---
license: mit
language:
- ko
---

# Korean FastSpeech 2 - Pytorch Implementation


# Dependencies
python=3.9,
[pytorch](https://pytorch.org/)=1.13, [ffmpeg](https://ffmpeg.org/) [g2pk](https://github.com/Kyubyong/g2pK)
```
sudo apt update
sudo apt install ffmpeg
pip install g2pk
pip install -r requirements.txt
```

# Useage
Data propress
Train VocGAN model
Train Fastspeech2 model

# References
- [FastSpeech 2: Fast and High-Quality End-to-End Text to Speech](https://arxiv.org/abs/2006.04558), Y. Ren, *et al*.
- [HGU-DLLAB](https://github.com/HGU-DLLAB/Korean-FastSpeech2-Pytorch)
- [rishikksh20's VocGAN implementation](https://github.com/rishikksh20/VocGAN)