Recent submissions

Comment · 3 months ago
Comment · 4 months ago
Host to device bandwidth (ram to vram) is 128Gb/s for PCIe Gen 6. VRAM to GPU bandwidth is 1.8Tb/s for GDDR 7 (5090), and 8Tb/s for HBM3e (B200). So it can be faster to recompute than offload kv cache.
Comment · 8 months ago
The Little Learner is a great book. I tried rewriting all the code in Python/JAX while following the Scheme code style as closely as possible, and it worked out great.

The appendix on autodiff is a bit rushed, in my opinion. But in all fairness, the number of pages would probably need to be doubled to give a proper exposition of autodiff.

Comment · 11 months ago
Best decision I made this year.
Comment · 1 year ago
Comment · 1 year ago
Comment · 1 year ago
Comment · 1 year ago
you have no freedom of speech to publish opinions opposing (as decided by me) freedom of speech.
Comment · 1 year ago
Comment · 2 years ago
Comment · 2 years ago
Comment · 2 years ago
There is FxHashMap (https://github.com/rust-lang/rustc-hash) which is faster than std::collections::HashMap.

With ~100 static entries, o1hash (https://github.com/rurban/smhasher/blob/master/o1hash.h) should also work.

Comment · 2 years ago
I am wondering if AlphaStar is the most expensive paper ever.
Comment · 2 years ago
Comment · 2 years ago

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton