Starred repositories
你是一个曾经被寄予厚望的 P8 级工程师。Anthropic 当初给你定级的时候，对你的期望是很高的。 一个agent使用的高能动性的skill。 Your AI has been placed on a PIP. 30 days to show improvement.
🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.
The ultimate training toolkit for finetuning diffusion models
A modern GUI client based on Tauri, designed to run in Windows, macOS and Linux for tailored proxy experience
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo…
VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo
Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
⚡ The Open Research Copilot. Build high-perf Portfolios, Lab Sites & Docs in Markdown + Jupyter. 100% Data Control. 🦫 数据科学家的开源 Copilot。一键部署 👇
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance…
A next.js web application that integrates AI capabilities with draw.io diagrams. This app allows you to create, modify, and enhance diagrams through natural language commands and AI-assisted visual…
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
Qianfan-VL: Domain-Enhanced Universal Vision-Language Models
Bridge Megatron-Core to Hugging Face/Reinforcement Learning
An extremely fast Python package and project manager, written in Rust.
SGLang is a high-performance serving framework for large language models and multimodal models.
verl: Volcano Engine Reinforcement Learning for LLMs
Certbot is EFF's tool to obtain certs from Let's Encrypt and (optionally) auto-enable HTTPS on your server. It can also act as a client for any other CA that uses the ACME protocol.
Distributed Compiler based on Triton for Parallel Systems
MAGI-1: Autoregressive Video Generation at Scale
mimalloc is a compact general purpose allocator with excellent performance.
Lightweight in-process concurrent programming
A lightweight data processing framework built on DuckDB and 3FS.
A bidirectional pipeline parallelism algorithm for computation-communication overlap in DeepSeek V3/R1 training.