Skip to content

Instantly share code, notes, and snippets.

View lxynov's full-sized avatar

Xingyuan Lin lxynov

  • Celonis
  • New York
  • 07:20 (UTC -04:00)
View GitHub Profile
@lxynov
lxynov / FAQ.md
Created January 27, 2025 17:55 — forked from ngxson/FAQ.md
convert ARM NEON to WASM SIMD prompt

What is your setup?

Just chat.deepseek.com with prompts adapted from this gist.

Does it work in one-shot or I have to prompt it multiple times?

  • For the qX_0 variants, they are actually quite straight-forward so deepseek can come up with a correct result in 1 shot. It is already crazy enough, given that ChatGPT and Claude have never produced a working result for me.
  • For the qX_K it's more complicated, I would say most of the time I need to re-prompt it 4 to 8 more times.
  • The most difficult was q6_K, the code never works until I ask it to only optimize one specific part, while leaving the rest intact (so it does not mess up everything)