HiFiGAN implementation

Installation guide

To install all the dependencies, run

cd hifigan-vocoder
pip install -r ./requirements.txt

To train a model, run

python3 train.py -c src/configs/default.json

To download a checkpoint, run

python3 upload_model.py

To get audios for test texts, run

python3 test.py -r checkpoint.pth

The generated audios will appear in the folder "results".

Credits

This repository is based on a heavily modified fork of pytorch-template repository.