Spaces:

aadnk

/

whisper-webui

like 343

Running

Hello, thank you for your project.

I noticed a slow diarization process.
I can transcribe a 40 minute audio in 2-3 minutes with medium model without diarization using parallel true and a 3060 gpu
if i check diarization on the same file it took over 40 minutes.
There is something i can try to reduce this time?
Thank you again