yu3zhou4

6 ms·

1.
by yu3zhou4 14d ago
I found that in LLMs the self-reference defined as "referring in generated text to itself" (so a bit different than in Scott's blog) comes largely from the chat template https://openreview.net/pdf?id=3O2A23MhN
2.
Offpunk Manifesto(blog.ayom.media)
3 points by yu3zhou4 14d ago|0 comments
3.
by yu3zhou4 21d ago
Also PageRank as a Markov chain [0] https://math.libretexts.org/Bookshelves/Linear_Algebra/Under...
4.
The Hacker's Renaissance (2025)(phrack.org)
125 points by yu3zhou4 1mo ago|103 comments
5.
by yu3zhou4 1mo ago
I’m happy it helps you!
6.
by yu3zhou4 1mo ago
If you prefer C++ and CUDA, then there's also tiny-vllm of mine [0] - recently we broke 1k gh stars [0] https://github.com/jmaczan/tiny-vllm
7.
by yu3zhou4 2mo ago
My fav zine
8.
by yu3zhou4 2mo ago
It's a very useful guide, I used it when working on torch-webgpu some time ago. Elie also published few packages that help with using some WebGPU related things, I don't remember exactly what was that but I recall it saved me lots
9.
by yu3zhou4 3mo ago
Print ondemand can be ordered for less than $5 per zine in Lulu, since 2022 or so. I guess Zach doesn’t make a cut from it since this $5 probably barely covers the cost of print and Lulu fee :(
10.
Exapunks (2018)(zachtronics.com)
333 points by yu3zhou4 3mo ago|123 comments
11.
William Thurston: On Proof and Progress in Mathematics (1994) [pdf](ams.org)
1 points by yu3zhou4 3mo ago|0 comments
12.
AMD contributes their GPU support to tiny-vLLM(github.com)
2 points by yu3zhou4 3mo ago|0 comments
13.
by yu3zhou4 3mo ago
PyTorch?
14.
The Cypherpunk Library(cypherpunkbooks.com)
378 points by yu3zhou4 3mo ago|97 comments
15.
by yu3zhou4 4mo ago
README is in my opinion (author here) the most interesting - I wrote it to help others build useful mental model to be able to recreate the project yourself, without need to even read my code
16.
Show HN: Tiny-vLLM – high performance LLM inference engine in C++ and CUDA(github.com)
205 points by yu3zhou4 4mo ago|18 comments
17.
by yu3zhou4 4mo ago
There was onivim that was a bit hyped a few years ago but unfortunately it died
18.
by yu3zhou4 4mo ago
Just learning math and trying myself in ml research
19.
by yu3zhou4 4mo ago
Laughed hard about Collegium Tumanum
20.
by yu3zhou4 4mo ago
Poland was sort of occupied until 1989
21.
by yu3zhou4 5mo ago
Good idea, too! Why do you see my explanation as cynical, though?
22.
by yu3zhou4 5mo ago
Maybe it’s in order to have an external provider to blame for failures and shift the blame/responsibility?
23.
by yu3zhou4 5mo ago
Adding a support for new hardware to PyTorch is actually quite convenient. I did that with WebGPU using the same PrivateUse1 mechanism TorchTPU used. Every hardware has its own slot and identifier, and when you want to add a support for a n
24.
by yu3zhou4 5mo ago
They write that they use PrivateUse1, so it’s a custom out-of-tree backend
25.
TorchWebGPU: Running PyTorch Natively on WebGPU(github.com)
2 points by yu3zhou4 5mo ago|0 comments
26.
by yu3zhou4 5mo ago
The system like iOS is still closed. Once Apple ends support for a device, being able to swap a battery won’t help much
27.
Ask HN: Did Claude lowered its usage limits?
7 points by yu3zhou4 5mo ago|2 comments
28.
by yu3zhou4 5mo ago
Thx, great read about Yemen from Maciej Cegłowski https://idlewords.com/2014/07/sana_a.htm
29.
Benchmark LLM Inference on WebGPU(arxiv.org)
1 points by yu3zhou4 5mo ago|0 comments
30.
by yu3zhou4 5mo ago
An open course on building high performance LLM inference engine! Hope to finish by the end of April https://github.com/jmaczan/tiny-vllm

More ›