Add Audio Feature Extraction Task

[ecyht2]

Currently, the Feature Extraction task includes both models for audio and text feature extraction (it is officially placed under the NLP modality). I think it would be nice to have a new task for Audio Feature Extraction task just like Image Feature Extraction to better label models.

Some Audio Feature Extraction models in Feature Extractions:

• facebook/encodec
• kyutai/mimi

[osanseviero]

Yes, having an audio-feature-extraction has been discussed few times in the past and there's agreement having it could make sense. cc @Vaibhavs10 @pcuenca

[merveenoyan]

I think it makes sense to have given we have a separate one for image too @Vaibhavs10

[Vaibhavs10]

Hey @ecyht2 - sorry for the delay in responding to you - this is in the works, we're in the process of adding this as a recognised task in transformers and then follow it up with a task page. Cheers!