Hyper parameters tuning for Speaker Diarization

[SrinivasKalagotla]

Hi,
I want to know how to tune the Speaker Diarization model. I saw a piece of code from the huggingface website in which hyper parameters are instantiated.

I have tried the same way for speaker diarization and it throwed an error :
ValueError: parameter 'onset' does not exist

Can you please let me know how to tune the Speaker Diarization.

Thank you

[hbredin]

There is a whole tutorial in the documentation about that.

[SrinivasKalagotla]

Thank you for the response.
I have a 120 seconds audio file with 3 speakers. I ran the Speaker Diarization on it and it is giving results with more than 120 seconds. Can you please explain why it is happening.

I think there is some issue with overlapping speech.