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

[idefics] fix vision's hidden_act #25787

Merged
merged 1 commit into from
Aug 28, 2023
Merged

[idefics] fix vision's hidden_act #25787

merged 1 commit into from
Aug 28, 2023

Conversation

stas00
Copy link
Contributor

@stas00 stas00 commented Aug 27, 2023

Thanks to @rwightman's discovery this PR is fixing vision config's hidden_act to gelu.

It looks like we messed things up when splitting the original config into 3 groups during the porting and inherited clip's default config. Whereas the model used during training was using gelu as can be seen here. https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K/blob/main/config.json

Thank you, @rwightman

@stas00 stas00 requested a review from rwightman August 27, 2023 21:56
@stas00 stas00 marked this pull request as ready for review August 27, 2023 21:56
@HuggingFaceDocBuilderDev
Copy link

HuggingFaceDocBuilderDev commented Aug 27, 2023

The documentation is not available anymore as the PR was closed or merged.

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! Think the value online should also be specified

@stas00 stas00 merged commit 50573c6 into main Aug 28, 2023
3 checks passed
@stas00 stas00 deleted the idefics-fixes-2 branch August 28, 2023 14:37
Narsil pushed a commit to huggingface/text-generation-inference that referenced this pull request Sep 1, 2023
parambharat pushed a commit to parambharat/transformers that referenced this pull request Sep 26, 2023
[idefics] fix vision's hidden_act
verdant621 pushed a commit to verdant621/text-generation-inference that referenced this pull request Oct 19, 2023
blbadger pushed a commit to blbadger/transformers that referenced this pull request Nov 8, 2023
[idefics] fix vision's hidden_act
EduardoPach pushed a commit to EduardoPach/transformers that referenced this pull request Nov 18, 2023
[idefics] fix vision's hidden_act
cr313 added a commit to cr313/text-generation-inference-load-test that referenced this pull request Apr 19, 2024
alfredgui2 pushed a commit to mlsys-io/kv.run that referenced this pull request Jul 6, 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.

4 participants