SageMakerCreateTrainingJobProps¶

class aws_cdk.aws_stepfunctions_tasks.SageMakerCreateTrainingJobProps(*, comment=None, heartbeat=None, input_path=None, integration_pattern=None, output_path=None, result_path=None, result_selector=None, timeout=None, algorithm_specification, input_data_config, output_data_config, training_job_name, enable_network_isolation=None, environment=None, hyperparameters=None, resource_config=None, role=None, stopping_condition=None, tags=None, vpc_config=None)¶

Bases: aws_cdk.aws_stepfunctions.TaskStateBaseProps

Properties for creating an Amazon SageMaker training job.

Parameters

comment (Optional[str]) – An optional description for this state. Default: - No comment
heartbeat (Optional[Duration]) – Timeout for the heartbeat. Default: - None
input_path (Optional[str]) – JSONPath expression to select part of the state to be the input to this state. May also be the special value JsonPath.DISCARD, which will cause the effective input to be the empty object {}. Default: - The entire task input (JSON path ‘$’)
integration_pattern (Optional[IntegrationPattern]) – AWS Step Functions integrates with services directly in the Amazon States Language. You can control these AWS services using service integration patterns Default: - IntegrationPattern.REQUEST_RESPONSE for most tasks. IntegrationPattern.RUN_JOB for the following exceptions: BatchSubmitJob, EmrAddStep, EmrCreateCluster, EmrTerminationCluster, and EmrContainersStartJobRun.
output_path (Optional[str]) – JSONPath expression to select select a portion of the state output to pass to the next state. May also be the special value JsonPath.DISCARD, which will cause the effective output to be the empty object {}. Default: - The entire JSON node determined by the state input, the task result, and resultPath is passed to the next state (JSON path ‘$’)
result_path (Optional[str]) – JSONPath expression to indicate where to inject the state’s output. May also be the special value JsonPath.DISCARD, which will cause the state’s input to become its output. Default: - Replaces the entire input with the result (JSON path ‘$’)
result_selector (Optional[Mapping[str, Any]]) – The JSON that will replace the state’s raw result and become the effective result before ResultPath is applied. You can use ResultSelector to create a payload with values that are static or selected from the state’s raw result. Default: - None
timeout (Optional[Duration]) – Timeout for the state machine. Default: - None
algorithm_specification (Union[AlgorithmSpecification, Dict[str, Any]]) – Identifies the training algorithm to use.
input_data_config (Sequence[Union[Channel, Dict[str, Any]]]) – Describes the various datasets (e.g. train, validation, test) and the Amazon S3 location where stored.
output_data_config (Union[OutputDataConfig, Dict[str, Any]]) – Identifies the Amazon S3 location where you want Amazon SageMaker to save the results of model training.
training_job_name (str) – Training Job Name.
enable_network_isolation (Optional[bool]) – Isolates the training container. No inbound or outbound network calls can be made to or from the training container. Default: false
environment (Optional[Mapping[str, str]]) – Environment variables to set in the Docker container. Default: - No environment variables
hyperparameters (Optional[Mapping[str, Any]]) – Algorithm-specific parameters that influence the quality of the model. Set hyperparameters before you start the learning process. For a list of hyperparameters provided by Amazon SageMaker Default: - No hyperparameters
resource_config (Union[ResourceConfig, Dict[str, Any], None]) – Specifies the resources, ML compute instances, and ML storage volumes to deploy for model training. Default: - 1 instance of EC2 M4.XLarge with 10GB volume
role (Optional[IRole]) – Role for the Training Job. The role must be granted all necessary permissions for the SageMaker training job to be able to operate. See https://docs.aws.amazon.com/fr_fr/sagemaker/latest/dg/sagemaker-roles.html#sagemaker-roles-createtrainingjob-perms Default: - a role will be created.
stopping_condition (Union[StoppingCondition, Dict[str, Any], None]) – Sets a time limit for training. Default: - max runtime of 1 hour
tags (Optional[Mapping[str, str]]) – Tags to be applied to the train job. Default: - No tags
vpc_config (Union[VpcConfig, Dict[str, Any], None]) – Specifies the VPC that you want your training job to connect to. Default: - No VPC

