---

date	title
08/27/2026	optimizing sgemm in cuda
02/24/2026	llm inference engine
12/30/2025	nanogpt: let's reproduce gpt-2 (124m)