Comment by coolThingsFirst

I still don’t understand the process of learning ML, like sure we build micrograd but is it only didactic exercise or can we use it to train it to do something serious on our own hardware?