This is a project bout GraphRAG Chat UI/input and configuration This initiative is an adaptation of Microsoft's GraphRAG core. This project features a comprehensive and interactive user interface designed to enhance your experience using graphrag.
- Interactive User Interfaces: Enjoy a rich and engaging UI for all your needs.
- Prompt-Tuning: Fine-tune prompts for optimal performance with ui integrated.
- Querying: Efficiently query and retrieve information.
Configure your settings easily in the .env file.
Search information within the chat UI.
To get started with this project, please follow the instructions below:
-
Clone the Repository:
git clone https://github.com/thethepai/Simple-Graphrag.git
-
Install Dependencies:
poetry is strongly recommended to load and develop this project
cd Simple-Graphrag pip install poetry poetry install
-
Run the Application:
cd Simple-Graphrag poetry run poe run_app
For detailed usage, please check the start page in the app and refer to the microsoft graphrag wiki:Welcome to GraphRAG (microsoft.github.io)
-
Fastapi support
- Streamlit has high overhead and poor performance for backend tasks. FastAPI offers better performance and scalability.
-
Nextjs frontend
- Streamlit's frontend capabilities are limited and not suitable for complex UI requirements. Next.js provides a more robust and flexible frontend framework.
-
Neo4j ui
- use neo4j to manage graph data easily
-
Query rewrite and multi session improve
- Rewriting user queries can improve recall rates. Enhancing multi-turn conversations with reference replacement can improve user experience.
-
File processor about xlsx
- Handle multiple tables, complete missing table information, and split tables into sections in a smart way.