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

Llama family, fix use_cache=False generation #30380

Merged
merged 4 commits into from
Apr 22, 2024
Merged

Conversation

ArthurZucker
Copy link
Collaborator

What does this PR do?

fixes #30347

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

@ArthurZucker ArthurZucker requested a review from gante April 22, 2024 08:54
Copy link
Member

@gante gante left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, thank you for fixing! 💛

Suggestion: add a fast llama + generate + no cache test

@ArthurZucker
Copy link
Collaborator Author

Yeah, but it feels like this should be part of the generation mixin tests (and AFAIR it is?)

@ArthurZucker ArthurZucker merged commit 2d92db8 into main Apr 22, 2024
17 of 19 checks passed
@ArthurZucker ArthurZucker deleted the fix-llama-no-cache branch April 22, 2024 12:42
zafstojano pushed a commit to zafstojano/transformers that referenced this pull request Apr 22, 2024
* nit to make sure cache positions are not sliced

* fix other models

* nit

* style
itazap pushed a commit that referenced this pull request May 14, 2024
* nit to make sure cache positions are not sliced

* fix other models

* nit

* style
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

Successfully merging this pull request may close these issues.

generate method does not work with generation_config.use_cache=False in Llama-2 model
3 participants