Skip to content

RAG Engineering

Retrieval-Augmented Generation (RAG) is a core technique for combining external knowledge with LLMs. This module covers RAG architecture design, vector database practices, and evaluation optimization.

Contents

Topic Description
RAG Architecture Design Evolution from Naive RAG to Modular RAG
Vector Database in Practice Embedding model selection and vector DB comparison
RAG Evaluation and Optimization RAGAS framework, hybrid search, reranking optimization

Learning Path

  1. Architecture Understanding — Master design choices at each stage of the RAG pipeline
  2. Technology Selection — Understand vector database and embedding model options
  3. Evaluation and Optimization — Learn RAG system evaluation metrics and optimization methods

评论 #