search

meronym / speaker-transcription

Transcription with speaker diarization pipeline
MIT License
80 stars 17 forks source link

Allow specifying whisper model #4

Open arnab opened 1 year ago

arnab commented 1 year ago

Similar to #2 - it would be great if we could specify the whisper model to use: large/base etc.

Using smaller models would probably be fine for my workflow right now, and would save run-time (and cost) on Replicate.com.