探索Azure AI Search：实现强大信息检索的利器

akhfuiigabv

于 2024-10-07 18:06:12 发布

阅读量192

点赞数 2

文章标签： azure 人工智能 flask python

本文链接：https://blog.csdn.net/akhfuiigabv/article/details/142744025

版权

引言

Azure AI Search提供了云搜索服务，使开发者能够在大规模的环境下进行向量、关键字和混合查询的信息检索。本文将指导你如何使用Azure AI Search进行信息检索，并整合其他工具如OpenAI Embeddings。

主要内容

安装必需的软件包

要使用Azure AI Search，你需要安装langchain-community和azure-search-documents软件包。

pip install -qU langchain-community
pip install --upgrade --quiet azure-search-documents azure-identity

导入所需库

根据你不同的OpenAI提供商账户选择导入相应的Embedding库。

import os
from langchain_community.vectorstores.azuresearch import AzureSearch
from langchain_openai import AzureOpenAIEmbeddings, OpenAIEmbeddings

配置OpenAI设置

无论是OpenAI账户还是Azure OpenAI账户，都需要配置相应的API密钥。

openai_api_key = "YOUR_API_KEY"
openai_api_version = "2023-05-15"
model = "text-embedding-ada-002"

azure_endpoint = "YOUR_AZURE_OPENAI_ENDPOINT"
azure_openai_api_key = "YOUR_AZURE_OPENAI_KEY"
azure_openai_api_version = "2023-05-15"
azure_deployment = "text-embedding-ada-002"

配置向量存储设置

需要Azure订阅和Azure AI Search服务。

vector_store_address = "YOUR_AZURE_SEARCH_ENDPOINT"
vector_store_password = "YOUR_AZURE_SEARCH_ADMIN_KEY"

创建Embedding和向量存储实例

embeddings = OpenAIEmbeddings(
    openai_api_key=openai_api_key, 
    openai_api_version=openai_api_version, 
    model=model
)

embeddings = AzureOpenAIEmbeddings(
    azure_deployment=azure_deployment,
    openai_api_version=azure_openai_api_version,
    azure_endpoint=azure_endpoint,
    api_key=azure_openai_api_key
)

index_name = "langchain-vector-demo"
vector_store = AzureSearch(
    azure_search_endpoint=vector_store_address,
    azure_search_key=vector_store_password,
    index_name=index_name,
    embedding_function=embeddings.embed_query
)

代码示例

from langchain_community.document_loaders import TextLoader
from langchain_text_splitters import CharacterTextSplitter

# 加载和分割文档
loader = TextLoader("path/to/document.txt", encoding="utf-8")
documents = loader.load()
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
docs = text_splitter.split_documents(documents)

# 添加文档到向量存储
vector_store.add_documents(documents=docs)

# 执行相似性搜索
results = vector_store.similarity_search(
    query="President's statement on Ketanji Brown Jackson", 
    k=3,
    search_type="similarity"
)
print(results[0].page_content)