A simple, fast and robust program-aware agentic inference system.

🐹 Clean, uninstall, analyze, optimize, and monitor your Mac from the terminal.

MoonEP: A Perfectly Balanced Expert Parallelism Library via Dynamic Redundant Experts

Peek preview window for LSP locations in Neovim

Use the w, e, b motions like a spider. Move by subwords and skip insignificant punctuation.

slime is an LLM post-training framework for RL Scaling.

The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.

Intuitive, easy CLIs based on python type hints.

AI agents running research on single-GPU nanochat training automatically

🚀 Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support

A set of vim, zsh, git, and tmux configuration files.

A tutorial on modern GPU programming for machine learning systems

Learning TileLang with 10 puzzles!

Ongoing research training transformer models at scale

compiler learning resources collect.

Open Machine Learning Compiler Framework

Agentic Kernel Optimization for All — automated GPU kernel optimization for any kernel, any hardware, any language

Conveniently export torch.compile compiled products into self-contained Python files

why not Rewrite It In Rust

cuDNN Frontend is NVIDIA's modern, open-source entry point to the cuDNN library and a growing collection of high-performance open-source kernels.

Wrap Antigravity, ChatGPT Codex, Claude Code, Grok Build as an OpenAI/Gemini/Claude/Codex compatible API service, allowing you to enjoy the free Gemini 3.1 Pro, GPT 5.6 Series, Grok 4.5, Claude mod…

Improve viewing Markdown in Neovim

A hackable markdown, Typst, latex, html(inline) & Asciidoc previewer for Neovim

high-performance linear attention kernel library built on TileLang

Causal depthwise conv1d in CUDA, with a PyTorch interface

A kernel library written in tilelang

FlashKDA: high-performance Kimi Delta Attention kernels