(default: openai/whisper-large-v3) --task {transcribe,translate ... audio clips --diarization_model DIARIZATION_MODEL Name of the pretrained model/ checkpoint to perform diarization. (default: ...