A program designed to convert knowledge graph data into vector representations.

Graph2Vector serves as a bridge between knowledge graph data and vectorized representations, enabling LLM-based RAG question-answering functionality.

1. Download the source code: git clone https://github.com/GrisaiaEvy/graph2vector.git
2. Configure: Set up the necessary parameters and configurations according to your requirements in main.rs.
3. Run: Execute main.rs to initialize the program and interact with the AI for question-answering capabilities.

Neo4j: v4.4.4

Embedding Model: Xenova/bge-small-zh-v1.5

Milvus: v2.3.1

LLM: gpt-3.5-turbo

Test on chinese knowledge graph.

• Add Reranker Support
• Enhance Graph Database Support
• Enhance Embedding Models Support

MIT license