What is the best state of the art text to speech model suitable for converting an English written book to an audiobook (mp3) usable from Hugging Face transformers library? 

I don't want to use reader apps, because their performance is not so good, I can leave it running overnight and prefer better intonation and more natural sounding voice.