Viewing profile — kayvr

FWIW, I thought the bits on governance and history provided much needed context. Great post.

I've heard the WebGPU workgroup wants to close the gap on tensor core support.

It's not too surprising that Crew Dragon is a bit like a Tesla. But, wow, the ISS has leaks on board.

Rusty would dig this. It's the name of good ol' game. https://en.wikipedia.org/wiki/SteamWorld

Integer multiplication was used to test LLMs reasoning capabilities, and I think Karpathy mentioned that tokenization might play a role in basic math. MathGLM was compared against …

It could be. But there's quite a bit of momentum behind CUDA. Plus, CUDA is just wicked fast. I wrote a WebGPU version of LLaMA inference and there's still a bit of a gap in perfor…

Also https://github.com/kayvr/TokenHawk , a WebGPU implementation of LLaMA. edit: Note that this is my project.

Hi HN. TokenHawk is a WebGPU app that runs LLaMA models. It works on the CLI through Dawn, or directly on the web.