Question

You have authored an Azure Machine Learning training pipeline job as a YAML pipeline. The business requires the pipeline to run automatically every Monday at 06:00 UTC so that a fresh model is retrained on the latest weekly data. You do not want to invoke the pipeline manually or from an external orchestrator. What is the most appropriate way to meet this requirement in Azure Machine Learning?

Accepted Answer

Create a schedule with a recurrence trigger that references the pipeline job, specifying a weekly frequency on Monday at 06:00 UTC

Answer

Configure a data drift monitor on the training dataset and let it launch the pipeline when drift is detected

Answer

Add a cron expression to the compute cluster's autoscale settings so the cluster starts the job weekly

Answer

Enable MLflow autologging on the pipeline so runs are automatically re-created on a weekly basis