Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Strikethrough text (using a V100 graphics card) #27

Closed
Petilia opened this issue Apr 18, 2023 · 3 comments
Closed

Strikethrough text (using a V100 graphics card) #27

Petilia opened this issue Apr 18, 2023 · 3 comments

Comments

@Petilia
Copy link

Petilia commented Apr 18, 2023

I am using a V100 graphics card (it does not support int8 operations, when I try to run low_resource=True there is an error with cuda_setup).

I put low_resource=False and run it, but I get crossed out and wrong text

I used the right weights (git-lfs clone https://huggingface.co/lmsys/vicuna-13b-delta-v0)
image
image

@Andy1621
Copy link

I also met the same issue...

@stevezhang88
Copy link

https://huggingface.co/lmsys/vicuna-13b-delta-v0) . I think this weight should be merged with LLaMa 13B weights. Then works .

@TsuTikgiau
Copy link
Collaborator

Thanks for your interest in our work! People in this issue show similar problems to your case, and their reason looks like they don't add the delta weight of vicuna back to llama weight to get the final working weight. Note that the vicuna weight you directly download is a delta version. And you need to follow their instruction to add this delta weight back to the original llama weight, to get the final working weight of vicuna. The reason why we need this complex process is that llama doesn't allow us to distribute their weights, so vicuna has to release a delta weight version to bypass this rule

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants