Microsoft.MachineLearning webServices 2016-05-01-preview
Bicep 資源定義
WebServices 資源類型可以使用目標作業來部署:
- 資源群組 - 請參閱 資源群組部署命令
如需每個 API 版本中已變更屬性的清單,請參閱 變更記錄檔。
資源格式
若要建立 Microsoft.MachineLearning/webServices 資源,請將下列 Bicep 新增至範本。
resource symbolicname 'Microsoft.MachineLearning/webServices@2016-05-01-preview' = {
location: 'string'
name: 'string'
properties: {
assets: {
{customized property}: {
id: 'string'
inputPorts: {
{customized property}: {
type: 'string'
}
}
locationInfo: {
credentials: 'string'
uri: 'string'
}
metadata: {
{customized property}: 'string'
}
name: 'string'
outputPorts: {
{customized property}: {
type: 'string'
}
}
parameters: [
{
modeValuesInfo: {
{customized property}: {
interfaceString: 'string'
parameters: [
...
]
}
}
name: 'string'
parameterType: 'string'
}
]
type: 'string'
}
}
commitmentPlan: {
id: 'string'
}
description: 'string'
diagnostics: {
expiry: 'string'
level: 'string'
}
exampleRequest: {
globalParameters: {
{customized property}: any(Azure.Bicep.Types.Concrete.AnyType)
}
inputs: {
{customized property}: [
[
any(Azure.Bicep.Types.Concrete.AnyType)
]
]
}
}
exposeSampleData: bool
input: {
description: 'string'
properties: {
{customized property}: {
description: 'string'
format: 'string'
properties: {
{customized property}: {
enum: [
any(Azure.Bicep.Types.Concrete.AnyType)
]
format: 'string'
type: 'string'
x-ms-isnullable: bool
x-ms-isordered: bool
}
}
title: 'string'
type: 'string'
}
}
title: 'string'
type: 'string'
}
keys: {
primary: 'string'
secondary: 'string'
}
machineLearningWorkspace: {
id: 'string'
}
output: {
description: 'string'
properties: {
{customized property}: {
description: 'string'
format: 'string'
properties: {
{customized property}: {
enum: [
any(Azure.Bicep.Types.Concrete.AnyType)
]
format: 'string'
type: 'string'
x-ms-isnullable: bool
x-ms-isordered: bool
}
}
title: 'string'
type: 'string'
}
}
title: 'string'
type: 'string'
}
parameters: {
{customized property}: 'string'
}
readOnly: bool
realtimeConfiguration: {
maxConcurrentCalls: int
}
storageAccount: {
key: 'string'
name: 'string'
}
title: 'string'
packageType: 'string'
// For remaining properties, see WebServiceProperties objects
}
tags: {
{customized property}: 'string'
}
}
WebServiceProperties 物件
設定 packageType 屬性,以指定對象的類型。
針對 Graph,請使用:
{
package: {
edges: [
{
sourceNodeId: 'string'
sourcePortId: 'string'
targetNodeId: 'string'
targetPortId: 'string'
}
]
graphParameters: {
{customized property}: {
description: 'string'
links: [
{
nodeId: 'string'
parameterKey: 'string'
}
]
type: 'string'
}
}
nodes: {
{customized property}: {
assetId: 'string'
inputId: 'string'
outputId: 'string'
parameters: {
{customized property}: 'string'
}
}
}
}
packageType: 'Graph'
}
屬性值
AssetItem
名字 | 描述 | 價值 |
---|---|---|
id | 資產的標識碼。 | 字串 |
inputPorts | 資產輸入埠的相關信息。 | AssetItemInputPorts |
locationInfo | 存取資產的資訊。 | AssetLocation (必要) |
元數據 | 如果資產是自定義模組,這會保存模組的元數據。 | AssetItemMetadata |
名字 | 資產的易記名稱。 | 字串 (必要) |
outputPorts | 資產輸出埠的相關信息。 | AssetItemOutputPorts |
參數 | 如果資產是自定義模組,這會保存模組的參數。 | ModuleAssetParameter[] |
類型 | 資產的類型。 | 'Module' 'Resource' (必要) |
AssetItemInputPorts
名字 | 描述 | 價值 |
---|
AssetItemMetadata
名字 | 描述 | 價值 |
---|
AssetItemOutputPorts
名字 | 描述 | 價值 |
---|
AssetLocation
名字 | 描述 | 價值 |
---|---|---|
憑據 | 如果適用,請存取資產的認證(例如記憶體帳戶連接字串 + Blob URI 所指定的資產) | 字串 |
