README.md

TTS HiFiGAN project

Aksenov Yaroslav

Installation guide

Download model to the folder final_model

Create folder "data/LJSpeech-1.1" with LJSpeech dataset

pip install -r ./requirements.txt

Launching guide

Testing:

python test.py \
   -c src/train_config.json \
   -o generated \
   -r final_model/hifigan_ckpt.pth \
   -wp wavs

Training:

python train.py \
   -c src/train_config.json \
   -wk "YOUR_WANDB_API_KEY"

Test outputs

Audios for test melspecs are available in generated directory