@fleek

Something new coming soon

fleek.sh
Joined October 2018
  • Tweets3.4K
  • Following192
  • Followers86.9K
587Media
please excuse the silence. we've been cooking up something cool and are excited to share more details soon
41
1
1
43
7,442
NVIDIA just dropped benchmarks showing 4-bit inference loses less than 1 point vs BF16 on most tasks. It's not accuracy per request that you should be measuring. It's tasks completed per dollar. And at that metric, 4-bit wins by a landslide. Read the full blog 👇
Fleek
@fleekJan 29

NVIDIA Just Killed the "Quantization = Quality Loss" Myth

Article · @fleek
33
6
42
11,603

NVIDIA Just Killed the "Quantization = Quality Loss" Myth

Article · @fleek
16
3
2
29
18,297
1/ Yesterday we announced mdspan-cute: C++23 std::mdspan syntax with CUTLASS cute layouts. One header. Zero overhead. Here's how it works 🧵
8
4
22
3,658
more replies
7/ Layout algebra is formalized in Lean 4. 26 theorems, 0 sorry. Properties extracted to RapidCheck tests. The art/ directory has 23 SVG visualizations - we drew pictures until we understood.
2
8
2,339
8/ Check out the code: github.com/weyl-ai/mdspan-cu… Check out the Proofs: github.com/weyl-ai/mdspan-cu… /end
7
2,048
💿 Open Source Release 💿 mdspan-cute: a zero-overhead bridge between C++23 std::mdspan and CUTLASS cute layouts. One header. Swizzled memory. No bank conflicts. Read the blog and check out the repo (links in reply)
3
1
11
2,929
Read the blog: weyl.ai/plan/mdspan-cute/ Check out the repo: github.com/weyl-ai/mdspan-cu…

mdspan-cute: Zero-Overhead Bridge to CUTLASS | Weyl

C++23 std::mdspan meets CUTLASS cute layouts. One header. Zero cost. 26 theorems. 0 sorry.

weyl.ai
1
6
1,799
5/ Quantized RoPE already runs in: → LLaMA → Mistral → Most open source inference stacks This isn't obscure. It's foundational.
1
4
809
6/ On "bit augmentation": Log/exp is a bijection. Information in = information out. You can't create precision from a reversible transformation. Thermodynamics doesn't allow it.
3
712
1/Yesterday we announced nix2gpu - a NixOS package for portable GPU containers. Portable containers prevent vendor inference lock-in. Here's why it's a big deal. #Nix #AIInfra
1
1
2
13
2,553
more replies
7/ Why it matters: Makes distributed GPU compute easy and deterministic. Philosophy: It's just Linux with libs - complexity is optional. Open-source, MIT-licensed; production-tested on Fleek machines.
1
2
946
8/ Check out more info on nix2gpu: Full blog: weyl.ai/plan/portable-nix-gp… Repo: github.com/fleek-sh/nix2gpu Quickstart in README - test and send feedback! /End

Ruining GPU Market Owners' Day with the Power of Nix | Weyl

Build containers with nix2gpu that run on any GPU market

weyl.ai
2
879
Load more