ModelExplainabilityJobDefinition

The ModelExplainabilityJobDefinition resource allows you to create and manage model explainability jobs in AWS SageMaker for your machine learning models. For more detailed information, please refer to the AWS SageMaker ModelExplainabilityJobDefinitions documentation.

Minimal Example

This example demonstrates how to create a basic ModelExplainabilityJobDefinition with required properties and a common optional property.

import AWS from "alchemy/aws/control";

const modelExplainabilityJob = await AWS.SageMaker.ModelExplainabilityJobDefinition("basicExplainabilityJob", {
  ModelExplainabilityJobOutputConfig: {
    S3OutputPath: "s3://my-bucket/explainability-output/"
  },
  RoleArn: "arn:aws:iam::123456789012:role/SageMakerExecutionRole",
  ModelExplainabilityJobInput: {
    EndpointName: "my-endpoint",
    S3DataSource: {
      S3DataType: "ManifestFile",
      S3Uri: "s3://my-bucket/data/input/"
    }
  },
  JobResources: {
    ClusterConfig: {
      InstanceType: "ml.m5.large",
      InstanceCount: 1,
      VolumeSizeInGB: 30
    }
  }
});

Advanced Configuration

This example demonstrates how to configure the ModelExplainabilityJobDefinition with additional advanced settings, such as a stopping condition and network configuration.

const advancedExplainabilityJob = await AWS.SageMaker.ModelExplainabilityJobDefinition("advancedExplainabilityJob", {
  ModelExplainabilityJobOutputConfig: {
    S3OutputPath: "s3://my-bucket/explainability-output/"
  },
  RoleArn: "arn:aws:iam::123456789012:role/SageMakerExecutionRole",
  ModelExplainabilityJobInput: {
    EndpointName: "my-endpoint",
    S3DataSource: {
      S3DataType: "ManifestFile",
      S3Uri: "s3://my-bucket/data/input/"
    }
  },
  JobResources: {
    ClusterConfig: {
      InstanceType: "ml.m5.large",
      InstanceCount: 1,
      VolumeSizeInGB: 30
    }
  },
  StoppingCondition: {
    MaxRuntimeInSeconds: 3600
  },
  NetworkConfig: {
    EnableInterContainerTrafficEncryption: true,
    VpcConfig: {
      SecurityGroupIds: ["sg-0123456789abcdef0"],
      Subnets: ["subnet-0123456789abcdef0"]
    }
  }
});

Using Baseline Configuration

This example shows how to include a ModelExplainabilityBaselineConfig for baseline metrics to compare against.

const baselineExplainabilityJob = await AWS.SageMaker.ModelExplainabilityJobDefinition("baselineExplainabilityJob", {
  ModelExplainabilityJobOutputConfig: {
    S3OutputPath: "s3://my-bucket/explainability-output/"
  },
  RoleArn: "arn:aws:iam::123456789012:role/SageMakerExecutionRole",
  ModelExplainabilityJobInput: {
    EndpointName: "my-endpoint",
    S3DataSource: {
      S3DataType: "ManifestFile",
      S3Uri: "s3://my-bucket/data/input/"
    }
  },
  ModelExplainabilityBaselineConfig: {
    BaselineConfig: {
      S3Uri: "s3://my-bucket/baseline/",
      BaselineType: "Model"
    }
  },
  JobResources: {
    ClusterConfig: {
      InstanceType: "ml.m5.large",
      InstanceCount: 1,
      VolumeSizeInGB: 30
    }
  }
});