Show HN: Flash Attention in ~100 lines of CUDA (github.com) | HN Screenshots

YHacker Newsscreenshots

Screenshot of github.com: Show HN: Flash Attention in ~100 lines of CUDA

First screen, captured 2 years ago.

Looks similar

Screenshot of github.com: Code for the Byte Pair Encoding algorithm, commonly used in LLM tokenization

Code for the Byte Pair Encoding algorithm, commonly used in LLM tokenization (github.com)

65 points by magoghm 2 years ago | 19 comments

Screenshot of github.com: Show HN: Minimal – Open-Source Community driven Hardened Container Images

Show HN: Minimal – Open-Source Community driven Hardened Container Images (github.com)

98 points by ritvikarya98 8 months ago | 27 comments

Screenshot of github.com: Micrograd-CUDA: adapting Karpathy's tiny autodiff engine for GPU acceleration

Micrograd-CUDA: adapting Karpathy's tiny autodiff engine for GPU acceleration (github.com)

128 points by cataPhil 2 years ago | 12 comments

Screenshot of github.com: AI-powered reverse-engineering of Rosetta 2 (for Linux VM)

AI-powered reverse-engineering of Rosetta 2 (for Linux VM)(github.com)

32 points by inoki 8 months ago | 10 comments

Screenshot of github.com: Git-fetch-file – Sync files from other repos with commit tracking and safety

Git-fetch-file – Sync files from other repos with commit tracking and safety (github.com)

27 points by andrewmcwatters 1 year ago | 14 comments

Screenshot of github.com: Bashbro – Make Any Comp a Web-Based File Server

Bashbro – Make Any Comp a Web-Based File Server (github.com)

84 points by ReversedChaos 2 years ago | 38 comments