CoolFace
Apppublic

abadesalex/DocuRAG

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes
App README

This project implements a simple RAG system using hybrid search techniques. The system allows you to:

  • Upload a document, which is automatically chunked and stored in a vectorized database.
  • Perform hybrid search using both keyword-based and semantic search (cosine similarity).
  • Query an LLM (paraphrase-MiniLM-L6-v2) to retrieve relevant information from the document.

The document chunks can be previewed, and the system provides responses based on the uploaded content, enabling efficient and accurate document-based information retrieval.