// event · library

Hugging Face Transformers 5.18.0 adds Nemotron 3 Diarization, NemotronH Omni, HyperCLOVAX Vision V2, and GTE

library

vendor: huggingfacepublished: 12d agoconfidence: 0.90id: a5fa4fdf-3171-43fe-a737-41dabf5d3b2a

Transformers 5.18.0 adds four new model families: Nemotron 3 Diarization (streaming speaker diarization up to 8 speakers with configurable 80ms–30.4s latency), NemotronH Omni (Mamba-Transformer LM + RADIO vision + Parakeet audio for joint text/image/video/audio reasoning), HyperCLOVAX Vision V2 (NAVER multimodal VLM with think tokens), and GTE (RoPE BERT-style long-context multilingual embeddings backing gte-*-v1.5 and Snowflake arctic-embed-m-v2.0).

capability delta

an agent can now use Nemotron 3 Diarization for streaming speaker "who spoke when" detection, Nemotron H Omni for joint text/image/video/audio reasoning, HyperCLOVAX Vision V2, and GTE multilingual embeddings directly via Hugging Face Transformers.

curator note

Major Transformers minor release adding four new model families spanning diarization, omni-multimodal reasoning, vision-language, and embeddings.

// install

pip install transformers==5.18.0

pypi · transformers @ 5.18.0

// example

python

from transformers import AutoModel, AutoProcessor

# Streaming speaker diarization with Nemotron 3
model = AutoModel.from_pretrained("nvidia/nemotron3-diarization")
processor = AutoProcessor.from_pretrained("nvidia/nemotron3-diarization")

inputs = processor(audio_array, sampling_rate=16000, return_tensors="pt")
outputs = model(**inputs)
# outputs contain per-frame speaker labels ordered by arrival

// links

transformershuggingfacenemotrondiarizationmultimodalembeddingshyperclovaxgte

source: github_release · https://github.com/huggingface/transformers/releases/tag/v5.18.0