Installation

Ensure you have:

• Python >= 3.6
• Pytorch 1 with CUDA

Then install the rest with pip:

pip install -r requirements.txt

Quick Start

If you want to use TTS functionality immediately you can simply use:

python quick_start.py

This will generate everything in the default sentences.txt file and output to a new 'quick_start' folder where you can playback the wav files and take a look at the attention plots

You can also use that script to generate custom tts sentences and/or use '-u' to generate unbatched (better audio quality):

python quick_start.py -u --input_text "What will happen if I run this command?"