SNAC 🍿

Multi-Scale Neural Audio Codec (SNAC) compressess audio into discrete codes at a low bitrate.

👉 This model was primarily trained on music data, and its recommended use case is music (and SFX) generation. See below for other pretrained models.

🔗 GitHub repository: https://github.com/hubertsiuzdak/snac/

Overview

SNAC encodes audio into hierarchical tokens similarly to SoundStream, EnCodec, and DAC. However, SNAC introduces a simple change where coarse tokens are sampled less frequently, covering a broader time span.

This model compresses 44 kHz audio into discrete codes at a 2.6 kbps bitrate. It uses 4 RVQ levels with token rates of 14, 29, 57, and 115 Hz.

Pretrained models

Currently, all models support only single audio channel (mono).

Model	Bitrate	Sample Rate	Params	Recommended use case
hubertsiuzdak/snac_24khz	0.98 kbps	24 kHz	19.8 M	🗣️ Speech
hubertsiuzdak/snac_32khz	1.9 kbps	32 kHz	54.5 M	🎸 Music / Sound Effects
hubertsiuzdak/snac_44khz (this model)	2.6 kbps	44 kHz	54.5 M	🎸 Music / Sound Effects

Usage

Install it using:

pip install snac

To encode (and decode) audio with SNAC in Python, use the following code:

import torch
from snac import SNAC

model = SNAC.from_pretrained("hubertsiuzdak/snac_44khz").eval().cuda()
audio = torch.randn(1, 1, 44100).cuda()  # B, 1, T

with torch.inference_mode():
    codes = model.encode(audio)
    audio_hat = model.decode(codes)

You can also encode and reconstruct in a single call:

with torch.inference_mode():
    audio_hat, codes = model(audio)

⚠️ Note that codes is a list of token sequences of variable lengths, each corresponding to a different temporal resolution.

>>> [code.shape[1] for code in codes]
[16, 32, 64, 128]

Acknowledgements

Module definitions are adapted from the Descript Audio Codec.