uri | 可從存取資產的 URI(例如系統資產的 aml://abc,或用戶資產 https://xyz | 字串 (必要) |
ColumnSpecification
名字 | 描述 | 價值 |
---|---|---|
列舉 | 如果數據類型是類別目錄,這會提供已接受的類別清單。 | any[] |
格式 | 數據類型的其他格式資訊。 | 'Byte' 'Char' 'Complex128' 'Complex64' 'Date-time' 'Date-timeOffset' 'Double' 'Duration' 'Float' 'Int16' 'Int32' 'Int64' 'Int8' 'Uint16' 'Uint32' 'Uint64' 'Uint8' |
類型 | 數據行的數據類型。 | 'Boolean' 'Integer' 'Number' 'String' (必要) |
x-ms-isnullable | 指出型別是否支援 Null 值的旗標。 | bool |
x-ms-isordered | 指出類別是否被視為已排序集合的旗標,如果這是類別數據行。 | bool |
CommitmentPlanAutoGenerated
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的承諾用量方案的 Azure Resource Manager 識別符。 | 字串 (必要) |
DiagnosticsConfiguration
名字 | 描述 | 價值 |
---|---|---|
滿期 | 指定記錄停止的日期和時間。 如果為 null,診斷收集不會有時間限制。 | 字串 |
水準 | 指定診斷輸出的詳細資訊。 有效值為:無 - 停用追蹤;錯誤 - 只收集錯誤 (stderr) 追蹤;All - 收集所有追蹤(stdout 和 stderr)。 | 'All' 'Error' 'None' (必要) |
ExampleRequest
名字 | 描述 | 價值 |
---|---|---|
globalParameters | Web 服務全域參數的範例輸入數據 | ExampleRequestGlobalParameters |
輸入 | 以輸入名稱的形式提供給 Web 服務的輸入資料範例輸入數據,以取樣輸入值矩陣對應。 | ExampleRequestInputs |
ExampleRequestGlobalParameters
名字 | 描述 | 價值 |
---|
ExampleRequestInputs
名字 | 描述 | 價值 |
---|
GraphEdge
名字 | 描述 | 價值 |
---|---|---|
sourceNodeId | 來源圖形節點的標識碼。 | 字串 |
sourcePortId | 邊緣所連線來源節點埠的標識碼。 | 字串 |
targetNodeId | 目的地圖形節點的標識碼。 | 字串 |
targetPortId | 邊緣所連線目的地節點埠的標識碼。 | 字串 |
GraphNode
名字 | 描述 | 價值 |
---|---|---|
assetId | 這個節點所代表資產的標識碼。 | 字串 |
inputId | 這個節點所表示之輸入專案的標識碼。 | 字串 |
outputId | 這個節點所表示之輸出項目的標識碼。 | 字串 |
參數 | 如果適用,則為節點的參數。 全域圖形參數會對應至這些參數,並在運行時間設定值。 | GraphNodeParameters |
GraphNodeParameters
名字 | 描述 | 價值 |
---|
GraphPackage
名字 | 描述 | 價值 |
---|---|---|
邊緣 | 組成圖形的邊緣清單。 | GraphEdge[] |
graphParameters | 圖形的全局參數集合,指定為 GraphParameter 地圖的全局參數名稱。 這裡的每個參數都有 1:1 比對,與 WebServiceProperties 層級宣告的全域參數值對應。 | GraphPackageGraphParameters |
節點 | 組成圖形的節點集,提供做為 GraphNode 地圖的 nodeId | GraphPackageNodes |
GraphPackageGraphParameters
名字 | 描述 | 價值 |
---|
GraphPackageNodes
名字 | 描述 | 價值 |
---|
GraphParameter
名字 | 描述 | 價值 |
---|---|---|
描述 | 這個圖表參數的描述。 | 字串 |
連結 | 此參數與圖形中節點的關聯連結。 | GraphParameterLink[] (必要) |
類型 | Graph 參數的類型。 | 'Boolean' 'ColumnPicker' 'Credential' 'DataGatewayName' 'Double' 'Enumerated' 'Float' 'Int' 'Mode' 'ParameterRange' 'Script' 'String' (必要) |
GraphParameterLink
名字 | 描述 | 價值 |
---|---|---|
nodeId | 圖形節點的標識碼 | 字串 (必要) |
parameterKey | 全域參數所對應的節點參數標識碼。 | 字串 (必要) |
InputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
MachineLearningWorkspace
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的機器學習工作區工作區標識碼 | 字串 (必要) |
Microsoft.MachineLearning/webServices
名字 | 描述 | 價值 |
---|---|---|
位置 | 指定資源的位置。 | 字串 (必要) |
名字 | 資源名稱 | 字串 (必要) |
性能 | 包含描述 Web 服務的屬性承載。 | WebServiceProperties (必要) |
標籤 | 資源標籤 | 標記名稱和值的字典。 請參閱範本中的 標籤 |
ModeValueInfo
名字 | 描述 | 價值 |
---|---|---|
interfaceString | 巢狀參數的介面字串名稱。 | 字串 |
參數 | 參數的定義。 | ModuleAssetParameter[] |
ModuleAssetParameter
名字 | 描述 | 價值 |
---|---|---|
modeValuesInfo | 如果這是複雜的模塊參數,則為巢狀介面參數的定義。 | ModuleAssetParameterModeValuesInfo |
名字 | 參數名稱。 | 字串 |
parameterType | 參數類型。 | 字串 |
ModuleAssetParameterModeValuesInfo
名字 | 描述 | 價值 |
---|
OutputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
RealtimeConfiguration
名字 | 描述 | 價值 |
---|---|---|
maxConcurrentCalls | 指定可對 Web 服務進行的最大並行呼叫。 最小值:4,最大值:200。 | int 約束: 最小值 = 4 最大值 = 200 |
ResourceTags
名字 | 描述 | 價值 |
---|
ServiceInputOutputSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構的描述。 | 字串 |
性能 | 指定集合,其中包含 Web 服務每個輸入或輸出的數據行架構。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecificationProperties (必要) |
標題 | Swagger 架構的標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 一律為 『object』。 | 字串 (必要) |
ServiceInputOutputSpecificationProperties
名字 | 描述 | 價值 |
---|
StorageAccount
名字 | 描述 | 價值 |
---|---|---|
鑰匙 | 指定用來存取記憶體帳戶的金鑰。 | 字串 |
名字 | 指定記憶體帳戶的名稱。 | 字串 |
TableSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構描述。 | 字串 |
格式 | 格式,如果 'type' 不是 'object' | 字串 |
性能 | 數據表內的數據行集。 | TableSpecificationProperties |
標題 | Swagger 架構標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 | 字串 (必要) |
TableSpecificationProperties
名字 | 描述 | 價值 |
---|
WebServiceKeys
名字 | 描述 | 價值 |
---|---|---|
主要 | 主要存取金鑰。 | 字串 |
二 次 | 次要存取金鑰。 | 字串 |
WebServiceProperties
名字 | 描述 | 價值 |
---|---|---|
資產 | 包含描述 Web 服務資產的使用者定義屬性。 屬性會以索引鍵/值組表示。 | WebServicePropertiesAssets |
commitmentPlan | 包含與此 Web 服務相關聯的承諾方案。 在建立時設定。 設定之後,就無法變更此值。 注意:承諾用量計劃不會從對 GET 作業的呼叫傳回。 | CommitmentPlanAutoGenerated |
描述 | Web 服務的描述。 | 字串 |
診斷 | 控制 Web 服務的診斷追蹤集合的設定。 | DiagnosticsConfiguration |
exampleRequest | 定義一或多個服務輸入的範例輸入數據。 | ExampleRequest |
exposeSampleData | 當設定為 true 時,範例數據會包含在 Web 服務的 Swagger 定義中。 默認值為 true。 | bool |
輸入 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸入。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
鑰匙 | 包含 Web 服務佈建金鑰。 如果您未指定佈建密鑰,Azure Machine Learning 系統就會為您產生它們。 注意:金鑰不會從呼叫 GET 作業傳回。 | WebServiceKeys |
machineLearningWorkspace | 指定 Machine Learning 工作區,其中包含 Web 服務的來源實驗。 | MachineLearningWorkspace |
輸出 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸出。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
packageType | 將 類型設定為 'Graph',WebServicePropertiesForGraph。 | 'Graph' (必要) |
參數 | 為 Web 服務定義的一組全域參數值,指定為預設值對應的全域參數名稱。 如果未指定預設值,則會將 參數視為必要。 | WebServicePropertiesParameters |
readOnly | 當設定為 true 時,表示 Web 服務是唯讀的,而且無法再更新或修補,只能移除。 默認值為 false。 注意:一旦設定為 true,就無法變更其值。 | bool |
realtimeConfiguration | 包含 Web 服務端點的組態設定。 | RealtimeConfiguration |
storageAccount | 指定 Azure Machine Learning 用來儲存 Web 服務相關信息的記憶體帳戶。 只有記憶體帳戶的名稱會從呼叫 GET 作業傳回。 更新記憶體帳戶資訊時,您必須確定新的記憶體帳戶中提供所有必要的資產,或對 Web 服務的呼叫將會失敗。 | StorageAccount |
標題 | Web 服務的標題。 | 字串 |
WebServicePropertiesAssets
名字 | 描述 | 價值 |
---|
WebServicePropertiesForGraph
名字 | 描述 | 價值 |
---|---|---|
包 | 組成此 Web 服務的圖形套件定義。 | GraphPackage |
packageType | 指定封裝類型。 有效值為 Graph(指定透過 Machine Learning Studio 發佈的 Web 服務)和程式碼(指定使用 Python 之類的程式代碼發佈的 Web 服務)。 注意:目前不支持程序代碼。 | 'Graph' (必要) |
WebServicePropertiesParameters
名字 | 描述 | 價值 |
---|
ARM 樣本資源定義
WebServices 資源類型可以使用目標作業來部署:
- 資源群組 - 請參閱 資源群組部署命令
如需每個 API 版本中已變更屬性的清單,請參閱 變更記錄檔。
資源格式
若要建立 Microsoft.MachineLearning/webServices 資源,請將下列 JSON 新增至範本。
{
"type": "Microsoft.MachineLearning/webServices",
"apiVersion": "2016-05-01-preview",
"name": "string",
"location": "string",
"properties": {
"assets": {
"{customized property}": {
"id": "string",
"inputPorts": {
"{customized property}": {
"type": "string"
}
},
"locationInfo": {
"credentials": "string",
"uri": "string"
},
"metadata": {
"{customized property}": "string"
},
"name": "string",
"outputPorts": {
"{customized property}": {
"type": "string"
}
},
"parameters": [
{
"modeValuesInfo": {
"{customized property}": {
"interfaceString": "string",
"parameters": [
...
