File size: 520 Bytes
0abfcc2 a1eddb6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
---
language: en
datasets:
- ljspeech
tags:
- audio
- text-to-speech
---
# HiFi-GAN
[HiFi-GAN](https://arxiv.org/abs/2010.05646) vocoder trained on the [LJ Speech dataset](https://keithito.com/LJ-Speech-Dataset/). The modeling code is based on the [official implementation](https://github.com/jik876/hifi-gan) and the [fairseq adaptation](https://github.com/pytorch/fairseq).
## Usage
```python
from transformers import AutoModel
model = AutoModel.from_pretrained("jake/hifigan-lj-v1", trust_remote_code=True)
```
|