aws-cdk-lib.aws_stepfunctions_tasks.SageMakerCreateTrainingJobJsonataProps

SageMakerCreateTrainingJobJsonataProps Struct

Properties for creating an Amazon SageMaker training job using JSONata.

SageMakerCreateTrainingJobJsonataProps

extends

• TaskStateJsonataBaseProps

Source

aws-stepfunctions-tasks/lib/sagemaker/create-training-job.ts

Type last changed

v2.178.0 · 6 Feb 2025

Example

// The code below shows an example of how to instantiate this type.
// The values are placeholders you should change.
import * as cdk from 'aws-cdk-lib';
import { aws_ec2 as ec2 } from 'aws-cdk-lib';
import { aws_iam as iam } from 'aws-cdk-lib';
import { aws_kms as kms } from 'aws-cdk-lib';
import { aws_stepfunctions as stepfunctions } from 'aws-cdk-lib';
import { aws_stepfunctions_tasks as stepfunctions_tasks } from 'aws-cdk-lib';

declare const assign: any;
declare const dockerImage: stepfunctions_tasks.DockerImage;
declare const hyperparameters: any;
declare const instanceType: ec2.InstanceType;
declare const key: kms.Key;
declare const outputs: any;
declare const role: iam.Role;
declare const s3Location: stepfunctions_tasks.S3Location;
declare const size: cdk.Size;
declare const subnet: ec2.Subnet;
declare const subnetFilter: ec2.SubnetFilter;
declare const taskRole: stepfunctions.TaskRole;
declare const timeout: stepfunctions.Timeout;
declare const vpc: ec2.Vpc;
const sageMakerCreateTrainingJobJsonataProps: stepfunctions_tasks.SageMakerCreateTrainingJobJsonataProps = {
  algorithmSpecification: {
    algorithmName: 'algorithmName',
    metricDefinitions: [{
      name: 'name',
      regex: 'regex',
    }],
    trainingImage: dockerImage,
    trainingInputMode: stepfunctions_tasks.InputMode.PIPE,
  },
  outputDataConfig: {
    s3OutputLocation: s3Location,

    // the properties below are optional
    encryptionKey: key,
  },
  trainingJobName: 'trainingJobName',

  // the properties below are optional
  assign: {
    assignKey: assign,
  },
  comment: 'comment',
  credentials: {
    role: taskRole,
  },
  enableNetworkIsolation: false,
  environment: {
    environmentKey: 'environment',
  },
  heartbeat: cdk.Duration.minutes(30),
  heartbeatTimeout: timeout,
  hyperparameters: {
    hyperparametersKey: hyperparameters,
  },
  inputDataConfig: [{
    channelName: 'channelName',
    dataSource: {
      s3DataSource: {
        s3Location: s3Location,

        // the properties below are optional
        attributeNames: ['attributeNames'],
        s3DataDistributionType: stepfunctions_tasks.S3DataDistributionType.FULLY_REPLICATED,
        s3DataType: stepfunctions_tasks.S3DataType.MANIFEST_FILE,
      },
    },

    // the properties below are optional
    compressionType: stepfunctions_tasks.CompressionType.NONE,
    contentType: 'contentType',
    inputMode: stepfunctions_tasks.InputMode.PIPE,
    recordWrapperType: stepfunctions_tasks.RecordWrapperType.NONE,
    shuffleConfig: {
      seed: 123,
    },
  }],
  integrationPattern: stepfunctions.IntegrationPattern.REQUEST_RESPONSE,
  outputs: outputs,
  queryLanguage: stepfunctions.QueryLanguage.JSON_PATH,
  resourceConfig: {
    instanceCount: 123,
    instanceType: instanceType,
    volumeSize: size,

    // the properties below are optional
    volumeEncryptionKey: key,
  },
  role: role,
  stateName: 'stateName',
  stoppingCondition: {
    maxRuntime: cdk.Duration.minutes(30),
  },
  tags: {
    tagsKey: 'tags',
  },
  taskTimeout: timeout,
  timeout: cdk.Duration.minutes(30),
  vpcConfig: {
    vpc: vpc,

    // the properties below are optional
    subnets: {
      availabilityZones: ['availabilityZones'],
      onePerAz: false,
      subnetFilters: [subnetFilter],
      subnetGroupName: 'subnetGroupName',
      subnets: [subnet],
      subnetType: ec2.SubnetType.PRIVATE_ISOLATED,
    },
  },
};

Properties

algorithmSpecificationAlgorithmSpecificationalgorithm_specificationAlgorithmSpecificationalgorithmSpecificationAlgorithmSpecificationAlgorithmSpecificationAlgorithmSpecificationAlgorithmSpecificationAlgorithmSpecificationrequired#

Identifies the training algorithm to use.

assign?optional{ [key: string]: any }assign?optionaltyping.Mapping[str, Any]assign?optionalMap<String, Object>Assign?optionalIDictionary<string, object>Assign?optional*map[string]interface{}#

Workflow variables to store in this step.

Default: Not assign variables

Using workflow variables, you can store data in a step and retrieve that data in future steps.

comment?optionalstringcomment?optionalstrcomment?optionalStringComment?optionalstringComment?optional*string#

A comment describing this state.

Default: No comment

credentials?optionalCredentialscredentials?optionalCredentialscredentials?optionalCredentialsCredentials?optionalCredentialsCredentials?optionalCredentials#

Credentials for an IAM Role that the State Machine assumes for executing the task.

