Amazon SageMaker
Developer Guide

The AWS Documentation website is getting a new look!
Try it now and let us know what you think. Switch to the new look >>

You can return to the original look by selecting English in the language selector above.

CreateAlgorithm

Create a machine learning algorithm that you can use in Amazon SageMaker and list in the AWS Marketplace.

Request Syntax

{ "AlgorithmDescription": "string", "AlgorithmName": "string", "CertifyForMarketplace": boolean, "InferenceSpecification": { "Containers": [ { "ContainerHostname": "string", "Image": "string", "ImageDigest": "string", "ModelDataUrl": "string", "ProductId": "string" } ], "SupportedContentTypes": [ "string" ], "SupportedRealtimeInferenceInstanceTypes": [ "string" ], "SupportedResponseMIMETypes": [ "string" ], "SupportedTransformInstanceTypes": [ "string" ] }, "TrainingSpecification": { "MetricDefinitions": [ { "Name": "string", "Regex": "string" } ], "SupportedHyperParameters": [ { "DefaultValue": "string", "Description": "string", "IsRequired": boolean, "IsTunable": boolean, "Name": "string", "Range": { "CategoricalParameterRangeSpecification": { "Values": [ "string" ] }, "ContinuousParameterRangeSpecification": { "MaxValue": "string", "MinValue": "string" }, "IntegerParameterRangeSpecification": { "MaxValue": "string", "MinValue": "string" } }, "Type": "string" } ], "SupportedTrainingInstanceTypes": [ "string" ], "SupportedTuningJobObjectiveMetrics": [ { "MetricName": "string", "Type": "string" } ], "SupportsDistributedTraining": boolean, "TrainingChannels": [ { "Description": "string", "IsRequired": boolean, "Name": "string", "SupportedCompressionTypes": [ "string" ], "SupportedContentTypes": [ "string" ], "SupportedInputModes": [ "string" ] } ], "TrainingImage": "string", "TrainingImageDigest": "string" }, "ValidationSpecification": { "ValidationProfiles": [ { "ProfileName": "string", "TrainingJobDefinition": { "HyperParameters": { "string" : "string" }, "InputDataConfig": [ { "ChannelName": "string", "CompressionType": "string", "ContentType": "string", "DataSource": { "FileSystemDataSource": { "DirectoryPath": "string", "FileSystemAccessMode": "string", "FileSystemId": "string", "FileSystemType": "string" }, "S3DataSource": { "AttributeNames": [ "string" ], "S3DataDistributionType": "string", "S3DataType": "string", "S3Uri": "string" } }, "InputMode": "string", "RecordWrapperType": "string", "ShuffleConfig": { "Seed": number } } ], "OutputDataConfig": { "KmsKeyId": "string", "S3OutputPath": "string" }, "ResourceConfig": { "InstanceCount": number, "InstanceType": "string", "VolumeKmsKeyId": "string", "VolumeSizeInGB": number }, "StoppingCondition": { "MaxRuntimeInSeconds": number, "MaxWaitTimeInSeconds": number }, "TrainingInputMode": "string" }, "TransformJobDefinition": { "BatchStrategy": "string", "Environment": { "string" : "string" }, "MaxConcurrentTransforms": number, "MaxPayloadInMB": number, "TransformInput": { "CompressionType": "string", "ContentType": "string", "DataSource": { "S3DataSource": { "S3DataType": "string", "S3Uri": "string" } }, "SplitType": "string" }, "TransformOutput": { "Accept": "string", "AssembleWith": "string", "KmsKeyId": "string", "S3OutputPath": "string" }, "TransformResources": { "InstanceCount": number, "InstanceType": "string", "VolumeKmsKeyId": "string" } } } ], "ValidationRole": "string" } }

Request Parameters

For information about the parameters that are common to all actions, see Common Parameters.

The request accepts the following data in JSON format.

AlgorithmDescription

A description of the algorithm.

Type: String

Length Constraints: Maximum length of 1024.

Pattern: [\p{L}\p{M}\p{Z}\p{S}\p{N}\p{P}]*

Required: No

AlgorithmName

The name of the algorithm.

Type: String

Length Constraints: Minimum length of 1. Maximum length of 63.

Pattern: ^[a-zA-Z0-9](-*[a-zA-Z0-9])*$

Required: Yes

CertifyForMarketplace

Whether to certify the algorithm so that it can be listed in AWS Marketplace.

Type: Boolean

Required: No

InferenceSpecification

Specifies details about inference jobs that the algorithm runs, including the following:

  • The Amazon ECR paths of containers that contain the inference code and model artifacts.

  • The instance types that the algorithm supports for transform jobs and real-time endpoints used for inference.

  • The input and output content formats that the algorithm supports for inference.

Type: InferenceSpecification object

Required: No

TrainingSpecification

Specifies details about training jobs run by this algorithm, including the following:

  • The Amazon ECR path of the container and the version digest of the algorithm.

  • The hyperparameters that the algorithm supports.

  • The instance types that the algorithm supports for training.

  • Whether the algorithm supports distributed training.

  • The metrics that the algorithm emits to Amazon CloudWatch.

  • Which metrics that the algorithm emits can be used as the objective metric for hyperparameter tuning jobs.

  • The input channels that the algorithm supports for training data. For example, an algorithm might support train, validation, and test channels.

Type: TrainingSpecification object

Required: Yes

ValidationSpecification

Specifies configurations for one or more training jobs and that Amazon SageMaker runs to test the algorithm's training code and, optionally, one or more batch transform jobs that Amazon SageMaker runs to test the algorithm's inference code.

Type: AlgorithmValidationSpecification object

Required: No

Response Syntax

{ "AlgorithmArn": "string" }

Response Elements

If the action is successful, the service sends back an HTTP 200 response.

The following data is returned in JSON format by the service.

AlgorithmArn

The Amazon Resource Name (ARN) of the new algorithm.

Type: String

Length Constraints: Minimum length of 1. Maximum length of 2048.

Pattern: arn:aws[a-z\-]*:sagemaker:[a-z0-9\-]*:[0-9]{12}:algorithm/.*

Errors

For information about the errors that are common to all actions, see Common Errors.

See Also

For more information about using this API in one of the language-specific AWS SDKs, see the following: