audiobox-aesthetics (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
audiobox-aesthetics.gguf	396.2 MB	1fbccc854cfcae56b5554ab4b8d67d8cc25b9c793821f8e738cbbace40b0e9dd

Usage

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

vokra-cli run --model audiobox-aesthetics.gguf --input input.wav

Provenance

Field	Value
Architecture	audiobox-aesthetics
Tensors	324
Upstream source	facebook/audiobox-aesthetics (wav2vec2 backbone + 5-dim quality regression, cc-by-4.0)
Upstream licence	cc-by-4.0
Licence class	attribution-required
Registry model id	audiobox-aesthetics
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 Audiobox Aesthetics model (wav2vec2 SSL backbone + 5-layer projection MLP head predicting BALANCED / CONTENT_ENJOYMENT / CONTENT_USEFULNESS / PRODUCTION_COMPLEXITY / PRODUCTION_QUALITY audio-quality axes; arXiv:2502.05139). Model weights are licensed under CC-BY 4.0 (attribution required; commercial use permitted). Copyright (c) Meta / Facebook AI Research. Source: https://github.com/facebookresearch/audiobox-aesthetics / https://huggingface.co/facebook/audiobox-aesthetics

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 audiobox-aesthetics.gguf
# expect: 1fbccc854cfcae56b5554ab4b8d67d8cc25b9c793821f8e738cbbace40b0e9dd