This repository demonstrates how to build a vector database in Python for semantic search. Text is converted into vector embeddings using a pre-trained SentenceTransformer model and stored in ChromaDB. Queries return the most semantically similar documents instead of relying on exact keywords.
- Convert text into vector embeddings
- Store embeddings in a lightweight vector database (ChromaDB)
- Perform semantic search based on meaning
- Python 3.6+
sentence-transformerschromadb
Install dependencies with:
pip install sentence-transformers chromadbSee the included Python scripts for examples of:
- Loading a pre-trained model
- Adding documents to the database
- Performing semantic search