Skip to content

xcmyz/tacotron2.xcmyz

Repository files navigation

Tacotron2

Use NVIDIA's tacotron2 code, and modify training mode and data loader.

Usage

  1. Download and unzip LJSpeech dataset in data
  2. Run python3 preprocess.py
  3. Run python3 train.py
  4. Download waveglow pretrained model in waveglow/pretrained_model
  5. Run python3 eval.py --step (checkpoint step)
  6. Samples here (step: 30000; batch size: 128; vocoder: waveglow)

About

new version of tacotron2 (old version: https://github.com/xcmyz/Tacotron2-Pytorch)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages