[Narsil]

What does this PR do?

Fixes # (issue)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[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.

[ydshieh]

run-slow: bert, gpt2, vit, whisper, llama, t5, siglip, bart, clip

[github-actions]

This comment contains run-slow, running the specified jobs:

models: ['models/bart', 'models/bert', 'models/clip', 'models/gpt2', 'models/llama', 'models/siglip', 'models/t5', 'models/vit', 'models/whisper']
quantizations: [] ...

[ydshieh]

There are several failures which seems to new from this PR

bert

https://github.com/huggingface/transformers/actions/runs/15680114584/job/44169418138

vit

https://github.com/huggingface/transformers/actions/runs/15680114584/job/44169418172

actually, all the models I triggered have this failures

[Narsil]

run-slow: bert, gpt2, vit, whisper, llama, t5, siglip, bart, clip

[ydshieh]

run-slow: bert, gpt2, vit, whisper, llama, t5, siglip, bart, clip

[github-actions]

This comment contains run-slow, running the specified jobs:

models: ['models/bart', 'models/bert', 'models/clip', 'models/gpt2', 'models/llama', 'models/siglip', 'models/t5', 'models/vit', 'models/whisper']
quantizations: [] ...

[ydshieh]

the failing tests match the results of main's results, so good

[ArthurZucker]

Thanks for the PR, and sorry it didn't make it in! This branch is based on v4, and v5 is very different, so rather than rebasing we'd love to invite you to open a new PR against main using our latest tools and good practices:

• Transformers tenets
• mlinter and the modeling rules it enforces (make typing)
• Modular transformers
• Contributing guide

Closing this one in the meantime, thanks again!