Code for Generating diverse and natural text-to-speech samples using a quantized fine-grained VAE and auto-regressive prosody prior

Add code button
avatar
https://google.github.io/tacotron/publications/prosody_prior