Default: None (Task is executed using the State Machine's execution role)

This enables cross-account resource invocations.

enableNetworkIsolation?optionalbooleanenable_network_isolation?optionalboolenableNetworkIsolation?optionalBooleanEnableNetworkIsolation?optionalboolEnableNetworkIsolation?optional*bool= false#

Isolates the training container.

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

environment?optional{ [key: string]: string }environment?optionaltyping.Mapping[str, str]environment?optionalMap<String, String>Environment?optionalIDictionary<string, string>Environment?optional*map[string]*string#

Environment variables to set in the Docker container.

Default: No environment variables

heartbeat?optionalDurationheartbeat?optionalDurationheartbeat?optionalDurationHeartbeat?optionalDurationHeartbeat?optionalDuration= Nonedeprecated#

use `heartbeatTimeout`

heartbeatTimeout?optionalTimeoutheartbeat_timeout?optionalTimeoutheartbeatTimeout?optionalTimeoutHeartbeatTimeout?optionalTimeoutHeartbeatTimeout?optionalTimeout= None#

Timeout for the heartbeat.

[disable-awslint:duration-prop-type] is needed because all props interface in aws-stepfunctions-tasks extend this interface

hyperparameters?optional{ [key: string]: any }hyperparameters?optionaltyping.Mapping[str, Any]hyperparameters?optionalMap<String, Object>Hyperparameters?optionalIDictionary<string, object>Hyperparameters?optional*map[string]interface{}#

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

Default: No hyperparameters

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

inputDataConfig?optionalChannel[]input_data_config?optionaltyping.List[Channel]inputDataConfig?optionalList<Channel>InputDataConfig?optionalChannel[]InputDataConfig?optional*[]Channel#

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

Default: No inputDataConfig

integrationPattern?optionalIntegrationPatternintegration_pattern?optionalIntegrationPatternintegrationPattern?optionalIntegrationPatternIntegrationPattern?optionalIntegrationPatternIntegrationPattern?optionalIntegrationPattern#

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

Default: IntegrationPattern.REQUEST_RESPONSE for most tasks. IntegrationPattern.RUN_JOB for the following exceptions: BatchSubmitJob, EmrAddStep, EmrCreateCluster, EmrTerminationCluster, and EmrContainersStartJobRun.

You can control these AWS services using service integration patterns.

Depending on the AWS Service, the Service Integration Pattern availability will vary.

outputDataConfigOutputDataConfigoutput_data_configOutputDataConfigoutputDataConfigOutputDataConfigOutputDataConfigOutputDataConfigOutputDataConfigOutputDataConfigrequired#

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

outputs?optionalanyoutputs?optionalAnyoutputs?optionalObjectOutputs?optionalobjectOutputs?optionalinterface{}#

Used to specify and transform output from the state.

Default: $states.result or $states.errorOutput

notes

When specified, the value overrides the state output default. The output field accepts any JSON value (object, array, string, number, boolean, null). Any string value, including those inside objects or arrays, will be evaluated as JSONata if surrounded by {% %} characters. Output also accepts a JSONata expression directly.

queryLanguage?optionalQueryLanguagequery_language?optionalQueryLanguagequeryLanguage?optionalQueryLanguageQueryLanguage?optionalQueryLanguageQueryLanguage?optionalQueryLanguage= JSONPath#

The name of the query language used by the state.

If the state does not contain a queryLanguage field, then it will use the query language specified in the top-level queryLanguage field.

resourceConfig?optionalResourceConfigresource_config?optionalResourceConfigresourceConfig?optionalResourceConfigResourceConfig?optionalResourceConfigResourceConfig?optionalResourceConfig#

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?optionalIRolerole?optionalIRolerole?optionalIRoleRole?optionalIRoleRole?optionalIRole#

Role for the Training Job.

Default: a role will be created.

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

stateName?optionalstringstate_name?optionalstrstateName?optionalStringStateName?optionalstringStateName?optional*string#

Optional name for this state.

Default: The construct ID will be used as state name

stoppingCondition?optionalStoppingConditionstopping_condition?optionalStoppingConditionstoppingCondition?optionalStoppingConditionStoppingCondition?optionalStoppingConditionStoppingCondition?optionalStoppingCondition#

Sets a time limit for training.

Default: max runtime of 1 hour

tags?optional{ [key: string]: string }tags?optionaltyping.Mapping[str, str]tags?optionalMap<String, String>Tags?optionalIDictionary<string, string>Tags?optional*map[string]*string#

Tags to be applied to the train job.

Default: No tags

taskTimeout?optionalTimeouttask_timeout?optionalTimeouttaskTimeout?optionalTimeoutTaskTimeout?optionalTimeoutTaskTimeout?optionalTimeout= None#

Timeout for the task.

[disable-awslint:duration-prop-type] is needed because all props interface in aws-stepfunctions-tasks extend this interface

timeout?optionalDurationtimeout?optionalDurationtimeout?optionalDurationTimeout?optionalDurationTimeout?optionalDuration= Nonedeprecated#

use `taskTimeout`

trainingJobNamestringtraining_job_namestrtrainingJobNameStringTrainingJobNamestringTrainingJobName*stringrequired#

Training Job Name.

vpcConfig?optionalVpcConfigvpc_config?optionalVpcConfigvpcConfig?optionalVpcConfigVpcConfig?optionalVpcConfigVpcConfig?optionalVpcConfig#

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

Default: No VPC