obtain timestamps in transcription mode
#35
by
hanifanggawi
- opened
is it possible to obtain timestamps in transcription mode when using .apply_transcription_request? it seems to only return the transcription text
Interested in the answer as well.
From the technical report, it could be that the audio is pre-segmented and individually transcribed.
hanifanggawi
changed discussion title from
obatain timestamps in transcription mode
to obtain timestamps in transcription mode
Yeah, I'm wondering the same. How does it work?
Any updates in this? I am trying to realise speaker diarization witch pyannote 3.1. it works well in API call Transkription but on local inference with vllm i cannot achieve this.