How can i find the training details with pyannote-audio?

[xuanjihe]

Hi, I trained an overlap speech detection with my own data by pyannote-audio(version: 2.0) and get outstanding performance. Now I want to learn more about the details of the training, but I can't find the entry of the training script, can you tell me where the function entry corresponding to the bin of pyannote-audio is, so that I can better understand the training process?

[hbredin]

Thanks for the positive feedback!

Would you mind sharing which command(s) you ran? Difficult to answer without this information.

[xuanjihe]

I use pyannote-audio to train an overlap speech detection model on my own data, and I run the command:
pyannote-audio ovl train --subset=train --to=1200 --parallel=20 --gpu --from 100 alimeet/ovd AMI.SpeakerDiarization.MixHeadset