That should be used together with ML models. Image compression for example or voice and video compression like what nvidia does. But i do like this implementa

That should be used together with ML models. Image compression for example or voice and video compression like what nvidia does. But i do like this implementation focusing on libs, why not?