ExampleMetadata

infused

Example:

tasks.SageMakerCreateTrainingJob(self, "TrainSagemaker",
    training_job_name=sfn.JsonPath.string_at("$.JobName"),
    algorithm_specification=tasks.AlgorithmSpecification(
        algorithm_name="BlazingText",
        training_input_mode=tasks.InputMode.FILE
    ),
    input_data_config=[tasks.Channel(
        channel_name="train",
        data_source=tasks.DataSource(
            s3_data_source=tasks.S3DataSource(
                s3_data_type=tasks.S3DataType.S3_PREFIX,
                s3_location=tasks.S3Location.from_json_expression("$.S3Bucket")
            )
        )
    )],
    output_data_config=tasks.OutputDataConfig(
        s3_output_location=tasks.S3Location.from_bucket(s3.Bucket.from_bucket_name(self, "Bucket", "mybucket"), "myoutputpath")
    ),
    resource_config=tasks.ResourceConfig(
        instance_count=1,
        instance_type=ec2.InstanceType(sfn.JsonPath.string_at("$.InstanceType")),
        volume_size=Size.gibibytes(50)
    ),  # optional: default is 1 instance of EC2 `M4.XLarge` with `10GB` volume
    stopping_condition=tasks.StoppingCondition(
        max_runtime=Duration.hours(2)
    )
)

Attributes

algorithm_specification¶

Identifies the training algorithm to use.

Return type: AlgorithmSpecification

comment¶

An optional description for this state.

Default

No comment

Return type

Optional[str]

enable_network_isolation¶

Isolates the training container.

No inbound or outbound network calls can be made to or from the training container.

Default: false
Return type: Optional[bool]

environment¶

Environment variables to set in the Docker container.

Default

No environment variables

Return type

Optional[Mapping[str, str]]

heartbeat¶

Timeout for the heartbeat.

Default

None

Return type

Optional[Duration]

hyperparameters¶

Algorithm-specific parameters that influence the quality of the model.

Set hyperparameters before you start the learning process. For a list of hyperparameters provided by Amazon SageMaker

Default

No hyperparameters

See

https://docs.aws.amazon.com/sagemaker/latest/dg/algos.html

Return type

Optional[Mapping[str, Any]]

input_data_config¶

Describes the various datasets (e.g. train, validation, test) and the Amazon S3 location where stored.

Return type: List[Channel]

input_path¶

JSONPath expression to select part of the state to be the input to this state.

May also be the special value JsonPath.DISCARD, which will cause the effective input to be the empty object {}.

Default

The entire task input (JSON path ‘$’)

Return type

Optional[str]

integration_pattern¶

AWS Step Functions integrates with services directly in the Amazon States Language.

You can control these AWS services using service integration patterns

Default

IntegrationPattern.REQUEST_RESPONSE for most tasks.

IntegrationPattern.RUN_JOB for the following exceptions: BatchSubmitJob, EmrAddStep, EmrCreateCluster, EmrTerminationCluster, and EmrContainersStartJobRun.

See: https://docs.aws.amazon.com/step-functions/latest/dg/connect-to-resource.html#connect-wait-token
Return type: Optional[IntegrationPattern]

output_data_config¶

Identifies the Amazon S3 location where you want Amazon SageMaker to save the results of model training.

Return type: OutputDataConfig

output_path¶

JSONPath expression to select select a portion of the state output to pass to the next state.

May also be the special value JsonPath.DISCARD, which will cause the effective output to be the empty object {}.

Default

The entire JSON node determined by the state input, the task result,

and resultPath is passed to the next state (JSON path ‘$’)

Return type: Optional[str]

resource_config¶

Specifies the resources, ML compute instances, and ML storage volumes to deploy for model training.

Default

1 instance of EC2 M4.XLarge with 10GB volume

Return type

Optional[ResourceConfig]

result_path¶

JSONPath expression to indicate where to inject the state’s output.

May also be the special value JsonPath.DISCARD, which will cause the state’s input to become its output.

Default

Replaces the entire input with the result (JSON path ‘$’)

Return type

Optional[str]

result_selector¶

The JSON that will replace the state’s raw result and become the effective result before ResultPath is applied.

You can use ResultSelector to create a payload with values that are static or selected from the state’s raw result.

Default

None

See

https://docs.aws.amazon.com/step-functions/latest/dg/input-output-inputpath-params.html#input-output-resultselector

Return type

Optional[Mapping[str, Any]]

role¶

Role for the Training Job.

The role must be granted all necessary permissions for the SageMaker training job to be able to operate.

See https://docs.aws.amazon.com/fr_fr/sagemaker/latest/dg/sagemaker-roles.html#sagemaker-roles-createtrainingjob-perms

Default

a role will be created.

Return type

Optional[IRole]

stopping_condition¶

Sets a time limit for training.

Default

max runtime of 1 hour

Return type

Optional[StoppingCondition]

tags¶

Tags to be applied to the train job.

Default

No tags

Return type

Optional[Mapping[str, str]]

timeout¶

Timeout for the state machine.

Default

None

Return type

Optional[Duration]

training_job_name¶

Training Job Name.

Return type: str

vpc_config¶

Specifies the VPC that you want your training job to connect to.

Default

No VPC

Return type

Optional[VpcConfig]