Patrick C. Toulme

pyptx. A Python DSL to write Nvidia PTX for Hopper and Blackwell in JAX and PyTorch

justabyte. Code snippets and reproductions from JustAByte

harnessgym. Iterative agent harness improvement: run a coding agent on a hard task, generate the reusable tooling it was missing, qualify it, and replay fresh sessions with it activated. Works with Codex and Claude Code.

xla. A machine learning compiler for GPUs, CPUs, and ML accelerators

axlearn. Python