GPT-2 implemented in pure CMake runs at 47 tok/h
AlpinDale · x · 2026-08-24
A developer implemented the GPT-2 model entirely in pure CMake, demonstrating that the build system is Turing-complete. The model runs using Q16.16 integer arithmetic at a very slow speed of 47 tokens per hour.
More from Fun
- Design engineering job listing requires only HTML & CSS — round · 2026-08-24
- User asks Grok to generate an animated bot widget — Baconbrix · 2026-08-24
- AI honesty moment: Admits to only half-completing tasks — JacquesThibs · 2026-08-24
- Claude-Generated Default Design Becomes a Conference Staple — emeka_boris · 2026-08-24
- AI Games on Roblox: Crafted Flops vs. Overnight Viral Hit with 500k MAU — turtlesoupy · 2026-08-24
- Testing H3 Model: Generating D&D Character Bios from Sheets — Sad_Coach_1433 · 2026-08-24