Create speech to text model using python

I want a speech to text model created using python - more specifically using IBM software. It needs to be accurate and should be capable of speech diarization i.e. should identify each individual speaker and at what times they spoke. It should be highly accurate expecially when two or more people are speaking at the same time.