Amazon SageMaker Pipeline introduces a automatic hyperparameter tuning step
Amazon SageMaker Pipelines, the first purpose-built continuous integration and continuous delivery (CI/CD) service for machine learning (ML), is now integrated with SageMaker's automatic model tuning capability. Customers can add a model tuning step (TuningStep) in their SageMaker Pipelines which will automatically invoke a hyperparameter tuning job. The hyperparameter tuning finds the best version of a model by running many training jobs on the dataset using the algorithm and the ranges of hyperparameters specified by the customer. They can then register the best version of the model into the model registry using the RegisterModel step.
Amazon SageMaker Pipelines, the first purpose-built continuous integration and continuous delivery (CI/CD) service for machine learning (ML), is now integrated with SageMaker's automatic model tuning capability. Customers can add a model tuning step (TuningStep) in their SageMaker Pipelines which will automatically invoke a hyperparameter tuning job. The hyperparameter tuning finds the best version of a model by running many training jobs on the dataset using the algorithm and the ranges of hyperparameters specified by the customer. They can then register the best version of the model into the model registry using the RegisterModel step.
The native support for the SageMaker model tuning as a step in Pipelines enables customers to incorporate automatic model tuning as part of the model building workflow without writing custom integration code. Also, information about the TuningStep such as the location of the data sources and model artifacts are automatically stored by Amazon SageMaker ML Lineage tracking, a service that creates and stores information about the steps of a ML workflow. To learn more, visit our documentation page.