Editar

Compartilhar via

New-AzMLWorkspaceBatchDeployment

  • Reference
Module:
Az.MachineLearningServices

Creates/updates a batch inference deployment (asynchronous).

Syntax

New-AzMLWorkspaceBatchDeployment
   -EndpointName <String>
   -Name <String>
   -ResourceGroupName <String>
   -WorkspaceName <String>
   [-SubscriptionId <String>]
   -Location <String>
   [-CodeId <String>]
   [-CodeScoringScript <String>]
   [-ComputeId <String>]
   [-DeploymentConfigurationType <BatchDeploymentConfigurationType>]
   [-Description <String>]
   [-EndpointDeploymentProperties <Hashtable>]
   [-EnvironmentId <String>]
   [-EnvironmentVariable <Hashtable>]
   [-ErrorThreshold <Int32>]
   [-IdentityType <ManagedServiceIdentityType>]
   [-IdentityUserAssigned <Hashtable>]
   [-Kind <String>]
   [-LoggingLevel <BatchLoggingLevel>]
   [-MaxConcurrencyPerInstance <Int32>]
   [-MiniBatchSize <Int64>]
   [-Model <IAssetReferenceBase>]
   [-OutputAction <BatchOutputAction>]
   [-OutputFileName <String>]
   [-ResourceInstanceCount <Int32>]
   [-ResourceInstanceType <String>]
   [-ResourceProperty <Hashtable>]
   [-RetryMax <Int32>]
   [-RetryTimeout <TimeSpan>]
   [-SkuCapacity <Int32>]
   [-SkuFamily <String>]
   [-SkuName <String>]
   [-SkuSize <String>]
   [-SkuTier <SkuTier>]
   [-Tag <Hashtable>]
   [-DefaultProfile <PSObject>]
   [-AsJob]
   [-NoWait]
   [-WhatIf]
   [-Confirm]
   [<CommonParameters>]

Description

Creates/updates a batch inference deployment (asynchronous).

Examples

Example 1: Creates/updates a batch inference deployment (asynchronous)

# The Reference Type includes Data Path, Output Path and Id.
# You can use following command to create it then pass it as value to Property parameter of the New-AzMLWorkspaceBatchDeployment cmdlet.
# New-AzMLWorkspaceIdAssetReferenceObject
# New-AzMLWorkspaceDataPathAssetReferenceObject
# New-AzMLWorkspaceOutputPathAssetReferenceObject
$model = New-AzMLWorkspaceIdAssetReferenceObject -AssetId '/subscriptions/11111111-2222-3333-4444-123456789101/resourceGroups/joyer-test/providers/Microsoft.MachineLearningServices/workspaces/mlworkspace-test002/models/openai-embeddings/versions/1' -ReferenceType 'Id'
New-AzMLWorkspaceBatchDeployment -ResourceGroupName ml-rg-test -WorkspaceName mlworkspace-cli01 -EndpointName batch-pwsh03 -Name nonmlflowdp -Location "eastus" `
-CodeId "/subscriptions/11111111-2222-3333-4444-123456789101/resourceGroups/ml-rg-test/providers/Microsoft.MachineLearningServices/workspaces/mlworkspace-cli01/codes/bd430754-fba7-4a63-a6b8-8ea8635767f3/versions/1" -CodeScoringScript "digit_identification.py" `
-EnvironmentId "/subscriptions/11111111-2222-3333-4444-123456789101/resourceGroups/ml-rg-test/providers/Microsoft.MachineLearningServices/workspaces/mlworkspace-cli01/environments/CliV2AnonymousEnvironment/versions/5d230430f302e7876f9b64710733f68e" `
-Model $model `
-ComputeId "/subscriptions/11111111-2222-3333-4444-123456789101/resourceGroups/ml-rg-test/providers/Microsoft.MachineLearningServices/workspaces/mlworkspace-cli01/computes/batch-cluster"

Location Name        SystemDataCreatedAt SystemDataCreatedBy                 SystemDataCreatedByType SystemDataLastModifiedAt SystemDataLastModifiedBy SystemDataLastModifiedByType  Kind ResourceGroupName
-------- ----        ------------------- -------------------                 ----------------------- ------------------------ ------------------------ ----------------------------  ---- -----------------
eastus   nonmlflowdp 6/1/2022 6:19:16 AM UserName (Example)                  6/1/2022 6:19:16 AM                                                                                          ml-rg-test

Creates/updates a batch inference deployment (asynchronous)

Parameters

-AsJob

Run the command as a job

Type:SwitchParameter
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-CodeId

ARM resource ID of the code asset.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-CodeScoringScript

[Required] The script to execute on startup. eg. "score.py"

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-ComputeId

Compute target for batch inference operation.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Confirm

Prompts you for confirmation before running the cmdlet.

