2noise/ChatTTS ? reverse-engineered prompt

Build me a local text to speech project for dialogue style voices, based on this repo. I want something I can run in Python with a simple web page and a command line option, where I can type English or Chinese text and get back natural sounding speech files. It should feel conversational, not like a robotic narrator, and it should support things like different speakers plus simple control over pauses, laughter, and speaking style when possible.

Please wire up the basic install, model loading, and inference flow so it actually works end to end, including saving output audio files and a very simple example script. If streaming audio generation is already supported here, include that too. Keep it practical and easy to try locally, with clear defaults and a small demo.

Also make sure the project notes that the released model is for research or academic use and not commercial use. If anything in the repo is outdated, look up current docs online if you need to.