pyannote-v3-segmentation-GGUF

Pyannote Segmentation 3.0 — GGUF

Native GGUF port of pyannote/segmentation-3.0 for speaker diarization.

Model details

The model performs joint voice-activity detection, speaker segmentation, and overlapped-speech detection on short audio chunks. Downstream clustering then produces full-file speaker diarization.

Usage with CrispASR

crispasr \
  --diarize-method pyannote \
  --sherpa-segment-model pyannote-seg-3.0.gguf \
  audio.wav

Provenance

Weights were exported directly from the original PyTorch checkpoint (pyannote/segmentation-3.0) into GGUF format, preserving full F32 precision across all 41 tensors.

License

MIT — same as the original pyannote-audio segmentation-3.0 model.

Provenance and EU AI Act Art. 53 note

• Upstream model: pyannote/segmentation-3.0 — published by pyannote.
• Upstream licence: mit. This repository redistributes under the same terms; it grants no rights the upstream licence does not.
• What was done here: format conversion and/or quantisation only (GGUF/GGML). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs.
• Training data: documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository.
• Provider status: under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream.