Apppublic

antuka/graphrag

sourceHugging Faceupdated 5mo agoView on Hugging Face

0likes

App README

🕸️ GraphRAG Intelligence Engine

GraphRAG Master Hero

![Deployment Pipeline](https://github.com/Anteneh-T-Tessema/graphrag/actions) ![License: MIT](https://opensource.org/licenses/MIT) ![Local AI](https://ollama.com)

A production-grade, local-first implementation of the GraphRAG architecture. This system transforms raw documents into a structured, community-summarized Knowledge Graph, enabling deep semantic reasoning without any recurring API costs.

📚 Documentation Hub

Explore the system in extreme detail through our specialized guides:

Guide	Description	Highlights
🖼️ [Visual Guide](VISUAL_GUIDE.md)	Best for Concept Learning. Visualizes the entire engine.	3D Architecture, Search Mode Split, Community Detection.
📖 [Deep Dive](DOCUMENTATION.md)	Best for Technical Reference. How it works under the hood.	Step-by-step logic, API Endpoints, Configuration.
🚀 [Deployment Guide](docker-compose.yml)	Best for Production. Running in the cloud.	GHCR Image, Docker Compose, Port mapping.

✨ Key Features

—🧠 100% Local & Private: Powered by Ollama (Llama 3) and Sentence-Transformers. No data ever leaves your machine.
—👁️ Real-Time Watcher: A background service that monitors your data/ folder and automatically indexes new files in real-time.
—🖥️ Forever Background Service: Native macOS/Linux installers that run the intelligence engine as a persistent system utility.
—📊 Interactive D3 Visualization: A premium web dashboard to explore your knowledge graph, nodes, and relationships.
—🧬 Dual-Search Modes: Switch between Local Search (precise facts) and Global Search (thematic synthesis).

⚡ Quick Start

1. Setup Environment

bash

# Clone and enter
git clone https://github.com/Anteneh-T-Tessema/graphrag.git && cd graphrag

# Create venv and install
python3.11 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

2. Configure Local AI

Create a .env file based on .env.example and set LLM_PROVIDER=ollama. Ensure Ollama is running on your machine.

3. Activate Background Intelligence

bash

# Install as a native background service (Mac/Linux)
bash scripts/setup_service.sh

🏗️ Architecture at a Glance

mermaid

graph LR
    A[Documents] -->|Auto-Watcher| B(GraphRAG Core)
    B --> C[Ollama / Local LLM]
    B --> D[NetworkX / GraphML]
    B --> E[Vector Store]
    D --> F[D3.js Dashboard]
    E --> F

For the full architectural sequence, see the [Visual Guide](VISUAL_GUIDE.md).

🤝 Contributing

We welcome contributions! Please see our CONTRIBUTING.md for details on our code of conduct and the process for submitting pull requests.