hiyouga/FastEdit

Is there any way to apply this interesting algorithm to the chatGLM-6B or chatGLM2-6B models?

rayoa opened this issue · 2 comments

rayoa commented

thx

hiyouga commented

We have already validated the performance of model-editing on the ChatGLM models. Unfortunately, the effectiveness is extremely unsatisfactory. We will make our efforts to tackle this problem in the future.

  • 👍6
datalee commented

need too，thks