G0ldenSp00n

Been working through Andrej Karpathy’s ML lectures in Rust. The backprop one went pretty well, but I had to learn how to do type indirection and interior mutabilty because of the backprop graph structure. I’m now on the makemore lecture, but having a lot of trouble building the bi-gram model in Burn (the rust native ML framework), because it seems like directly incrementing the tensor values is insanely slow. His example that takes like 10 seconds to run in Python takes two and a half minutes in Rust with Burn, so trying to figure out how to optimize or speed that up.

So far I have only really scratch built backprop. And had severe performance problems with Burn trying to do something it probably wasn’t built to do. Once I get further in makemore I should have a better idea.

Are all the commenters paid by the video game industry, seriously what is going on there.

Abrash is a snake

What are you signing into where you need a password but don’t have internet?

Wow, didn’t expect all the winter drop features already.

I currently am compiling Arch for Arm manually using some third party packages to run in distrobox on Asahi Linux (M1 Mac Fedora based distro). Hopefully this should mean I won’t need the manual compile step anymore, and hopefully we start seeing more general Arm support for more Linux apps.

Rust is named after the fungus, not oxidized metal

What is so wrong with people being excited about a language they like? I have always found the Rust community extremely welcoming and caring.

They are AR, it is the Orion prototype.