Real Time Whisper Transcription and Diarizations

run with --model_path ./models --model medium --max_speakers_number 2 --language it