WaveRNN

This is a Pytorch implementation of WaveRNN provided:

Preparation

Requirements

• Python 3.6 or newer
• PyTorch with CUDA enabled

Preparing data

1. Set parameters in utils/audio.py, In particular, you should set sample_rate, hop_length, win_length
2. python process.py --wav_dir='wavs' --output='data'

Training

train.py is the entry point:

$ python train.py

Trained models are saved under the logdir directory.

Generating

generate.py is the entry point:

$ python generate.py --resume="ema_logdir"

audios are saved under the out directory.

Reference

1. fatchord/WaveRNN.
2. mkotha/WaveRNN.