How to use our own VAD or Speaker Embedding models in speaker diarization?

[yerzhan7orazayev]

Describe your question
Is it possible to use our own VAD or Speaker Embedding models in speaker diarization? If so, how one can do it properly?
Many thanks.

[fayejf]

Thanks for your question.
Of course! You can change the diarizer.vad.model_path and diarizer.speaker_embeddings.model_path in config file to use your own models.