Type:SwitchParameter
Aliases:cf
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-DefaultProfile

The DefaultProfile parameter is not functional. Use the SubscriptionId parameter when available if executing the cmdlet against a different subscription.

Type:PSObject
Aliases:AzureRMContext, AzureCredential
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-DeploymentConfigurationType

[Required] The type of the deployment

Type:BatchDeploymentConfigurationType
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Description

Description of the endpoint deployment.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-EndpointDeploymentProperties

Property dictionary. Properties can be added, but not removed or altered.

Type:Hashtable
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-EndpointName

Inference endpoint name

Type:String
Position:Named
Default value:None
Required:True
Accept pipeline input:False
Accept wildcard characters:False

-EnvironmentId

ARM resource ID or AssetId of the environment specification for the endpoint deployment.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-EnvironmentVariable

Environment variables configuration for the deployment.

Type:Hashtable
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-ErrorThreshold

Error threshold, if the error count for the entire input goes above this value,the batch inference will be aborted. Range is [-1, int.MaxValue].For FileDataset, this value is the count of file failures.For TabularDataset, this value is the count of record failures.If set to -1 (the lower bound), all failures during batch inference will be ignored.

Type:Int32
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-IdentityType

Type of managed service identity (where both SystemAssigned and UserAssigned types are allowed).

Type:ManagedServiceIdentityType
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-IdentityUserAssigned

The set of user assigned identities associated with the resource. The userAssignedIdentities dictionary keys will be ARM resource ids in the form: '/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.ManagedIdentity/userAssignedIdentities/{identityName}. The dictionary values can be empty objects ({}) in requests.

Type:Hashtable
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Kind

Metadata used by portal/tooling/etc to render different UX experiences for resources of the same type.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Location

The geo-location where the resource lives

Type:String
Position:Named
Default value:None
Required:True
Accept pipeline input:False
Accept wildcard characters:False

-LoggingLevel

Logging level for batch inference operation.

Type:BatchLoggingLevel
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-MaxConcurrencyPerInstance

Indicates maximum number of parallelism per instance.

Type:Int32
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-MiniBatchSize

Size of the mini-batch passed to each batch invocation.For FileDataset, this is the number of files per mini-batch.For TabularDataset, this is the size of the records in bytes, per mini-batch.

Type:Int64
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Model

Reference to the model asset for the endpoint deployment. To construct, see NOTES section for MODEL properties and create a hash table.

Type:IAssetReferenceBase
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Name

The identifier for the Batch inference deployment.

Type:String
Position:Named
Default value:None
Required:True
Accept pipeline input:False
Accept wildcard characters:False

-NoWait

Run the command asynchronously

Type:SwitchParameter
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-OutputAction

Indicates how the output will be organized.

Type:BatchOutputAction
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-OutputFileName

Customized output file name for append_row output action.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-ResourceGroupName

The name of the resource group. The name is case insensitive.

Type:String
Position:Named
Default value:None
Required:True
Accept pipeline input:False
Accept wildcard characters:False

-ResourceInstanceCount

Optional number of instances or nodes used by the compute target.

Type:Int32
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-ResourceInstanceType

Optional type of VM used as supported by the compute target.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-ResourceProperty

Additional properties bag.

Type:Hashtable
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-RetryMax

Maximum retry count for a mini-batch

Type:Int32
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-RetryTimeout

Invocation timeout for a mini-batch, in ISO 8601 format.

Type:TimeSpan
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SkuCapacity

If the SKU supports scale out/in then the capacity integer should be included. If scale out/in is not possible for the resource this may be omitted.

Type:Int32
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SkuFamily

If the service has different generations of hardware, for the same SKU, then that can be captured here.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SkuName

The name of the SKU. Ex - P3. It is typically a letter+number code

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SkuSize

The SKU size. When the name field is the combination of tier and some other value, this would be the standalone code.

Type:String
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SkuTier

This field is required to be implemented by the Resource Provider if the service has more than one tier, but is not required on a PUT.

Type:SkuTier
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-SubscriptionId

The ID of the target subscription.

Type:String
Position:Named
Default value:(Get-AzContext).Subscription.Id
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-Tag

Resource tags.

Type:Hashtable
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-WhatIf

Shows what would happen if the cmdlet runs. The cmdlet is not run.

Type:SwitchParameter
Aliases:wi
Position:Named
Default value:None
Required:False
Accept pipeline input:False
Accept wildcard characters:False

-WorkspaceName

Name of Azure Machine Learning workspace.

Type:String
Position:Named
Default value:None
Required:True
Accept pipeline input:False
Accept wildcard characters:False

Outputs

IBatchDeployment