runtime Package

Contains functionality for running automated ML in pipelines, working with model explainers, and creating ensembles.

Included in this package are classes for configuring and managing pipelines, and examining run output for automated machine learning experiments. For more information about automated machine learning in Azure, see the article What is automated machine learning?

To define a reusable machine learning workflow for automated machine learning, use AutoMLStep to create a Pipeline.

Modules

automl_step	DEPRECATED. Use the functionality in the automl_step module.
ensemble	Contains functionality for creating ensembles from previous automated machine learning iterations. Creating ensembles can improve machine learning results by combining multiple iterations that may provide better predictions compared to a single iteration. Configure an experiment to use ensembles with the AutoMLConfig object.
run	Contains functionality for managing automated ML runs in Azure Machine Learning. This module allows you to start or stop automated ML runs, monitor run status, and retrieve model output.

Classes

AutoMLStep	DEPRECATED. Use the AutoMLStep class. DEPRECATED.
AutoMLStepRun	DEPRECATED. Use the AutoMLStepRun class. DEPRECATED.
HTSInferenceParameters	Parameters for HTS inference pipeline.
HTSTrainParameters	Parameters for HTS train pipeline.
ManyModelsInferenceParameters	Parameters used for ManyModels inference pipeline.
ManyModelsTrainParameters	Parameters used for ManyModels train pipeline.