Omar Sanseviero

Please tag PRs so they are nicely linked: https://github.com/huggingface/huggingface.js/pull/614

Hey all! It might be good idea to sync about this topic as we're discussing with @julien-c @gary149 and @SBrandeis the unification of input/output expectations per library/task pair as well...

cc @sanchit-gandhi and @Vaibhavs10 for our audio experts :)

Sorry for my confusion, I read too quickly and did string matching with ASR :smiling_face_with_tear: Yes, this is indeed vision, In the past, OCR models have been tagged as `image-to-text`...

This looks simple for text but becomes trickier for audio/images or tasks such as zero-shot classification. If we do this, I would try to reuse the work done in https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/src/snippets/python.ts

Is this supported in the `pipeline` class already @NielsRogge ?

Cc @mishig25 and @OlivierDehaene maybe can help here (adding the slider in the widget and adding support in the inference API for this)

I think the reason is because it's not in https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/src/const.ts . We need to add this there. (cc @LysandreJik for vis)

Actually it's because there is no code snippet for it - https://github.com/huggingface/huggingface.js/blob/main/packages/tasks/src/snippets/python.ts

Hi there! I don't have a strong opinion, so I would love to hear what @merveenoyan has to say. In general, we always welcome new tasks given that: - They...