Getting started with AI Platform Pipelines

This quickstart provides a brief introduction to AI Platform Pipelines. In this guide, you install Kubeflow Pipelines with TensorFlow Extended on a new Google Kubernetes Engine cluster, then run an example pipeline.

This topic is intended for users who are new to AI Platform Pipelines.

Before you begin

Before following this guide, check that your Google Cloud project is correctly set up.

  1. Sign in to your Google Account.

    If you don't already have one, sign up for a new account.

  2. In the Cloud Console, on the project selector page, select or create a Cloud project.

    Go to the project selector page

  3. Make sure that billing is enabled for your Google Cloud project. Learn how to confirm billing is enabled for your project.

Set up your AI Platform Pipelines instance

Use the following instructions to set up AI Platform Pipelines on a new GKE cluster.

  1. Open AI Platform Pipelines in Google Cloud Console.

    Go to AI Platform Pipelines

  2. Select the Google Cloud project you want to use for this quickstart, then click Continue.

  3. In the AI Platform Pipelines toolbar, click New instance. Kubeflow Pipelines opens in Google Cloud Marketplace.

  4. Click Configure. A form where you can configure your Kubeflow Pipelines deployment opens.

  5. If your Google Cloud project has existing GKE clusters, click Create a new cluster. Otherwise, continue to the next step.

  6. Select us-central1-a as the Cluster zone where your GKE cluster should be created.

  7. Check Allow access to the following Cloud APIs to grant applications that run on your GKE cluster access to Google Cloud resources. Granting your cluster access to Google Cloud resources in this manner saves you the effort of creating a Kubernetes secret.

  8. Click Create cluster to create your GKE cluster.

  9. After your cluster has been created, supply the following information:

    • Namespace: Select default as the namespace.
    • App instance name: Enter hosted-pipelines-quickstart as the instance name.
  10. Click Deploy to deploy Kubeflow Pipelines onto your new GKE cluster.

The deployment process takes several minutes to complete. After the deployment process is finished, continue to the next section.

Run an example pipeline

Use the following instructions to run an example pipeline in your new AI Platform Pipelines instance.

  1. Open AI Platform Pipelines in Google Cloud Console.

    Go to AI Platform Pipelines

  2. Find the AI Platform Pipelines cluster named hosted-pipelines-quickstart, then click Open pipelines dashboard to open Kubeflow Pipelines. The Kubeflow Pipelines dashboard opens, displaying the Getting Started page.

  3. You can use the Getting Started page in the dashboard to learn more about the demonstration and tutorial pipelines that are provided with Kubeflow Pipelines, or learn more about how to create a pipeline.

    In the left navigation panel, click Pipelines.

  4. Kubeflow Pipelines provides several example pipelines. Click [Tutorial] Data passing in python components. A graph displaying the steps in the pipeline opens.

  5. To run the pipeline once, click Create run. A form where you can enter the run details opens.

  6. Enter Quickstart pipeline run as the Run name.

  7. Click Start. The pipelines dashboard displays a list of pipeline runs.

  8. Click the run named Quickstart pipeline run. The graph of your run is displayed. While your run is still in progress, the graph changes as each step executes.

  9. Click the pipeline steps to explore your run's inputs, outputs, logs, etc.

You have now run an example pipeline in your AI Platform Pipelines instance.

Clean up

To avoid incurring more charges to your Google Cloud account, use the following instructions to delete the AI Platform Pipelines instance and GKE cluster that you created in the preceding sections.

  1. Open AI Platform Pipelines in Google Cloud Console.

    Go to AI Platform Pipelines

  2. Select the checkbox for the AI Platform Pipelines instance named hosted-pipelines-quickstart.

  3. In the AI Platform Pipelines toolbar, click Delete.

  4. In the Delete Kubeflow Pipelines from cluster dialog, select the Delete cluster checkbox. Selecting this checkbox indicates that you want to delete the GKE cluster you created for this quickstart.

  5. Click Delete to delete your AI Platform Pipelines instance and GKE cluster.

What's next