Build me a Text to Speech system TTS using Festival Framework for a new language

I'm looking to build a text-to-speech system using an open-source framework called Festival. I have collected data for that language but I need someone to set up the environment for the framework and preprocess the data according to festivals requirement, then train the model. This models needs to produce a speech that's at least 80% understandable speech.
There's another tool developed by Google that builds a voice that can be used by the Festival framework to build the TTS synthesis system.
https://www.cstr.ed.ac.uk/projects/festival/
https://github.com/google/voice-builder