Implementation of HiFi-GAN vocoder.

Installation guide

As in previous projects the installation process is also pretty simple.

You can use standart Kaggle environment and the following minimal requirements to run the code.

pip install wandb
pip install librosa
pip install inflect
pip install speechbrain

This would be enough to run the inference.

Use the following code for donwloading and running:

git clone https://github.com/Pe4enkazAMI/tts_project_2.git
cd tts_project_2
python download_weights.py
python test.py -r  ../MyanFi280.pth -t ../tts_project_2/test_data -c ../tts_project/config.json

Details

This is also just an implementation of the HiFi-GAN.