Is transcription with JAX-able models and subsequent labeling with whisper-at possible?
Or does the transcription result need to be from whisper-at? Im wondering because of how fast whisper-jax is, I could use it to transcribe but use your model to label (speech, laughter ect.) afterward.
Is transcription with JAX-able models and subsequent labeling with whisper-at possible? Or does the transcription result need to be from whisper-at? Im wondering because of how fast whisper-jax is, I could use it to transcribe but use your model to label (speech, laughter ect.) afterward.