]
}
},
"name": "string",
"parameterType": "string"
}
],
"type": "string"
}
},
"commitmentPlan": {
"id": "string"
},
"description": "string",
"diagnostics": {
"expiry": "string",
"level": "string"
},
"exampleRequest": {
"globalParameters": {
"{customized property}": {}
},
"inputs": {
"{customized property}": [
[ {} ]
]
}
},
"exposeSampleData": "bool",
"input": {
"description": "string",
"properties": {
"{customized property}": {
"description": "string",
"format": "string",
"properties": {
"{customized property}": {
"enum": [ {} ],
"format": "string",
"type": "string",
"x-ms-isnullable": "bool",
"x-ms-isordered": "bool"
}
},
"title": "string",
"type": "string"
}
},
"title": "string",
"type": "string"
},
"keys": {
"primary": "string",
"secondary": "string"
},
"machineLearningWorkspace": {
"id": "string"
},
"output": {
"description": "string",
"properties": {
"{customized property}": {
"description": "string",
"format": "string",
"properties": {
"{customized property}": {
"enum": [ {} ],
"format": "string",
"type": "string",
"x-ms-isnullable": "bool",
"x-ms-isordered": "bool"
}
},
"title": "string",
"type": "string"
}
},
"title": "string",
"type": "string"
},
"parameters": {
"{customized property}": "string"
},
"readOnly": "bool",
"realtimeConfiguration": {
"maxConcurrentCalls": "int"
},
"storageAccount": {
"key": "string",
"name": "string"
},
"title": "string",
"packageType": "string"
// For remaining properties, see WebServiceProperties objects
},
"tags": {
"{customized property}": "string"
}
}
WebServiceProperties 物件
設定 packageType 屬性,以指定對象的類型。
針對 Graph,請使用:
{
"package": {
"edges": [
{
"sourceNodeId": "string",
"sourcePortId": "string",
"targetNodeId": "string",
"targetPortId": "string"
}
],
"graphParameters": {
"{customized property}": {
"description": "string",
"links": [
{
"nodeId": "string",
"parameterKey": "string"
}
],
"type": "string"
}
},
"nodes": {
"{customized property}": {
"assetId": "string",
"inputId": "string",
"outputId": "string",
"parameters": {
"{customized property}": "string"
}
}
}
},
"packageType": "Graph"
}
屬性值
AssetItem
名字 | 描述 | 價值 |
---|---|---|
id | 資產的標識碼。 | 字串 |
inputPorts | 資產輸入埠的相關信息。 | AssetItemInputPorts |
locationInfo | 存取資產的資訊。 | AssetLocation (必要) |
元數據 | 如果資產是自定義模組,這會保存模組的元數據。 | AssetItemMetadata |
名字 | 資產的易記名稱。 | 字串 (必要) |
outputPorts | 資產輸出埠的相關信息。 | AssetItemOutputPorts |
參數 | 如果資產是自定義模組,這會保存模組的參數。 | ModuleAssetParameter[] |
類型 | 資產的類型。 | 'Module' 'Resource' (必要) |
AssetItemInputPorts
名字 | 描述 | 價值 |
---|
AssetItemMetadata
名字 | 描述 | 價值 |
---|
AssetItemOutputPorts
名字 | 描述 | 價值 |
---|
AssetLocation
名字 | 描述 | 價值 |
---|---|---|
憑據 | 如果適用,請存取資產的認證(例如記憶體帳戶連接字串 + Blob URI 所指定的資產) | 字串 |
uri | 可從存取資產的 URI(例如系統資產的 aml://abc,或用戶資產 https://xyz | 字串 (必要) |
ColumnSpecification
名字 | 描述 | 價值 |
---|---|---|
列舉 | 如果數據類型是類別目錄,這會提供已接受的類別清單。 | any[] |
格式 | 數據類型的其他格式資訊。 | 'Byte' 'Char' 'Complex128' 'Complex64' 'Date-time' 'Date-timeOffset' 'Double' 'Duration' 'Float' 'Int16' 'Int32' 'Int64' 'Int8' 'Uint16' 'Uint32' 'Uint64' 'Uint8' |
類型 | 數據行的數據類型。 | 'Boolean' 'Integer' 'Number' 'String' (必要) |
x-ms-isnullable | 指出型別是否支援 Null 值的旗標。 | bool |
x-ms-isordered | 指出類別是否被視為已排序集合的旗標,如果這是類別數據行。 | bool |
CommitmentPlanAutoGenerated
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的承諾用量方案的 Azure Resource Manager 識別符。 | 字串 (必要) |
DiagnosticsConfiguration
名字 | 描述 | 價值 |
---|---|---|
滿期 | 指定記錄停止的日期和時間。 如果為 null,診斷收集不會有時間限制。 | 字串 |
水準 | 指定診斷輸出的詳細資訊。 有效值為:無 - 停用追蹤;錯誤 - 只收集錯誤 (stderr) 追蹤;All - 收集所有追蹤(stdout 和 stderr)。 | 'All' 'Error' 'None' (必要) |
ExampleRequest
名字 | 描述 | 價值 |
