WU Junyan comments

Repositories
Issues
Comments

Results 3 comments of


                                            WU Junyan

step3 answer is not correct

I met the same error, and I found that the ema model has the same weight as my SFT model while actor model saved well with different weights I guess...

Resizing model embedding when loading the model

we usually resize the model when using add speical tokens

hi, could you provide a code like llama3?

Maybe you are looking for [https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite/blob/main/modeling_deepseek.py](https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite/blob/main/modeling_deepseek.py)