Skip to content

Superpoint imports fix#29898

Merged
gante merged 1 commit intohuggingface:mainfrom
zucchini-nlp:quick_fix
Apr 3, 2024
Merged

Superpoint imports fix#29898
gante merged 1 commit intohuggingface:mainfrom
zucchini-nlp:quick_fix

Conversation

@zucchini-nlp
Copy link
Member

What does this PR do?

I have been seeing this errors for a few days (when make quality) and seems like there is no PR fixing it. I decided to make it a separate PR, to merge quickly independently of other changes.

@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.

That's a bit weird, ci's are green on main. Locally I don't have an issue so most probably a version of ruff?

@zucchini-nlp
Copy link
Member Author

I noticed now that it does not fail in every make quality. Guess that depends on what local changes were made 🤔

The bug here is that the import is done from incorrect place, because the requires_backend should be imported from utils. It is already done so for all image processors in library. I am not sure how the Superpoint model tests passed when adding the model, but currently trying to get the processor as below fails in main

processor = AutoImageProcessor.from_pretrained("magic-leap-community/superpoint")

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 for explaining!

@zucchini-nlp
Copy link
Member Author

@gante this can be merged

@gante gante merged commit c10b5dd into huggingface:main Apr 3, 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