---|---|---|
globalParameters | Web 服務全域參數的範例輸入數據 | ExampleRequestGlobalParameters |
輸入 | 以輸入名稱的形式提供給 Web 服務的輸入資料範例輸入數據,以取樣輸入值矩陣對應。 | ExampleRequestInputs |
ExampleRequestGlobalParameters
名字 | 描述 | 價值 |
---|
ExampleRequestInputs
名字 | 描述 | 價值 |
---|
GraphEdge
名字 | 描述 | 價值 |
---|---|---|
sourceNodeId | 來源圖形節點的標識碼。 | 字串 |
sourcePortId | 邊緣所連線來源節點埠的標識碼。 | 字串 |
targetNodeId | 目的地圖形節點的標識碼。 | 字串 |
targetPortId | 邊緣所連線目的地節點埠的標識碼。 | 字串 |
GraphNode
名字 | 描述 | 價值 |
---|---|---|
assetId | 這個節點所代表資產的標識碼。 | 字串 |
inputId | 這個節點所表示之輸入專案的標識碼。 | 字串 |
outputId | 這個節點所表示之輸出項目的標識碼。 | 字串 |
參數 | 如果適用,則為節點的參數。 全域圖形參數會對應至這些參數,並在運行時間設定值。 | GraphNodeParameters |
GraphNodeParameters
名字 | 描述 | 價值 |
---|
GraphPackage
名字 | 描述 | 價值 |
---|---|---|
邊緣 | 組成圖形的邊緣清單。 | GraphEdge[] |
graphParameters | 圖形的全局參數集合,指定為 GraphParameter 地圖的全局參數名稱。 這裡的每個參數都有 1:1 比對,與 WebServiceProperties 層級宣告的全域參數值對應。 | GraphPackageGraphParameters |
節點 | 組成圖形的節點集,提供做為 GraphNode 地圖的 nodeId | GraphPackageNodes |
GraphPackageGraphParameters
名字 | 描述 | 價值 |
---|
GraphPackageNodes
名字 | 描述 | 價值 |
---|
GraphParameter
名字 | 描述 | 價值 |
---|---|---|
描述 | 這個圖表參數的描述。 | 字串 |
連結 | 此參數與圖形中節點的關聯連結。 | GraphParameterLink[] (必要) |
類型 | Graph 參數的類型。 | 'Boolean' 'ColumnPicker' 'Credential' 'DataGatewayName' 'Double' 'Enumerated' 'Float' 'Int' 'Mode' 'ParameterRange' 'Script' 'String' (必要) |
GraphParameterLink
名字 | 描述 | 價值 |
---|---|---|
nodeId | 圖形節點的標識碼 | 字串 (必要) |
parameterKey | 全域參數所對應的節點參數標識碼。 | 字串 (必要) |
InputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
MachineLearningWorkspace
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的機器學習工作區工作區標識碼 | 字串 (必要) |
Microsoft.MachineLearning/webServices
名字 | 描述 | 價值 |
---|---|---|
apiVersion | API 版本 | '2016-05-01-preview' |
位置 | 指定資源的位置。 | 字串 (必要) |
名字 | 資源名稱 | 字串 (必要) |
性能 | 包含描述 Web 服務的屬性承載。 | WebServiceProperties (必要) |
標籤 | 資源標籤 | 標記名稱和值的字典。 請參閱範本中的 標籤 |
類型 | 資源類型 | 'Microsoft.MachineLearning/webServices' |
ModeValueInfo
名字 | 描述 | 價值 |
---|---|---|
interfaceString | 巢狀參數的介面字串名稱。 | 字串 |
參數 | 參數的定義。 | ModuleAssetParameter[] |
ModuleAssetParameter
名字 | 描述 | 價值 |
---|---|---|
modeValuesInfo | 如果這是複雜的模塊參數,則為巢狀介面參數的定義。 | ModuleAssetParameterModeValuesInfo |
名字 | 參數名稱。 | 字串 |
parameterType | 參數類型。 | 字串 |
ModuleAssetParameterModeValuesInfo
名字 | 描述 | 價值 |
---|
OutputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
RealtimeConfiguration
名字 | 描述 | 價值 |
---|---|---|
maxConcurrentCalls | 指定可對 Web 服務進行的最大並行呼叫。 最小值:4,最大值:200。 | int 約束: 最小值 = 4 最大值 = 200 |
ResourceTags
名字 | 描述 | 價值 |
---|
ServiceInputOutputSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構的描述。 | 字串 |
性能 | 指定集合,其中包含 Web 服務每個輸入或輸出的數據行架構。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecificationProperties (必要) |
標題 | Swagger 架構的標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 一律為 『object』。 | 字串 (必要) |
ServiceInputOutputSpecificationProperties
名字 | 描述 | 價值 |
---|
StorageAccount
名字 | 描述 | 價值 |
---|---|---|
鑰匙 | 指定用來存取記憶體帳戶的金鑰。 | 字串 |
名字 | 指定記憶體帳戶的名稱。 | 字串 |
TableSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構描述。 | 字串 |
格式 | 格式,如果 'type' 不是 'object' | 字串 |
性能 | 數據表內的數據行集。 | TableSpecificationProperties |
標題 | Swagger 架構標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 | 字串 (必要) |
TableSpecificationProperties
名字 | 描述 | 價值 |
---|
WebServiceKeys
名字 | 描述 | 價值 |
---|---|---|
主要 | 主要存取金鑰。 | 字串 |
二 次 | 次要存取金鑰。 | 字串 |
WebServiceProperties
名字 | 描述 | 價值 |
---|---|---|
資產 | 包含描述 Web 服務資產的使用者定義屬性。 屬性會以索引鍵/值組表示。 | WebServicePropertiesAssets |
commitmentPlan | 包含與此 Web 服務相關聯的承諾方案。 在建立時設定。 設定之後,就無法變更此值。 注意:承諾用量計劃不會從對 GET 作業的呼叫傳回。 | CommitmentPlanAutoGenerated |
