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

Gemma2: fix FA2 generation #32553

Merged
merged 2 commits into from
Aug 9, 2024
Merged

Conversation

zucchini-nlp
Copy link
Member

@zucchini-nlp zucchini-nlp commented Aug 9, 2024

What does this PR do?

As reported in #32309 (comment) Gemma2 FA2 is broken in the latest release due to applying 4D mask when preparing inputs. FA2 expects a 2D mask so we can skip that step if FA2 is set

Ran the FA2 generate tests, everything passing, except for the Phi3 which is already fixed in another PR for patch release

@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.

Copy link
Collaborator

@ArthurZucker ArthurZucker left a comment

Choose a reason for hiding this comment

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

thanks, I don't remember where this is from but we need a patch!

@zucchini-nlp
Copy link
Member Author

Yes, I'll add link to this PR in the internal slack thread for patches :)

@zucchini-nlp zucchini-nlp merged commit 838d141 into huggingface:main Aug 9, 2024
21 checks passed
@ArthurZucker
Copy link
Collaborator

Already added 😉

ArthurZucker pushed a commit that referenced this pull request Aug 16, 2024
ArthurZucker pushed a commit that referenced this pull request Aug 20, 2024
ArthurZucker pushed a commit that referenced this pull request Aug 20, 2024
ArthurZucker pushed a commit that referenced this pull request Aug 20, 2024
stevhliu pushed a commit to stevhliu/transformers that referenced this pull request Aug 21, 2024
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.

3 participants