Speaker Diarization to parakeet v3 and nemotron-3.5-asr-streaming-0.6b

**Is your feature request related to a problem? Please describe.**

Currently there is no Speaker Diarization in any of the mentioned models, meaning that you cannot differentiate between voices.

**Describe the solution you'd like**

Add Speaker Diarization support

**Describe alternatives you've considered**

I heard that you can pipeline it? I am unsure


