kayvr

6 ms·

1. 1.▲
   by kayvr 2y ago

   FWIW, I thought the bits on governance and history provided much needed context. Great post.
2. 2.▲
   by kayvr 2y ago

   I've heard the WebGPU workgroup wants to close the gap on tensor core support.
3. 3.▲
   by kayvr 2y ago

   It's not too surprising that Crew Dragon is a bit like a Tesla. But, wow, the ISS has leaks on board.
4. 4.▲
   by kayvr 2y ago

   Rusty would dig this. It's the name of good ol' game. https://en.wikipedia.org/wiki/SteamWorld
5. 5.▲
   by kayvr 2y ago

   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 GPT-4 in the article, but I couldn't figure out if Math
6. 6.▲
   by kayvr 3y ago

   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 performance between WebGPU and CUDA. Admittedly, WebGPU
7. 7.▲
   by kayvr 3y ago

   Also https://github.com/kayvr/TokenHawk , a WebGPU implementation of LLaMA. edit: Note that this is my project.
8. 8.▲
   Show HN: TokenHawk, WebGPU Running LLaMA(github.com)

   2 points by kayvr 3y ago|0 comments