Key Info
NVIDIA has open-sourced Nemotron 3 Diarization, a model that can reliably track speakers in live conversations under a commercial-friendly license, with day-zero integration with Transformers.
Highlights
- Model tracks speakers in real-time, addressing a key gap in voice agents.
- Tested with one-second speech chunks, showing good quality.
- Includes day-zero integration with Hugging Face Transformers.
- Demonstrated on Reachy Mini and DGX Spark.