NVIDIA Releases Nemotron 3 Diarization Model for Overlapping Speakers

NVIDIA's Nemotron 3 Diarization model (100M parameters) tracks who speaks when for up to 8 speakers, handling overlapping speech. Available on Hugging Face.

NVIDIA's Nemotron 3 Diarization model (100M parameters) tracks who speaks when for up to 8 speakers, handling overlapping speech. Available on Hugging Face.