GanjinZero / RRHF

[NIPS2023] RRHF & Wombat
780 stars 49 forks source link

Error when try to inference #12

Closed oasis-0927 closed 1 year ago

oasis-0927 commented 1 year ago

Traceback (most recent call last): File "single_sentence_inference.py", line 7, in from stanford_alpaca.train import smart_tokenizer_and_embedding_resize, DEFAULT_PAD_TOKEN, DEFAULT_EOS_TOKEN, DEFAULT_BOS_TOKEN, DEFAULT_UNK_TOKEN ModuleNotFoundError: No module named 'stanford_alpaca'

Seems like stanford_alpaca should be cloned into the directory .

GanjinZero commented 1 year ago

You are correct. I will fix it.

GanjinZero commented 1 year ago

Fix it. Please check if it worked.

oasis-0927 commented 1 year ago

Thanks for the quick response and yes it worked. Another thing, I suppose Chinese is not currently supported? Since I got this really weird response when try to get a Chinese answer.

image
GanjinZero commented 1 year ago

Wombat-7B is not trained by Chinese. We will train a Wombat-7B-Chinese and release it later.

oasis-0927 commented 1 year ago

Got it. Thanks!