huggingface/transformers v5.18.0: Release 5.18.0
Nemotron 3 Diarization is an open-weight streaming speaker diarization model designed to determine "who spoke when" in real-world audio.
ProofCode ↗
Key points
- A single checkpoint supports configurable latency profiles, from an 80 ms input buffer to a 30.4 s offline-style buffer, and configurable output frame resolution in multiples of 10 ms.
- With chunked inference, the maximum audio duration is not limited.
- embedding space at the <image> / <video> context-token positions; audio clips are projected in the same way at
- HyperCLOVAX Vision V2 is a multimodal vision-language model developed by NAVER.
Sources (1)
- [1]huggingface/transformers v5.18.0: Release 5.18.0GitHub: huggingface/transformers · Sep 30, 04:46 PM
Nemotron 3 Diarization is an open-weight streaming speaker diarization model designed to determine "who spoke when" in real-world audio.
A single checkpoint supports configurable latency profiles, from an 80 ms input buffer to a 30.4 s offline-style buffer, and configurable output frame resolution in multiples of 10 ms.
Extractive summary: sentences quoted from the sources.
Before this
- Sep 9, 2026huggingface/transformers v5.17.0: Release 5.17.0
- Sep 9, 2026vllm-project/vllm v0.29.0
- Aug 22, 2026sgl-project/sglang v0.5.18
- Jul 10, 2026sgl-project/sglang v0.5.15
- Jul 3, 2026huggingface/transformers v5.13.0: Release v5.13.0
- Jun 10, 2026huggingface/transformers v5.11.0: Release v5.11.0
