George Hotz: evaluation

Transcript context

…And have you looked into the other approaches like PyTorch Lightning to accelerate PyTorch itself? Well, PyTorch Lightning, my understanding is, it's mostly a framework around PyTorch, right? PyTorch Lightning is not going to fix this fundamental problem of I multiply six tensors together. It's not going to fix it going to memory any more than a single read from each and a single write to the output. There are lower level things in PyTorch that are, I'm not exactly sure what Dynamo does, but I know they're generating some Triton stuff, which is going to generate the kernels on the fly. But, you know, PyTorch Lightning is at a higher level of abstraction. So TinyGrad's front-end stuff looks like PyTorch. I made a few tweaks. There's a few things I don't like about PyTorch. Why is Relu a class? Really, what's the state? You make a class, and there's a state. Everything should just be Torch functional and then Relu, but just dot Relu on the tensor. There's things in Torch where you have to do tensor dot and not a tensor dot. It just shows an API that's not perfectly refined. But when you're doing stuff TinyGrad style where you don't have lines, well, it has to work this way. Because even the lines to express the, well, you can't use the where operator in PyTorch. Why is it true case, condition, false case? Ugh, that's how Python expresses ifs. It's disgusting. Turner operators are much nicer. It should be, I can do my like a less than zero dot where a comma one, right? The very pandas-like API?…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.