A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

Standalone Flash Attention v2 kernel without libtorch dependency

Fast, Flexible and Portable Structured Generation

Universal LLM Deployment Engine with ML Compilation

An ASGI web server, for Python. 🦄

FastAPI framework, high performance, easy to learn, fast to code, ready for production

FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.

Extended pickling support for Python objects

An open-source framework for making universal native apps with React. Expo runs on Android, iOS, and the web.

User-friendly Desktop Client App for AI Models/LLMs (GPT, Claude, Gemini, Ollama...)

A modern GUI client based on Tauri, designed to run in Windows, macOS and Linux for tailored proxy experience

The PCI Utilities

verl: Volcano Engine Reinforcement Learning for LLMs

🚀 Efficient implementations of state-of-the-art linear attention models

Model Context Protocol Servers

ZeroMQ core engine in C++, implements ZMTP/3.1

Longformer: The Long-Document Transformer

Empowering everyone to host fast and efficient Minecraft servers.

Microsoft PowerToys is a collection of utilities that help you customize Windows and streamline everyday tasks

Flutter makes it easy and fast to build beautiful apps for mobile and beyond

NVIDIA Inference Xfer Library (NIXL)

A fast GPU memory copy library based on NVIDIA GPUDirect RDMA technology

Super-fast Structured Outputs

[MIRROR] Tiny HTTP server

Redis library for rust

PowerPoint-ist（/'pauəpɔintist/）, An online presentation application that replicates most of the commonly used features of MS PowerPoint, allowing for the editing and presentation of PPT online. Sup…

A framework for building native applications using React

The official repository for the gem5 computer-system architecture simulator.

A web framework for Rust.