This project contains implementation of a neural system for Text-to-Speech synthesis with explicit prosody modelling. The aim is to achieve human quality speech and compare different techniques of prosodic features modelling.
The information about the project's versions can be found in Changelog.md. The package should be considered unstable until it reaches version 1.0.0.