Audio-to-Audio
Safetensors
torch
File size: 1,525 Bytes
5c3331f
 
 
 
 
6c25ae7
78dc93d
 
5c3331f
 
0362a84
5c3331f
 
 
9e053ad
5c3331f
9e053ad
5c3331f
9e053ad
 
 
5c3331f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
78dc93d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
---
license: apache-2.0
base_model:
- microsoft/wavlm-large
pipeline_tag: audio-to-audio
library_name: torch
datasets:
- mythicinfinity/libritts
---

# ⚡ FocalCodec

A low-bitrate single-codebook 16 kHz speech codec based on [focal modulation](https://arxiv.org/abs/2203.11926).

This repository contains the **12.5 Hz checkpoint** trained on **LibriTTS 960**, as described in the preprint.

- 📜 **Preprint**: https://arxiv.org/abs/2502.04465

- 🌐 **Project Page**: https://lucadellalib.github.io/focalcodec-web/

- 💾 **GitHub**: https://github.com/lucadellalib/focalcodec

<img src="focalcodec.png" width="700">

---------------------------------------------------------------------------------------------------------

## ▶️ Quickstart

See the readme at: https://github.com/lucadellalib/focalcodec

---------------------------------------------------------------------------------------------------------

## @ Citing

```
@article{dellalibera2025focalcodec,
    title   = {{FocalCodec}: Low-Bitrate Speech Coding via Focal Modulation Networks},
    author  = {Luca {Della Libera} and Francesco Paissan and Cem Subakan and Mirco Ravanelli},
    journal = {arXiv preprint arXiv:2502.04465},
    year    = {2025},
}
```

---------------------------------------------------------------------------------------------------------

## 📧 Contact

[[email protected]](mailto:[email protected])

---------------------------------------------------------------------------------------------------------