warpfront / hipfire

RDNA-native LLM inference engine in Rust.

RyanCodrai / turbovec

A vector index built on TurboQuant, written in Rust with Python bindings

Bellman281 / esp32-tinyLLM

Rust inference engine for a 28.9M-parameter PLE TinyLM on the ESP32-S3 — 9.4 tok/s, faster than the C implementation it was ported from