描述 | Web 服務的描述。 | 字串 |
診斷 | 控制 Web 服務的診斷追蹤集合的設定。 | DiagnosticsConfiguration |
exampleRequest | 定義一或多個服務輸入的範例輸入數據。 | ExampleRequest |
exposeSampleData | 當設定為 true 時,範例數據會包含在 Web 服務的 Swagger 定義中。 默認值為 true。 | bool |
輸入 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸入。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
鑰匙 | 包含 Web 服務佈建金鑰。 如果您未指定佈建密鑰,Azure Machine Learning 系統就會為您產生它們。 注意:金鑰不會從呼叫 GET 作業傳回。 | WebServiceKeys |
machineLearningWorkspace | 指定 Machine Learning 工作區,其中包含 Web 服務的來源實驗。 | MachineLearningWorkspace |
輸出 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸出。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
packageType | 將 類型設定為 'Graph',WebServicePropertiesForGraph。 | 'Graph' (必要) |
參數 | 為 Web 服務定義的一組全域參數值,指定為預設值對應的全域參數名稱。 如果未指定預設值,則會將 參數視為必要。 | WebServicePropertiesParameters |
readOnly | 當設定為 true 時,表示 Web 服務是唯讀的,而且無法再更新或修補,只能移除。 默認值為 false。 注意:一旦設定為 true,就無法變更其值。 | bool |
realtimeConfiguration | 包含 Web 服務端點的組態設定。 | RealtimeConfiguration |
storageAccount | 指定 Azure Machine Learning 用來儲存 Web 服務相關信息的記憶體帳戶。 只有記憶體帳戶的名稱會從呼叫 GET 作業傳回。 更新記憶體帳戶資訊時,您必須確定新的記憶體帳戶中提供所有必要的資產,或對 Web 服務的呼叫將會失敗。 | StorageAccount |
標題 | Web 服務的標題。 | 字串 |
WebServicePropertiesAssets
名字 | 描述 | 價值 |
---|
WebServicePropertiesForGraph
名字 | 描述 | 價值 |
---|---|---|
包 | 組成此 Web 服務的圖形套件定義。 | GraphPackage |
packageType | 指定封裝類型。 有效值為 Graph(指定透過 Machine Learning Studio 發佈的 Web 服務)和程式碼(指定使用 Python 之類的程式代碼發佈的 Web 服務)。 注意:目前不支持程序代碼。 | 'Graph' (必要) |
WebServicePropertiesParameters
名字 | 描述 | 價值 |
---|
Terraform (AzAPI 提供者) 資源定義
WebServices 資源類型可以使用目標作業來部署:
- 資源群組
如需每個 API 版本中已變更屬性的清單,請參閱 變更記錄檔。
資源格式
若要建立 Microsoft.MachineLearning/webServices 資源,請將下列 Terraform 新增至範本。
resource "azapi_resource" "symbolicname" {
type = "Microsoft.MachineLearning/webServices@2016-05-01-preview"
name = "string"
location = "string"
tags = {
{customized property} = "string"
}
body = jsonencode({
properties = {
assets = {
{customized property} = {
id = "string"
inputPorts = {
{customized property} = {
type = "string"
}
}
locationInfo = {
credentials = "string"
uri = "string"
}
metadata = {
{customized property} = "string"
}
name = "string"
outputPorts = {
{customized property} = {
type = "string"
}
}
parameters = [
{
modeValuesInfo = {
{customized property} = {
interfaceString = "string"
parameters = [
...
]
}
}
name = "string"
parameterType = "string"
}
]
type = "string"
}
}
commitmentPlan = {
id = "string"
}
description = "string"
diagnostics = {
expiry = "string"
level = "string"
}
exampleRequest = {
globalParameters = {
{customized property} = ?
}
inputs = {
{customized property} = [
[
?
]
]
}
}
exposeSampleData = bool
input = {
description = "string"
properties = {
{customized property} = {
description = "string"
format = "string"
properties = {
{customized property} = {
enum = [
?
]
format = "string"
type = "string"
x-ms-isnullable = bool
x-ms-isordered = bool
}
}
title = "string"
type = "string"
}
}
title = "string"
type = "string"
}
keys = {
primary = "string"
secondary = "string"
}
machineLearningWorkspace = {
id = "string"
}
output = {
description = "string"
properties = {
{customized property} = {
description = "string"
format = "string"
properties = {
{customized property} = {
enum = [
?
]
format = "string"
type = "string"
x-ms-isnullable = bool
x-ms-isordered = bool
}
}
title = "string"
type = "string"
}
}
title = "string"
type = "string"
}
parameters = {
{customized property} = "string"
}
readOnly = bool
realtimeConfiguration = {
maxConcurrentCalls = int
}
storageAccount = {
key = "string"
name = "string"
}
