---
library_name: pytorch
tags:
  - audio
  - speech
  - soundstream
datasets:
  - librispeech_asr
---

# SoundStream Codec

SoundStream-style speech codec checkpoint trained on LibriSpeech at 16 kHz mono.

Training details and code will be available on GitHub:
https://github.com/daminovkamil/soundstream-codec/tree/main