Stars
A comprehensive collection of Agent Skills for context engineering, multi-agent architectures, and production agent systems. Use when building, optimizing, or debugging agent systems that require e…
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
AI for All: The First Systematic Vibe Coding Tutorial | From Zero to Full-Stack, Bring Your Ideas to Life | Live at: www.vibevibe.cn ；全民AI学习第一课，首个系统化 Vibe Coding 开源教程 | 零基础到全栈实战，让人人都能借助 AI 实现自己的想法与…
Unofficial plugin of Pillow(PIL), an image editing library for python
🧠 Make your agents learn from experience.
The absolute trainer to light up AI agents.
Fully open reproduction of DeepSeek-R1
FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI, Junie, Kiro, Leap.new, Lovable, Manus, NotionAI, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI…
A series of technical report on Slow Thinking with LLM
🕸️ A graph-augmented dense statute retriever. (EACL 2023)
This is the office repository for ACL 2024 paper "Learning or Self-aligning? Rethinking Instruction Fine-tuning"
Experiments on multi-hop reasoning with HotpotQA using GRPO fine-tuning in the OpenPipe ART framework. This repo contains training scripts, evaluation results, and inference demos for a Qwen-based …
Our library for RL environments + evals
Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.
Offical Code of MICCAI'25 Best-Paper-Shortlist paper "MedGround-R1: Advancing Medical Image Grounding via Spatial-Semantic Rewarded Group Relative Policy Optimization"
A Datacenter Scale Distributed Inference Serving Framework
An Open-Source Large-Scale Reinforcement Learning Project for Search Agents
Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.
🪄 Create rich visualizations with AI
LLM based data scientist, AI native data application. AI-driven infinite thinking redefines BI.
Graph-R1: Towards Agentic GraphRAG Framework via End-to-end Reinforcement Learning
A flexible and efficient codebase for training visually-conditioned language models (VLMs)
Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train OpenAI gpt-oss, DeepSeek, Qwen, Llama, Gemma, TTS 2x faster with 70% less VRAM.
《软件设计的哲学》中文翻译 | Chinese translation of A Philosophy of Software Design
Bash is all you need - A nano Claude Code–like agent, built from 0 to 1