共用方式為


AzureMachineLearningIndexChatExtensionConfiguration interface

使用 Azure OpenAI 聊天延伸模組時,Azure Machine Learning 向量索引可設定選項的特定表示法。

屬性

authentication

存取已定義數據源時要使用的驗證方法。 每個數據源類型都支援一組特定的可用驗證方法;如需支持的機制,請參閱數據源的檔。 如果未提供,在您的數據上會嘗試使用系統受控識別 (默認認證) 驗證。

filter

搜尋篩選。 只有在 Azure Machine Learning 向量索引的類型為 AzureSearch 時,才支援。

inScope

查詢是否應該限制為使用已編製索引的數據。

name

Azure Machine Learning 向量索引名稱。

projectResourceId

Azure Machine Learning 專案的資源標識碼。

roleInformation

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,並告訴它如何格式化回應。 其有 100 個令牌限制,而且會根據整體令牌限制計算。

strictness

搜尋相關性篩選的已設定嚴格性。 嚴格程度越高,精確度越高,但答案的回收率較低。

topNDocuments

所設定檔數目的頂端數目,可供設定的查詢使用。

type

設定 Azure OpenAI 聊天延伸模組時要使用的類型標籤。 這通常不應該從 Azure Machine Learning 向量索引的預設值變更。

version

Azure Machine Learning 向量索引的版本。

屬性詳細資料

authentication

存取已定義數據源時要使用的驗證方法。 每個數據源類型都支援一組特定的可用驗證方法;如需支持的機制,請參閱數據源的檔。 如果未提供,在您的數據上會嘗試使用系統受控識別 (默認認證) 驗證。

authentication?: OnYourDataAuthenticationOptionsUnion

屬性值

filter

搜尋篩選。 只有在 Azure Machine Learning 向量索引的類型為 AzureSearch 時,才支援。

filter?: string

屬性值

string

inScope

查詢是否應該限制為使用已編製索引的數據。

inScope?: boolean

屬性值

boolean

name

Azure Machine Learning 向量索引名稱。

name: string

屬性值

string

projectResourceId

Azure Machine Learning 專案的資源標識碼。

projectResourceId: string

屬性值

string

roleInformation

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,並告訴它如何格式化回應。 其有 100 個令牌限制,而且會根據整體令牌限制計算。

roleInformation?: string

屬性值

string

strictness

搜尋相關性篩選的已設定嚴格性。 嚴格程度越高,精確度越高,但答案的回收率較低。

strictness?: number

屬性值

number

topNDocuments

所設定檔數目的頂端數目,可供設定的查詢使用。

topNDocuments?: number

屬性值

number

type

設定 Azure OpenAI 聊天延伸模組時要使用的類型標籤。 這通常不應該從 Azure Machine Learning 向量索引的預設值變更。

type: "azure_ml_index"

屬性值

"azure_ml_index"

version

Azure Machine Learning 向量索引的版本。

version: string

屬性值

string