WaveGlow

A Tensorflow implementation of the WaveGlow: A Flow-based Generative Network for Speech Synthesis( https://arxiv.org/abs/1811.00002)

Preparing data

1. Set parameters in hparam.py
2. python process.py --wav_dir='wavs' --output='data'

Training

train.py is the entry point:

$ python train.py--wave_dir="data/train/audio" --lc_dir="data/train/mel"

Trained models are saved under the logdir/waveglow directory.

Generating

generate.py is the entry point:

$ python generate.py --lc_dir="data/test/mel" --out_dir="samples" --restore_from="logdir/waveglow"

Notes

• baseline model is trained using data form here(https://weixinxcxdb.oss-cn-beijing.aliyuncs.com/gwYinPinKu/BZNSYP.rar)