vocos-mel-24khz is an open-weight model from Charactr Inc., released under MIT License. Its published files total 54.4 MB. It draws 2.8M downloads a month.
Paper [[abs]](https://arxiv.org/abs/2306.00814) [[pdf]](https://arxiv.org/pdf/2306.00814.pdf) Vocos is a fast neural vocoder designed to synthesize audio waveforms from acoustic features.
SAVRN's Notes on vocos-mel-24khz
Fifty-four megabytes, one weight file, MIT license: vocos-mel-24khz is a neural vocoder that reconstructs 24 kHz audio from mel spectrograms in a single forward pass, and its 2.9M monthly downloads come from pipeline builders, not card renters. The listing publishes no parameter count, no context and no runs-on table because none are needed — this runs on a CPU, at a twentieth the footprint of even the small ASR checkpoints on this list.
The job definition matters: Vocos turns acoustic features into waveforms, so it pairs with whatever generates the mel — a TTS acoustic model, a codebook decoder — and its speed is the point. MIT permits commercial use and modification with the notice kept. The repository is four files and has been quiet since October 2023; pin it and forget it.
Model Card
By Charactr Inc., published under mit, revision 0feb3fdd929b.
Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis
Audio samples | Paper [abs] [pdf]
Vocos is a fast neural vocoder designed to synthesize audio waveforms from acoustic features. Trained using a Generative Adversarial Network (GAN) objective, Vocos can generate waveforms in a single forward pass. Unlike other typical GAN-based vocoders, Vocos does not model audio samples in the time domain. Instead, it generates spectral coefficients, facilitating rapid audio reconstruction through inverse Fourier transform.
Installation
To use Vocos only in inference mode, install it using:
pip install vocos
If you wish to train the model, install it with additional dependencies:
pip install vocos[train]
Usage
Reconstruct audio from mel-spectrogram
import torch
from vocos import Vocos
vocos = Vocos.from_pretrained("charactr/vocos-mel-24khz")
mel = torch.randn(1, 100, 256) # B, C, T
audio = vocos.decode(mel)
Copy-synthesis from a file:
Identity and Version
- Repository
- charactr/vocos-mel-24khz
- Publisher
- Charactr Inc.
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 0feb3fdd929bcd6649e0e7c5a688cf7dd012ef21
- First published
- 2023-06-11
- Last updated
- 2023-10-17
Files and Weights
4 files, 54.4 MB in total. The weights are 1 file totalling 54.4 MB in bin.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| pytorch_model.bin | Weights | 54.4 MB | 97ec976ad1fd |
| config.yaml | Configuration | 461 B | — |
| README.md | Documentation | 1.8 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- mit
- Access
- Open weights, no gate
- Download size
- 54.4 MB
Released by Charactr Inc. through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:2306.00814
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 54.4 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About vocos-mel-24khz
Can I use vocos-mel-24khz commercially?
Yes. vocos-mel-24khz is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.