
请问有baichuan-13b的modeling_baichuan_for_cls.py吗？ baichuan-13b和baichuan7b模型结构有些调整，直接基于7b的cls.py文件训练reward model会模型参数不一致的问题 感谢~ 
╭───────────────────────── Traceback (most recent call last) ──────────────────────────╮ │ /Project/lq_ChatGLM2-6B/LLM-Tuning/chatglm2_lora_tuning.py:172 in │ │ │ │ 169 │ │ 170 │ │ 171 if __name__ == "__main__": │ │ ❱ 172 │...
你好 博主！ 我使用lora进行微调ChatGLM2-6B会出现这样的错误 但是使用ptuning不会，你知道是什么原因吗，求解。 
Traceback (most recent call last): File "/data3/push_recall/LLM-Tuning/chatglm2_lora_tuning.py", line 153, in main() File "/data3/push_recall/LLM-Tuning/chatglm2_lora_tuning.py", line 93, in main model = AutoModel.from_pretrained( File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/models/auto/auto_factory.py", line 462, in from_pretrained return model_class.from_pretrained( File "/data3/env/miniconda3/envs/baichuan/lib/python3.9/site-packages/transformers/modeling_utils.py",...
