search

MusicLang / musiclang_predict

AI Prediction api of the MusicLang package
GNU General Public License v3.0
250 stars 16 forks source link

Audio Captioning #3

Closed zfarrell13 closed 2 months ago

zfarrell13 commented 6 months ago

this is an enhancement request rather than an issue, going to try to implement and work with this on my local machine today - but there is a tremendous use case for audio captioning. Based on what i can see via skimming your code, it seems like you all have the resources and data to build a captioning algorithm. would be really nice :)

floriangardin commented 6 months ago

Sounds very interesting. What's the specific use case you are thinking about, I don't fully get it ? You'll probably want to PR from the musiclang package directly as this one (musiclang_predict) is just the wrapper for prediction.

zfarrell13 commented 6 months ago

will do. use case is that for people building txt to audio, they will want to train their models on proprietary audio that hasnt been captioned yet and will need to apply captions before training.