faiss-vector-aggregator 0.1.0

A package to aggregate embeddings in a Faiss vector store based on metadata columns.

Faiss Vector Aggregator

faiss_vector_aggregator is a Python package designed to aggregate embeddings stored in a Faiss vector store based on a specified metadata column. This package simplifies the process of combining embeddings (e.g., averaging) associated with the same identifier, and creates a new Faiss index with the aggregated embeddings.

Features

• Flexible Aggregation: Aggregate embeddings based on any specified metadata field (e.g., id, name, year).
• Simple Interface: Provide input and output folder paths to automatically process Faiss and metadata files.
• Faiss Integration: Directly works with Faiss indexes for efficient similarity search.

Installation

Install the package using pip:

pip install faiss_vector_aggregator