{
  "model_type": "asr_with_diarization",
  "min_speakers": 2,
  "max_speakers": 5
}