Read the README and the code. This is a breathless pronouncement of a thin wrapper of some other wrappers. We really need to watch the hype, lest we kill the actually important stuff going on in ML
to be fair the author mentions in the README that the project was built to validate the idea. I think if you're just trying to validate an idea it's totally fine to be using thin wrappers
I also think it is a great way to learn from reading the code without noise and overhead added