Voice Cloning App

A Python/Pytorch app for easily synthesising human voices.

System Requirements

  • Windows 10 or Ubuntu 20.04+ operating system
  • 5GB+ Disk space
  • NVIDIA GPU with at least 4GB of memory & driver version 450.36+ (optional)

Key features

  • Automatic dataset generation
  • Local & remote training
  • Easy train start/stop
  • Tools for extracting kindle & audible as data sources
  • Data importing/exporting
  • Word replacement suggestion
  • Multi GPU support

Manual Guides

Experimental/In-Development features

Please try them out and give feedback on the discord so that they can be added to production.

Future Improvements

  • Add support for alternative models
  • Improved batch size estimation
  • AMD GPU support