RAG技术工作流程详细指南

# 示例代码：文档分割与向量化
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import OpenAIEmbeddings

text_splitter = RecursiveCharacterTextSplitter(
    chunk_size=1000,
    chunk_overlap=200
)
documents = text_splitter.split_documents(docs)

embeddings = OpenAIEmbeddings()
vectorstore = Chroma.from_documents(documents, embeddings)

# 示例代码：相似性检索
retriever = vectorstore.as_retriever(
    search_type="similarity",
    search_kwargs={"k": 4}
)
relevant_docs = retriever.get_relevant_documents(user_query)

# 示例代码：增强生成
from langchain.chat_models import ChatOpenAI
from langchain.schema import HumanMessage, SystemMessage

llm = ChatOpenAI(temperature=0.7)

prompt = f"""
基于以下上下文信息：
{context}

请回答这个问题：
{question}
"""

response = llm([
    SystemMessage(content="你是一个有帮助的助手"),
    HumanMessage(content=prompt)
])