Raudaschl/rag-fusion
RAG-Fusion: multi-query generation + Reciprocal Rank Fusion for better retrieval-augmented generation. Includes evaluation harness with NFCorpus/BEIR.
This helps scientists and researchers find more relevant information from their document collections. You input your original search query, and it generates multiple refined queries to cast a wider net. The output is a re-ranked list of documents, providing more comprehensive results than traditional searches. This is for anyone who struggles to find that crucial piece of information hidden deep within their documents using standard search tools.
908 stars.
Use this if you need to find more comprehensive and relevant information from a large set of documents, especially when your initial search queries might be too narrow.
Not ideal if you primarily need quick, straightforward answers to simple, unambiguous queries where basic keyword search is sufficient.
Stars
908
Forks
110
Language
Python
License
MIT
Category
Last pushed
Mar 07, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/Raudaschl/rag-fusion"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
HKUDS/LightRAG
[EMNLP2025] "LightRAG: Simple and Fast Retrieval-Augmented Generation"
beir-cellar/beir
A Heterogeneous Benchmark for Information Retrieval. Easy to use, evaluate your models across...
HKUDS/RAG-Anything
"RAG-Anything: All-in-One RAG Framework"
superlinear-ai/raglite
🥤 RAGLite is a Python toolkit for Retrieval-Augmented Generation (RAG) with DuckDB or PostgreSQL
illuin-tech/vidore-benchmark
Vision Document Retrieval (ViDoRe): Benchmark. Evaluation code for the ColPali paper.