search

bugbakery / audapolis

an editor for spoken-word audio with automatic transcription
GNU Affero General Public License v3.0
1.69k stars 40 forks source link

đŸĒŠ add whisper support #467

Open anuejn opened 1 year ago

anuejn commented 1 year ago

This is an experiment for adding support for the whisper models for improved accuracy and one step transcription with punctuation reconstruction. Still in an early stage and only model downloading and not inference is implemented.

ThiloteE commented 7 months ago

If I may ask, what's blocking this from merging?

pajowu commented 7 months ago

The inference part is still missing from what I can tell from a quick look at the code (i.e. the part that actually uses whisper to transcribe something)

ThiloteE commented 7 months ago

Closes #416