FastSpeech2

Installation guide

Installation guide is pretty simple.

Minimal requirements sufficient to run the code are just pytorch and the following libraries, make sure to run the following code

pip install wandb
pip install textgrid
pip install librosa
pip install pyworld
pip install inflect

Weights can be downloaded here https://www.kaggle.com/datasets/yanmaximov/myanchik

To run the inference please run the following code

python test.py -r WEIGHTS -c CONFIG

Details

This is just a fine implementation of FastSpeech2 model, without unnecessary details and things.

During project author tried to study optimization, however it did not workout. Though, the generations are pretty good.