Separating different speaking from audio file

[rikabi89]

Hello all,

I not sure if this is possible. But my use case in relation to dataset building of TTS models. If I have .wav which contains audio for two different speakers. Is it possible to separate to an accurate degree the audio into two batches depending on the ID of the speaker?

I've been trying find out my self and with the help with GPT I got this : https://github.com/rikabi89/diarization_script/blob/main/diarization_script.py
However the issue I found was that there was a lot of overlapping and this was not accurate. ps I don't know any coding and I would appreciate any steer here.