lucasnewman/vocos-mlx

Implementation of 'Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis', in MLX

33
/ 100
Emerging

Bridges time-domain and Fourier-based vocoding by reconstructing high-fidelity audio from either Mel spectrograms or EnCodec tokens, with dual inference pathways for flexible audio generation. Built on MLX framework optimized for Apple Silicon, supporting bandwidth-controlled EnCodec decoding and pretrained model checkpoints for 24kHz audio synthesis. Integrates with EnCodec for neural compression-based audio conditioning alongside traditional spectrogram-based approaches.

Used by 2 other packages. No commits in the last 6 months. Available on PyPI.

Stale 6m
Maintenance 0 / 25
Adoption 8 / 25
Maturity 18 / 25
Community 7 / 25

How are scores calculated?

Stars

24

Forks

2

Language

Python

License

MIT

Last pushed

Oct 30, 2024

Commits (30d)

0

Dependencies

4

Reverse dependents

2

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/lucasnewman/vocos-mlx"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.