💥 Fast State-of-the-Art Tokenizers optimized for Research and Production
Pure Rust Inference Engine
An extremely fast CSS parser, transformer, bundler, and minifier written in Rust.
Project Titania is a complete large language model system, from transformer to transistor, simple enough for one person to understand.
PSSA: a non-transformer language model written from scratch in Rust. Recurrent state-space layer, episodic memory, self-modifying weights, CUDA training, and a full terminal dashboard.
⚡ Pure-Rust WebGPU inference engine — OpenAI-API compatible, GGUF native, runs on any GPU. No Python. No llama.cpp. Single binary.