Use a prebuilt template from the Template Gallery

Preview

This feature is covered by the Pre-GA Offerings Terms of the Google Cloud Terms of Service. Pre-GA features might have limited support, and changes to pre-GA features might not be compatible with other pre-GA versions. For more information, see the launch stage descriptions.

The Vertex AI Pipelines Template Gallery contains Google-authored pipeline templates that you can use to create pipeline runs.

Note: The Template Gallery contains templates that are generally available (GA) as well as templates that are in preview. To understand the terms of service of each template, refer to its associated documentation. For more information, see the launch stage descriptions.

You can create a pipeline run using the following template from the Template Gallery:

• AutoML for Tabular Classification / Regression: A template for the Tabular Workflow for End-to-End AutoML pipeline. The Tabular Workflow for End-to-End AutoML is the complete AutoML pipeline for training a classification or regression model from a tabular dataset. Click Create run to create a pipeline run using this template. For more information about creating a pipeline run using this template, see Train a Model.

  For more information about the Tabular Workflow for End-to-End AutoML pipeline, see Tabular Workflow for End-to-End AutoML.

The following pipeline templates will be added soon:

• Wide & Deep training: For more information about the Wide & Deep training pipeline, see Tabular Workflow for Wide & Deep. To learn how to train a model using this pipeline, including links to Jupyter Notebook Tutorials, see Train a Model.

• TabNet training: For more information about the TabNet training pipeline, see Tabular Workflow for TabNet. To learn how to train a model using this pipeline, including links to Jupyter Notebook Tutorials, see Train a Model.

You can also create a pipeline run based on a custom pipeline template. For more information about how to create and use a pipeline template, see Create, upload, and use a pipeline template.