ljy0ustc / LLaRA

Apache License 2.0
94 stars 7 forks source link

About The Train #4

Open Albertchamberlain opened 8 months ago

Albertchamberlain commented 8 months ago

image

Will you help me to solve this problem ? 👆

Albertchamberlain commented 8 months ago

Waiting for the checkpoints😀

ljy0ustc commented 8 months ago

Hi~ Thanks for your attention to our work. The termination of the process doesn't seem to be due to a lack of checkpoints, as the training process doesn't require LLaRA checkpoints, but only needs the Hugging Face model of LLaMA2-7B. Could you provide more information about your configuration of the model path for the LLaMA-2 7B Hugging Face version?

Albertchamberlain commented 8 months ago

Hi~ Thanks for your attention to our work.嗨~感谢您对我们工作的关注。 The termination of the process doesn't seem to be due to a lack of checkpoints, as the training process doesn't require LLaRA checkpoints, but only needs the Hugging Face model of LLaMA2-7B. Could you provide more information about your configuration of the model path for the LLaMA-2 7B Hugging Face version?该过程的终止似乎不是由于缺乏检查点,因为训练过程不需要LLaRA检查点,只需要LLaMA2-7B的Hugging Face模型。您能提供有关LLaMA-2 7B Hugging Face版本模型路径配置的更多信息吗?

Thanks image

Albertchamberlain commented 8 months ago

It's seem‘s like no problem about the LLaMA-2 7B Hugging Face config path or version It‘s just be killed😂

ljy0ustc commented 8 months ago

It seems your program was killed while executing the line of code self.llama_model = LlamaForCausalLM.from_pretrained(llm_path, torch_dtype=torch.bfloat16) in mode/model_interface.py. Perhaps you could try directly loading the LLaMA-2 7B Hugging Face version model under your conda environment and see if it can be loaded normally.

Albertchamberlain commented 8 months ago

It seems your program was killed while executing the line of code self.llama_model = LlamaForCausalLM.from_pretrained(llm_path, torch_dtype=torch.bfloat16) in mode/model_interface.py. Perhaps you could try directly loading the LLaMA-2 7B Hugging Face version model under your conda environment and see if it can be loaded normally.似乎您的程序在执行 mode/model_interface.py 中的代码行 self.llama_model = LlamaForCausalLM.from_pretrained(llm_path, torch_dtype=torch.bfloat16) 时被终止。也许您可以尝试在您的 conda 环境下直接加载 LLaMA-2 7B Hugging Face 版本模型,看看是否可以正常加载。

Ok,got it,i will try later,Thx a lot !!!

Albertchamberlain commented 8 months ago

Yeah, as you said, I reset the LLaMA-2 7B Path in another machine. It's work image

Albertchamberlain commented 8 months ago

Hello,another Q,How can i get this file?

image

I have download the dataset image

lavinal712 commented 7 months ago

Yeah, as you said, I reset the LLaMA-2 7B Path in another machine. It's work image

Hello, I meet some problem about this code. Did you run the code successfully?