This repo is cloned and modified from Learning to Speak from Text for Low-Resource TTS to use as a baseline for my other project PRESENT.

Please refer to inference.ipynb if you want to run inference with the Takaaki-Saeki's pretrained multilingual model. I did it for Mandarin and provide the pinyin to IPA conversion here.