技术文章

1. 250131 LangChain调用DeepSeek大模型
2. 250212 LangChain实战开发-这坑爬不出来，止步不前
3. 250402 LangChain全面实用指南：如何利用RAG提升效率

核心代码块

切分文本

from langchain.text_splitter import RecursiveCharacterTextSplitter
import minge_config as config

document = config.text_documents

# 递归字符文本分割器
splitter = RecursiveCharacterTextSplitter(
    # separators=["\n\n", "\n"],
    chunk_size=100,
    chunk_overlap=10
)

# 分割文档
chunks = splitter.split_text(document)
for i,chunk in enumerate(chunks):
    print(f"Chunk {i+1}: \n{chunk}\n")

带语义召回功能的Agent

from langchain_huggingface import HuggingFaceEmbeddings
from langchain_community.vectorstores  import FAISS
from langchain_core.documents  import Document

def hg_embed():
    embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2") # download model 90M
    documents = [
        Document(page_content="现在使用Vibe Coding，还晚吗?我该怎么端好程序员这个饭碗，是在给你减负吗？"),
        Document(page_content="问题一： 效率降低、结果偏离、输出不稳定"),
        Document(page_content="问题二：应对蜜月期后的无力感"),
        Document(page_content="问题三：应对非技术人员冲击和价值焦虑"),
        Document(page_content="警惕AI的“过度设计”倾向"),
    ]
    # 构建向量数据库
    db = FAISS.from_documents(documents, embeddings)
    query = "有哪些问题，请先做好梳理"
    # 执行语义相似度召回
    results = db.similarity_search(query,k=2)
    # 打印结果
    for result in results:
        print(result.page_content)


if __name__ == "__main__":
    hg_embed()