hibiki-2b (Vokra GGUF)

Converted to the Vokra GGUF format for Vokra, a zero-dependency speech-AI inference runtime.

This is a conversion, not a new model. The weights are the upstream ones; Vokra re-packages them so its runtime can memory-map them directly. Credit for the model belongs upstream — see Source below.

Files

File	Size	SHA-256
hibiki-2b.gguf	5316.5 MB	62e70efbb54a612f51e3d7d35c9d901806331b642474d6e0857a73252aaf8934

Usage

# Download (any HTTP client works — the file is a plain GGUF)
curl -L -o hibiki-2b.gguf \
  https://huggingface.co/vokra/hibiki-2b/resolve/main/hibiki-2b.gguf

vokra-cli run --model hibiki-2b.gguf --input input.wav

Provenance

Field	Value
Architecture	hibiki
Tensors	327
Upstream source	kyutai/hibiki-2b-pytorch-bf16 (simultaneous FR↔EN S2S translation, cc-by-4.0)
Upstream licence	cc-by-4.0
Licence class	attribution-required
Registry model id	hibiki-2b
Vokra GGUF schema	1
Converted by	vokra-core 0.1.0-alpha.0

Every row above is read out of this file's own vokra.* metadata, so the card cannot claim something the artifact does not carry.

Licence

The weights are distributed under cc-by-4.0, unchanged from upstream. Conversion does not alter the licence, and your obligations run to the upstream author.

Attribution required

This application uses the Kyutai Hibiki-2B model (simultaneous speech-to-speech translation, Helium + Mimi codec). Model weights are licensed under CC-BY 4.0 (attribution required; commercial use permitted). Copyright (c) Kyutai. Source: https://huggingface.co/kyutai/hibiki-2b-pytorch-bf16

This licence obliges you to display the attribution above when you ship something built on these weights. Vokra surfaces it at runtime via vokra_model_attribution (C ABI) and a CLI banner.

Verifying this file

shasum -a 256 hibiki-2b.gguf
# expect: 62e70efbb54a612f51e3d7d35c9d901806331b642474d6e0857a73252aaf8934