共用方式為


資料來源 - Elasticsearch (預覽)

使用 Azure OpenAI On Your Data 時,Elasticsearch 的可設定選項。 API 第 2024-02-15-preview 版支援此資料來源。

名稱 類型​​ 必要 描述
parameters 參數 True 設定 Elasticsearch 時要使用的參數。
type 字串 True 必須是 elasticsearch

參數

姓名 類型​​ 必要 描述
endpoint string True Elasticsearch 資源所要使用的絕對端點路徑。
index_name 字串 True 在參考的 Elasticsearch 中要使用的索引名稱。
authentication KeyAndKeyIdAuthenticationOptionsEncodedApiKeyAuthenticationOptions 其中一個 True 存取已定義資料來源時要使用的驗證方法。
embedding_dependency One of DeploymentNameVectorizationSourceEndpointVectorizationSourceModelIdVectorizationSource 其中一個 False 向量搜尋的內嵌相依性。 當 query_typevector 時為必要。
fields_mapping FieldsMappingOptions False 與搜尋索引互動時要使用的自訂欄位對應行為。
in_scope boolean False 查詢是否應限制為使用已編製索引的資料。 預設值為 True
query_type QueryType False 要搭配 Elasticsearch 使用的查詢類型。 預設為 simple
role_information 字串 False 提供模型關於其行為方式的指示,以及產生回應時應該參考的任何內容。 您可以描述助理的特質,並對其告知如何格式化回應。
strictness 整數 False 已設定的搜尋相關性篩選嚴格度。 嚴格度越高,精確度就越高,但答案的召回率越低。 預設值為 3
top_n_documents 整數 False 要為設定的查詢顯示的已設定文件數目上限。 預設值為 5

金鑰和金鑰識別碼驗證選項

Azure OpenAI On Your Data 在使用 API 金鑰時的驗證選項。

名稱 類型​​ 必要 描述
key string True 要用於驗證的 Elasticsearch 金鑰。
key_id 字串 True 要用於驗證的 Elasticsearch 金鑰識別碼。
type 字串 True 必須是 key_and_key_id

編碼的 API 金鑰驗證選項

Azure OpenAI On Your Data 在使用 Elasticsearch 編碼 API 金鑰時的驗證選項。

名稱 類型​​ 必要 描述
encoded_api_key string True 要用於驗證的 Elasticsearch 編碼 API 金鑰。
type 字串 True 必須是 encoded_api_key

部署名稱向量化來源

套用向量搜尋時,Azure OpenAI On Your Data 所使用的向量化來源詳細資料。 此向量化來源是以相同 Azure OpenAI 資源中的內部內嵌模型部署名稱為基礎。 此向量化來源可讓您在沒有 Azure OpenAI API 金鑰的情況下使用向量搜尋,而不需要 Azure OpenAI 公用網路存取。

名稱 類型​​ 必要 描述
deployment_name string True 相同 Azure OpenAI 資源內的內嵌模型部署名稱。
type 字串 True 必須是 deployment_name

端點向量化來源

套用向量搜尋時,Azure OpenAI On Your Data 所使用的向量化來源詳細資料。 此向量化來源是以 Azure OpenAI 內嵌 API 端點為基礎。

名稱 類型​​ 必要 描述
endpoint string True 指定應從中擷取內嵌的資源端點 URL。 其格式應為 https://{YOUR_RESOURCE_NAME}.openai.azure.com/openai/deployments/YOUR_DEPLOYMENT_NAME/embeddings。 不允許 api-version 查詢參數。
authentication ApiKeyAuthenticationOptions True 指定從指定端點擷取內嵌時要使用的驗證選項。
type 字串 True 必須是 endpoint

模型識別碼向量化來源

套用向量搜尋時,Azure OpenAI On Your Data 所使用的向量化來源詳細資料。 此向量化來源是以 Elasticsearch 模型識別碼為基礎。

名稱 類型​​ 必要 描述
model_id string True 指定要用於向量化的模型識別碼。 此模型識別碼必須在 Elasticsearch 中定義。
type 字串 True 必須是 model_id

API 金鑰驗證選項

Azure OpenAI On Your Data 在使用 API 金鑰時的驗證選項。

名稱 類型​​ 必要 描述
key string True 要用於驗證的 API 金鑰。
type 字串 True 必須是 api_key

欄位對應選項

選擇性設定,可控制使用已設定的 Elasticsearch 資源時處理欄位的方式。

名稱 類型​​ 必要 描述
content_fields string[] False 應視為內容的索引欄位名稱。
vector_fields string[] False 代表向量資料的欄位名稱。
content_fields_separator 字串 False 內容欄位應使用的分隔符號模式。 預設值為 \n
filepath_field 字串 False 要用作檔案路徑的索引欄位名稱。
title_field 字串 False 要用作標題的索引欄位名稱。
url_field 字串 False 要用作 URL 的索引欄位名稱。

查詢類型

搭配 Azure OpenAI On Your Data 使用時所應執行的 Elasticsearch 擷取查詢類型。

列舉值 描述
simple 代表預設的簡單查詢剖析器。
vector 代表對計算資料的向量搜尋。

範例

先決條件:

  • 設定從使用者到 Azure OpenAI 資源的角色指派。 必要角色:Cognitive Services OpenAI User
  • 安裝 Az CLI,然後執行 az login
  • 定義下列環境變數:AzureOpenAIEndpointChatCompletionsDeploymentNameSearchEndpointIndexNameKeyKeyId
export AzureOpenAIEndpoint=https://example.openai.azure.com/
export ChatCompletionsDeploymentName=turbo
export SearchEndpoint='https://example.eastus.azurecontainer.io'
export IndexName=testindex
export Key='***'
export KeyId='***'

安裝最新的 pip 套件 openaiazure-identity

import os
from openai import AzureOpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

endpoint = os.environ.get("AzureOpenAIEndpoint")
deployment = os.environ.get("ChatCompletionsDeploymentName")
index_name = os.environ.get("IndexName")
search_endpoint = os.environ.get("SearchEndpoint")
key = os.environ.get("Key")
key_id = os.environ.get("KeyId")

token_provider = get_bearer_token_provider(
    DefaultAzureCredential(), "https://cognitiveservices.azure.com/.default")

client = AzureOpenAI(
    azure_endpoint=endpoint,
    azure_ad_token_provider=token_provider,
    api_version="2024-02-15-preview",
)

completion = client.chat.completions.create(
    model=deployment,
    messages=[
        {
            "role": "user",
            "content": "Who is DRI?",
        },
    ],
    extra_body={
        "data_sources": [
            {
                "type": "elasticsearch",
                "parameters": {
                    "endpoint": search_endpoint,
                    "index_name": index_name,
                    "authentication": {
                        "type": "key_and_key_id",
                        "key": key,
                        "key_id": key_id
                    }
                }
            }
        ]
    }
)

print(completion.model_dump_json(indent=2))