A user should be able to request speaker diarization without using a HF token, to allow for self-hosted model usage.
A user may pass an argument to force speaker diarization.
A user may pass an additional argument if speaker diarization is forced, which specifies whether or not the script will use HF or their own self-hosted model.
A user should be able to request speaker diarization without using a HF token, to allow for self-hosted model usage.
A user may pass an argument to force speaker diarization.
A user may pass an additional argument if speaker diarization is forced, which specifies whether or not the script will use HF or their own self-hosted model.
See https://github.com/pyannote/pyannote-audio/blob/develop/tutorials/community/offline_usage_speaker_diarization.ipynb for reference.