#audio #machine-learning #translation #model #real-time #metal #gen

app hibiki

Hibiki, a real-time speech-to-speech translation model

3 releases

new 0.1.2 Feb 9, 2025
0.1.1 Feb 9, 2025
0.1.0 Feb 6, 2025

#809 in Machine learning

Download history 91/week @ 2025-02-01 246/week @ 2025-02-08

337 downloads per month

MIT/Apache

32KB
470 lines

hibiki - rust

rust ci badge Latest version Documentation License

See the top-level README.md for more information.

This provides the Rust implementation for Hibiki, a real-time speech-to-speech translation model.

Requirements

You will need a recent version of the Rust toolchain. To compile GPU support, you will also need the CUDA properly installed for your GPU, in particular with nvcc.

Example

cd hibiki-rs
wget https://github.com/kyutai-labs/moshi/raw/refs/heads/main/data/sample_fr_hibiki_crepes.mp3
cargo run  --features metal -r -- gen sample_fr_hibiki_crepes.mp3 out_en.wav

License

The present code is provided under the Apache license.

Dependencies

~26–43MB
~809K SLoC