Stars
NSA Triton Kernels written with GPT5 and Opus 4.1
Ever used asyncio and wished you hadn't? A tiny (~300 lines) event loop for Python.
CLI tool for encoding utf8 text using language models
Official repository for our work on micro-budget training of large-scale diffusion models.
An open-source volume bot written in Node.JS for Raydium. Create fast and cheap volume for your token using JITO.
Simple repository for training small reasoning models
A comprehensive Rust SDK for seamless interaction with the PumpFun Solana program.
WIP! Reverse engineered Cursor Tab API in Neovim
Open source interpretability artefacts for R1.
Single File, Single GPU, From Scratch, Efficient, Full Parameter Tuning library for "RL for LLMs"
Compiling useful links, papers, benchmarks, ideas, etc.
A port of n0madic/twitter-scraper to Node.js.
Build your own visual reasoning model
Implementation of the HotStuff consensus protocol.
A version 1.1 of the Alexander Koch low cost robot arm with some small changes.
Prototyping the performance of various components of a theoretical faster Twitter
Fast parallel LLM inference for MLX
A very simple GRPO implement for reproducing r1-like LLM thinking.
Fast, secure, RFC-compliant WebSocket implementation for Rust, supporting compression
Reference implementation of the Transformer architecture optimized for Apple Neural Engine (ANE)