wangq8/ragflow
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
Iterative Multi-Granular RAG with Contextual Hierarchical Graph
Source code of LinearRAG at ICLR'26
the official repo for E^ 2GraphRAG: Streamlining Graph-based RAG for High Efficiency and Effectiveness
"MiniRAG: Making RAG Simpler with Small and Open-Sourced Language Models"
A modular graph-based Retrieval-Augmented Generation (RAG) system
This repository showcases various advanced techniques for Retrieval-Augmented Generation (RAG) systems. RAG systems combine information retrieval with generative models to provide accurate and contextually rich responses.
[EMNLP2025] "LightRAG: Simple and Fast Retrieval-Augmented Generation"
"RAG-Anything: All-in-One RAG Framework"
🔍大模型应用开发实战一:RAG 技术全栈指南,在线阅读地址:https://datawhalechina.github.io/all-in-rag/
Interview challenge for building a Payment Gateway - Python version
Open Source, Distributed, RESTful Search Engine
Fast math typesetting for the web.
Pure Ruby LDAP library
Simplified HTTP request client.