Interface ISageMakerCreateTransformJobProps
Properties for creating an Amazon SageMaker transform job task.
Inherited Members
Namespace: Amazon.CDK.AWS.StepFunctions.Tasks
Assembly: Amazon.CDK.Lib.dll
Syntax (csharp)
public interface ISageMakerCreateTransformJobProps : ITaskStateBaseProps
Syntax (vb)
Public Interface ISageMakerCreateTransformJobProps
Inherits ITaskStateBaseProps
Remarks
ExampleMetadata: infused
Examples
new SageMakerCreateTransformJob(this, "Batch Inference", new SageMakerCreateTransformJobProps {
TransformJobName = "MyTransformJob",
ModelName = "MyModelName",
ModelClientOptions = new ModelClientOptions {
InvocationsMaxRetries = 3, // default is 0
InvocationsTimeout = Duration.Minutes(5)
},
TransformInput = new TransformInput {
TransformDataSource = new TransformDataSource {
S3DataSource = new TransformS3DataSource {
S3Uri = "s3://inputbucket/train",
S3DataType = S3DataType.S3_PREFIX
}
}
},
TransformOutput = new TransformOutput {
S3OutputPath = "s3://outputbucket/TransformJobOutputPath"
},
TransformResources = new TransformResources {
InstanceCount = 1,
InstanceType = InstanceType.Of(InstanceClass.M4, InstanceSize.XLARGE)
}
});
Synopsis
Properties
BatchStrategy | Number of records to include in a mini-batch for an HTTP inference request. |
Environment | Environment variables to set in the Docker container. |
MaxConcurrentTransforms | Maximum number of parallel requests that can be sent to each instance in a transform job. |
MaxPayload | Maximum allowed size of the payload, in MB. |
ModelClientOptions | Configures the timeout and maximum number of retries for processing a transform job invocation. |
ModelName | Name of the model that you want to use for the transform job. |
Role | Role for the Transform Job. |
Tags | Tags to be applied to the train job. |
TransformInput | Dataset to be transformed and the Amazon S3 location where it is stored. |
TransformJobName | Transform Job Name. |
TransformOutput | S3 location where you want Amazon SageMaker to save the results from the transform job. |
TransformResources | ML compute instances for the transform job. |
Properties
BatchStrategy
Number of records to include in a mini-batch for an HTTP inference request.
virtual Nullable<BatchStrategy> BatchStrategy { get; }
Property Value
System.Nullable<BatchStrategy>
Remarks
Default: - No batch strategy
Environment
Environment variables to set in the Docker container.
virtual IDictionary<string, string> Environment { get; }
Property Value
System.Collections.Generic.IDictionary<System.String, System.String>
Remarks
Default: - No environment variables
MaxConcurrentTransforms
Maximum number of parallel requests that can be sent to each instance in a transform job.
virtual Nullable<double> MaxConcurrentTransforms { get; }
Property Value
System.Nullable<System.Double>
Remarks
Default: - Amazon SageMaker checks the optional execution-parameters to determine the settings for your chosen algorithm. If the execution-parameters endpoint is not enabled, the default value is 1.
MaxPayload
Maximum allowed size of the payload, in MB.
virtual Size MaxPayload { get; }
Property Value
Remarks
Default: 6
ModelClientOptions
Configures the timeout and maximum number of retries for processing a transform job invocation.
virtual IModelClientOptions ModelClientOptions { get; }
Property Value
Remarks
Default: - 0 retries and 60 seconds of timeout
ModelName
Name of the model that you want to use for the transform job.
string ModelName { get; }
Property Value
System.String
Role
Role for the Transform Job.
virtual IRole Role { get; }
Property Value
Remarks
Default: - A role is created with AmazonSageMakerFullAccess
managed policy
Tags
Tags to be applied to the train job.
virtual IDictionary<string, string> Tags { get; }
Property Value
System.Collections.Generic.IDictionary<System.String, System.String>
Remarks
Default: - No tags
TransformInput
Dataset to be transformed and the Amazon S3 location where it is stored.
ITransformInput TransformInput { get; }
Property Value
TransformJobName
Transform Job Name.
string TransformJobName { get; }
Property Value
System.String
TransformOutput
S3 location where you want Amazon SageMaker to save the results from the transform job.
ITransformOutput TransformOutput { get; }
Property Value
TransformResources
ML compute instances for the transform job.
virtual ITransformResources TransformResources { get; }
Property Value
Remarks
Default: - 1 instance of type M4.XLarge