Marker-Inc-Korea/AutoRAG
An open-source framework that automatically evaluates and optimizes RAG pipelines to find the best configuration for specific datasets using AutoML techniques.
Velocity · 7d
+5.5
★ / day
Trend
→steady
star history
AutoRAG is a RAG optimization tool that helps users find the optimal pipeline for their specific data and use case. It automates the evaluation of various RAG module combinations, allowing developers to systematically test and compare different retrieval and generation configurations. The framework supports benchmarking across multiple RAG components to identify the best-performing pipeline for a given dataset and evaluation criteria.