Documentation

Specifies the training algorithm to use in a CreateTrainingJob request.

For more information about algorithms provided by Amazon SageMaker, see Algorithms. For information about using your own algorithms, see Using Your Own Algorithms with Amazon SageMaker.

See: newAlgorithmSpecification smart constructor.

Create a value of AlgorithmSpecification with all optional fields omitted.

Use generic-lens or optics to modify other optional fields.

The following record fields are available, with the corresponding lenses provided for backwards compatibility:

$sel:enableSageMakerMetricsTimeSeries:AlgorithmSpecification', algorithmSpecification_enableSageMakerMetricsTimeSeries - To generate and save time-series metrics during training, set to true. The default is false and time-series metrics aren't generated except in the following cases:

• You use one of the Amazon SageMaker built-in algorithms

• You use one of the following Prebuilt Amazon SageMaker Docker Images:

  • Tensorflow (version >= 1.15)
  • MXNet (version >= 1.6)
  • PyTorch (version >= 1.3)
• You specify at least one MetricDefinition

$sel:algorithmName:AlgorithmSpecification', algorithmSpecification_algorithmName - The name of the algorithm resource to use for the training job. This must be an algorithm resource that you created or subscribe to on Amazon Web Services Marketplace. If you specify a value for this parameter, you can't specify a value for TrainingImage.

$sel:trainingImage:AlgorithmSpecification', algorithmSpecification_trainingImage - The registry path of the Docker image that contains the training algorithm. For information about docker registry paths for built-in algorithms, see Algorithms Provided by Amazon SageMaker: Common Parameters. Amazon SageMaker supports both registry/repository[:tag] and registry/repository[@digest] image path formats. For more information, see Using Your Own Algorithms with Amazon SageMaker.

$sel:metricDefinitions:AlgorithmSpecification', algorithmSpecification_metricDefinitions - A list of metric definition objects. Each object specifies the metric name and regular expressions used to parse algorithm logs. Amazon SageMaker publishes each metric to Amazon CloudWatch.

$sel:trainingInputMode:AlgorithmSpecification', algorithmSpecification_trainingInputMode - Undocumented member.

To generate and save time-series metrics during training, set to true. The default is false and time-series metrics aren't generated except in the following cases:

• You use one of the Amazon SageMaker built-in algorithms

• You use one of the following Prebuilt Amazon SageMaker Docker Images:

  • Tensorflow (version >= 1.15)
  • MXNet (version >= 1.6)
  • PyTorch (version >= 1.3)
• You specify at least one MetricDefinition

The name of the algorithm resource to use for the training job. This must be an algorithm resource that you created or subscribe to on Amazon Web Services Marketplace. If you specify a value for this parameter, you can't specify a value for TrainingImage.

The registry path of the Docker image that contains the training algorithm. For information about docker registry paths for built-in algorithms, see Algorithms Provided by Amazon SageMaker: Common Parameters. Amazon SageMaker supports both registry/repository[:tag] and registry/repository[@digest] image path formats. For more information, see Using Your Own Algorithms with Amazon SageMaker.

A list of metric definition objects. Each object specifies the metric name and regular expressions used to parse algorithm logs. Amazon SageMaker publishes each metric to Amazon CloudWatch.

Undocumented member.