RAG系统搭建实战

一、RAG 系统搭建实战

基于 LangChain 的检索增强生成系统完整搭建流程。

二、文档加载与分块

from langchain_community.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter

loader = PyPDFLoader("knowledge.pdf")
docs = loader.load()
splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
chunks = splitter.split_documents(docs)

三、向量存储

from langchain_openai import OpenAIEmbeddings
from langchain_community.vectorstores import Chroma
vectorstore = Chroma.from_documents(chunks, OpenAIEmbeddings())

四、生产级优化

• 混合检索：向量 + BM25
• 重排序：Cohere Reranker
• 缓存：Redis 缓存热门查询