无法从pinecone库导入Client类的技术问题求助

你当前使用的是Pinecone SDK 7.3.0版本，这个版本的API已做重大更新，不再存在Client类，这是报错的核心原因。以下是具体修复步骤：

1. 替换客户端初始化逻辑 #

旧版本（2.x及以下）的Client类已被替换为Pinecone类，修改导入和初始化代码：

# 替换原有的pinecone导入逻辑
from pinecone import Pinecone

# 初始化客户端
pinecone_api_key = os.getenv("PINECONE_API_KEY")
pinecone_environment = "us-east-1"
pc = Pinecone(api_key=pinecone_api_key, environment=pinecone_environment)

2. 调整索引操作代码 #

所有涉及客户端的索引操作改用新的Pinecone实例：

index_name = "gentlecat"

# 检查索引是否存在，不存在则创建
if index_name not in [idx.name for idx in pc.list_indexes()]:
    pc.create_index(name=index_name, dimension=384, metric="cosine")

# 连接目标索引
index = pc.Index(index_name)

3. 适配LangChain向量存储创建 #

LangChain的Pinecone.from_documents方法需适配新SDK，可选择两种方式：

• 直接传入API密钥和环境信息：

  vectorstore = Pinecone.from_documents(
      docs, 
      embeddings, 
      index_name=index_name, 
      pinecone_api_key=pinecone_api_key,
      pinecone_environment=pinecone_environment
  )
  

• 传入已初始化的Pinecone实例：

  vectorstore = Pinecone.from_documents(
      docs, 
      embeddings, 
      index_name=index_name, 
      client=pc
  )
  

完整修复后的代码 #

import os
from langchain_community.document_loaders import TextLoader
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import Pinecone
from langchain_community.llms import HuggingFaceHub
from langchain_core.prompts import PromptTemplate
from langchain.chains import RetrievalQA
from dotenv import load_dotenv 
from langchain_text_splitters import CharacterTextSplitter

# 加载环境变量
load_dotenv()

# 初始化Pinecone客户端（新版本SDK）
from pinecone import Pinecone
pinecone_api_key = os.getenv("PINECONE_API_KEY")
pinecone_environment = "us-east-1"
pc = Pinecone(api_key=pinecone_api_key, environment=pinecone_environment)

index_name = "gentlecat"

# 检查并创建索引
if index_name not in [idx.name for idx in pc.list_indexes()]:
    pc.create_index(name=index_name, dimension=384, metric="cosine")

# 连接索引
index = pc.Index(index_name)

# 加载并分割文档
loader = TextLoader(r"filepath", encoding="utf-8")
documents = loader.load()
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=4)
docs = text_splitter.split_documents(documents)

# 设置嵌入模型
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2") 

# 创建向量存储
vectorstore = Pinecone.from_documents(docs, embeddings, index_name=index_name, client=pc)

# 设置LLM
llm = HuggingFaceHub(
    repo_id="mistralai/Mixtral-8x7B-Instruct-v0.1",
    model_kwargs={"temperature": 0.8, "top_p": 0.8, "top_k": 50},
    huggingfacehub_api_token=os.getenv("HUGGINGFACE_API_KEY"),
)

# 设置提示模板
template = """You are a helpful assistant. Answer based on the Costco business context.

Context: {context}
Question: {question}
Answer:"""

prompt = PromptTemplate(template=template, input_variables=["context", "question"])

# 创建检索QA链
qa_chain = RetrievalQA.from_chain_type(
    llm=llm,
    chain_type="stuff",
    retriever=vectorstore.as_retriever(),
    chain_type_kwargs={"prompt": prompt},
)

# 示例查询
query = "What is Costco's primary business model?"
answer = qa_chain.run(query)
print("Answer:", answer)

额外说明：当前LangChain版本中，CharacterTextSplitter已迁移至langchain_text_splitters包，需提前安装：pip install langchain-text-splitters

内容的提问来源于stack exchange，提问作者ACR