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:

import torchaudio

y, sr = torchaudio.load(YOUR_AUDIO_FILE)
if y.size(0) > 1:  # mix to mono
    y = y.mean(dim=0, keepdim=True)
y = torchaudio.functional.resample(y, orig_freq=sr, new_freq=24000)
y_hat = vocos(y)

Reconstruct audio from EnCodec tokens

Additionally, you need to provide a bandwidth_id which corresponds to the embedding for bandwidth from the list: [1.5, 3.0, 6.0, 12.0].

vocos = Vocos.from_pretrained("charactr/vocos-encodec-24khz")

audio_tokens = torch.randint(low=0, high=1024, size=(8, 200))  # 8 codeboooks, 200 frames
features = vocos.codes_to_features(audio_tokens)
bandwidth_id = torch.tensor([2])  # 6 kbps

audio = vocos.decode(features, bandwidth_id=bandwidth_id)

Copy-synthesis from a file: It extracts and quantizes features with EnCodec, then reconstructs them with Vocos in a single forward pass.

y, sr = torchaudio.load(YOUR_AUDIO_FILE)
if y.size(0) > 1:  # mix to mono
    y = y.mean(dim=0, keepdim=True)
y = torchaudio.functional.resample(y, orig_freq=sr, new_freq=24000)

y_hat = vocos(y, bandwidth_id=bandwidth_id)

Integrate with 🐶 Bark text-to-audio model

See example notebook.

Pre-trained models

Model Name	Dataset	Training Iterations	Parameters
charactr/vocos-mel-24khz	LibriTTS	1M	13.5M
charactr/vocos-encodec-24khz	DNS Challenge	2M	7.9M

Training

Prepare a filelist of audio files for the training and validation set:

find $TRAIN_DATASET_DIR -name *.wav > filelist.train
find $VAL_DATASET_DIR -name *.wav > filelist.val

Fill a config file, e.g. vocos.yaml, with your filelist paths and start training with:

torchrun \
--nproc-per-node=1 \
train.py -c configs/vocos-v2.yaml \
--trainer.resume_from_checkpoint malaysian_vocos_mel_v2/last.ckpt

Run on single process first to download necessary models, after that scale up to multiGPUs,

torchrun \
--nproc-per-node=4 \
train.py -c configs/vocos-v2.yaml \
--trainer.resume_from_checkpoint malaysian_vocos_mel_v2/last.ckpt

Refer to Pytorch Lightning documentation for details about customizing the training pipeline.

Citation

If this code contributes to your research, please cite our work:

@article{siuzdak2023vocos,
  title={Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis},
  author={Siuzdak, Hubert},
  journal={arXiv preprint arXiv:2306.00814},
  year={2023}
}

License

The code in this repository is released under the MIT license as found in the LICENSE file.