Awesome AI for Infra › ChatOps & AI Assistants for Ops

BUAADreamer/EasyRAG

⭐ 640 Python added to this list on 2026-06-14 repository created 2024-06-04

EasyRAG is a lightweight and efficient Retrieval-Augmented Generation (RAG) framework specifically tailored for automated network operations. It provides accurate question answering for network-related queries through a sophisticated RAG scheme that includes a specific data processing workflow, dual-route sparse retrieval for coarse ranking, an LLM Reranker for improved relevance, and an LLM for answer generation and optimization. The framework emphasizes ease of deployment, primarily relying on BM25 retrieval and BGE-reranker, without requiring model fine-tuning, thus minimizing VRAM usage and ensuring high scalability. It offers a flexible code library with various search and generation strategies to facilitate custom implementations. Furthermore, EasyRAG features an efficient inference acceleration scheme that significantly reduces latency across the entire RAG process (coarse ranking, reranking, and generation) while maintaining high accuracy. Each acceleration component is designed to be plug-and-play, enhancing the overall efficiency of the RAG system. The project's success in the CCF AIOps International Challenge 2024 underscores its effectiveness in applying AI to operational tasks, particularly in the domain of network management.

https://github.com/BUAADreamer/EasyRAG

aiopsragretrieval-augmented-generationnetwork-operationsllmquestion-answeringautomationbm25bge-rerankerccf-aiops-challenge