Comment by lukaesch

whisper + pyannote + custom LLM pipeline to reduce errors, extract and attach speaker labels