title = "string"
packageType = "string"
// For remaining properties, see WebServiceProperties objects
}
})
}
WebServiceProperties 物件
設定 packageType 屬性,以指定對象的類型。
針對 Graph,請使用:
{
package = {
edges = [
{
sourceNodeId = "string"
sourcePortId = "string"
targetNodeId = "string"
targetPortId = "string"
}
]
graphParameters = {
{customized property} = {
description = "string"
links = [
{
nodeId = "string"
parameterKey = "string"
}
]
type = "string"
}
}
nodes = {
{customized property} = {
assetId = "string"
inputId = "string"
outputId = "string"
parameters = {
{customized property} = "string"
}
}
}
}
packageType = "Graph"
}
屬性值
AssetItem
名字 | 描述 | 價值 |
---|---|---|
id | 資產的標識碼。 | 字串 |
inputPorts | 資產輸入埠的相關信息。 | AssetItemInputPorts |
locationInfo | 存取資產的資訊。 | AssetLocation (必要) |
元數據 | 如果資產是自定義模組,這會保存模組的元數據。 | AssetItemMetadata |
名字 | 資產的易記名稱。 | 字串 (必要) |
outputPorts | 資產輸出埠的相關信息。 | AssetItemOutputPorts |
參數 | 如果資產是自定義模組,這會保存模組的參數。 | ModuleAssetParameter[] |
類型 | 資產的類型。 | 'Module' 'Resource' (必要) |
AssetItemInputPorts
名字 | 描述 | 價值 |
---|
AssetItemMetadata
名字 | 描述 | 價值 |
---|
AssetItemOutputPorts
名字 | 描述 | 價值 |
---|
AssetLocation
名字 | 描述 | 價值 |
---|---|---|
憑據 | 如果適用,請存取資產的認證(例如記憶體帳戶連接字串 + Blob URI 所指定的資產) | 字串 |
uri | 可從存取資產的 URI(例如系統資產的 aml://abc,或用戶資產 https://xyz | 字串 (必要) |
ColumnSpecification
名字 | 描述 | 價值 |
---|---|---|
列舉 | 如果數據類型是類別目錄,這會提供已接受的類別清單。 | any[] |
格式 | 數據類型的其他格式資訊。 | 'Byte' 'Char' 'Complex128' 'Complex64' 'Date-time' 'Date-timeOffset' 'Double' 'Duration' 'Float' 'Int16' 'Int32' 'Int64' 'Int8' 'Uint16' 'Uint32' 'Uint64' 'Uint8' |
類型 | 數據行的數據類型。 | 'Boolean' 'Integer' 'Number' 'String' (必要) |
x-ms-isnullable | 指出型別是否支援 Null 值的旗標。 | bool |
x-ms-isordered | 指出類別是否被視為已排序集合的旗標,如果這是類別數據行。 | bool |
CommitmentPlanAutoGenerated
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的承諾用量方案的 Azure Resource Manager 識別符。 | 字串 (必要) |
DiagnosticsConfiguration
名字 | 描述 | 價值 |
---|---|---|
滿期 | 指定記錄停止的日期和時間。 如果為 null,診斷收集不會有時間限制。 | 字串 |
水準 | 指定診斷輸出的詳細資訊。 有效值為:無 - 停用追蹤;錯誤 - 只收集錯誤 (stderr) 追蹤;All - 收集所有追蹤(stdout 和 stderr)。 | 'All' 'Error' 'None' (必要) |
ExampleRequest
名字 | 描述 | 價值 |
---|---|---|
globalParameters | Web 服務全域參數的範例輸入數據 | ExampleRequestGlobalParameters |
輸入 | 以輸入名稱的形式提供給 Web 服務的輸入資料範例輸入數據,以取樣輸入值矩陣對應。 | ExampleRequestInputs |
ExampleRequestGlobalParameters
名字 | 描述 | 價值 |
---|
ExampleRequestInputs
名字 | 描述 | 價值 |
---|
GraphEdge
名字 | 描述 | 價值 |
---|---|---|
sourceNodeId | 來源圖形節點的標識碼。 | 字串 |
sourcePortId | 邊緣所連線來源節點埠的標識碼。 | 字串 |
targetNodeId | 目的地圖形節點的標識碼。 | 字串 |
targetPortId | 邊緣所連線目的地節點埠的標識碼。 | 字串 |
GraphNode
名字 | 描述 | 價值 |
---|---|---|
assetId | 這個節點所代表資產的標識碼。 | 字串 |
inputId | 這個節點所表示之輸入專案的標識碼。 | 字串 |
outputId | 這個節點所表示之輸出項目的標識碼。 | 字串 |
參數 | 如果適用,則為節點的參數。 全域圖形參數會對應至這些參數,並在運行時間設定值。 | GraphNodeParameters |
GraphNodeParameters
名字 | 描述 | 價值 |
---|
GraphPackage
名字 | 描述 | 價值 |
---|---|---|
邊緣 | 組成圖形的邊緣清單。 | GraphEdge[] |
graphParameters | 圖形的全局參數集合,指定為 GraphParameter 地圖的全局參數名稱。 這裡的每個參數都有 1:1 比對,與 WebServiceProperties 層級宣告的全域參數值對應。 | GraphPackageGraphParameters |
節點 | 組成圖形的節點集,提供做為 GraphNode 地圖的 nodeId | GraphPackageNodes |
GraphPackageGraphParameters
名字 | 描述 | 價值 |
---|
GraphPackageNodes
名字 | 描述 | 價值 |
---|
GraphParameter
名字 | 描述 | 價值 |
---|---|---|
描述 | 這個圖表參數的描述。 | 字串 |
連結 | 此參數與圖形中節點的關聯連結。 | GraphParameterLink[] (必要) |
類型 | Graph 參數的類型。 | 'Boolean' 'ColumnPicker' 'Credential' 'DataGatewayName' 'Double' 'Enumerated' 'Float' 'Int' 'Mode' 'ParameterRange' 'Script' 'String' (必要) |
GraphParameterLink
名字 | 描述 | 價值 |
---|---|---|
nodeId | 圖形節點的標識碼 | 字串 (必要) |
parameterKey | 全域參數所對應的節點參數標識碼。 | 字串 (必要) |
InputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
MachineLearningWorkspace
名字 | 描述 | 價值 |
---|---|---|
id | 指定與 Web 服務相關聯的機器學習工作區工作區標識碼 | 字串 (必要) |
Microsoft.MachineLearning/webServices
名字 | 描述 | 價值 |
---|---|---|
位置 | 指定資源的位置。 | 字串 (必要) |
名字 | 資源名稱 | 字串 (必要) |
性能 | 包含描述 Web 服務的屬性承載。 | WebServiceProperties (必要) |
標籤 | 資源標籤 | 標記名稱和值的字典。 |
類型 | 資源類型 | “Microsoft.MachineLearning/webServices@2016-05-01-preview” |
ModeValueInfo
名字 | 描述 | 價值 |
---|---|---|
interfaceString | 巢狀參數的介面字串名稱。 | 字串 |
參數 | 參數的定義。 | ModuleAssetParameter[] |
ModuleAssetParameter
名字 | 描述 | 價值 |
---|---|---|
modeValuesInfo | 如果這是複雜的模塊參數,則為巢狀介面參數的定義。 | ModuleAssetParameterModeValuesInfo |
名字 | 參數名稱。 | 字串 |
parameterType | 參數類型。 | 字串 |
ModuleAssetParameterModeValuesInfo
名字 | 描述 | 價值 |
---|
OutputPort
名字 | 描述 | 價值 |
---|---|---|
類型 | 埠數據類型。 | 'Dataset' |
RealtimeConfiguration
名字 | 描述 | 價值 |
---|---|---|
maxConcurrentCalls | 指定可對 Web 服務進行的最大並行呼叫。 最小值:4,最大值:200。 | int 約束: 最小值 = 4 最大值 = 200 |
ResourceTags
名字 | 描述 | 價值 |
---|
ServiceInputOutputSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構的描述。 | 字串 |
性能 | 指定集合,其中包含 Web 服務每個輸入或輸出的數據行架構。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecificationProperties (必要) |
標題 | Swagger 架構的標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 一律為 『object』。 | 字串 (必要) |
ServiceInputOutputSpecificationProperties
名字 | 描述 | 價值 |
---|
StorageAccount
名字 | 描述 | 價值 |
---|---|---|
鑰匙 | 指定用來存取記憶體帳戶的金鑰。 | 字串 |
名字 | 指定記憶體帳戶的名稱。 | 字串 |
TableSpecification
名字 | 描述 | 價值 |
---|---|---|
描述 | Swagger 架構描述。 | 字串 |
格式 | 格式,如果 'type' 不是 'object' | 字串 |
性能 | 數據表內的數據行集。 | TableSpecificationProperties |
標題 | Swagger 架構標題。 | 字串 |
類型 | swagger 中所述之實體的類型。 | 字串 (必要) |
TableSpecificationProperties
名字 | 描述 | 價值 |
---|
WebServiceKeys
名字 | 描述 | 價值 |
---|---|---|
主要 | 主要存取金鑰。 | 字串 |
二 次 | 次要存取金鑰。 | 字串 |
WebServiceProperties
名字 | 描述 | 價值 |
---|---|---|
資產 | 包含描述 Web 服務資產的使用者定義屬性。 屬性會以索引鍵/值組表示。 | WebServicePropertiesAssets |
commitmentPlan | 包含與此 Web 服務相關聯的承諾方案。 在建立時設定。 設定之後,就無法變更此值。 注意:承諾用量計劃不會從對 GET 作業的呼叫傳回。 | CommitmentPlanAutoGenerated |
描述 | Web 服務的描述。 | 字串 |
診斷 | 控制 Web 服務的診斷追蹤集合的設定。 | DiagnosticsConfiguration |
exampleRequest | 定義一或多個服務輸入的範例輸入數據。 | ExampleRequest |
exposeSampleData | 當設定為 true 時,範例數據會包含在 Web 服務的 Swagger 定義中。 默認值為 true。 | bool |
輸入 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸入。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
鑰匙 | 包含 Web 服務佈建金鑰。 如果您未指定佈建密鑰,Azure Machine Learning 系統就會為您產生它們。 注意:金鑰不會從呼叫 GET 作業傳回。 | WebServiceKeys |
machineLearningWorkspace | 指定 Machine Learning 工作區,其中包含 Web 服務的來源實驗。 | MachineLearningWorkspace |
輸出 | 包含 Swagger 2.0 架構,描述一或多個 Web 服務的輸出。 如需詳細資訊,請參閱 Swagger 規格。 | ServiceInputOutputSpecification |
packageType | 將 類型設定為 'Graph',WebServicePropertiesForGraph。 | 'Graph' (必要) |
參數 | 為 Web 服務定義的一組全域參數值,指定為預設值對應的全域參數名稱。 如果未指定預設值,則會將 參數視為必要。 | WebServicePropertiesParameters |
readOnly | 當設定為 true 時,表示 Web 服務是唯讀的,而且無法再更新或修補,只能移除。 默認值為 false。 注意:一旦設定為 true,就無法變更其值。 | bool |
realtimeConfiguration | 包含 Web 服務端點的組態設定。 | RealtimeConfiguration |
storageAccount | 指定 Azure Machine Learning 用來儲存 Web 服務相關信息的記憶體帳戶。 只有記憶體帳戶的名稱會從呼叫 GET 作業傳回。 更新記憶體帳戶資訊時,您必須確定新的記憶體帳戶中提供所有必要的資產,或對 Web 服務的呼叫將會失敗。 | StorageAccount |
標題 | Web 服務的標題。 | 字串 |
WebServicePropertiesAssets
名字 | 描述 | 價值 |
---|
WebServicePropertiesForGraph
名字 | 描述 | 價值 |
---|---|---|
包 | 組成此 Web 服務的圖形套件定義。 | GraphPackage |
packageType | 指定封裝類型。 有效值為 Graph(指定透過 Machine Learning Studio 發佈的 Web 服務)和程式碼(指定使用 Python 之類的程式代碼發佈的 Web 服務)。 注意:目前不支持程序代碼。 | 'Graph' (必要) |
WebServicePropertiesParameters
名字 | 描述 | 價值 |
---|