Vertex AI Studio

Appearance

SOP-001: Initial Setup ​

DOCUMENT CONTROL

FieldValue
SOP IDSOP-001
Version1.0
StatusActive

INFO

Document Control Title: Initial Setup for Vertex AI Studio Version: 1.0 Last Updated: 2023-04-17

Purpose ​

This Standard Operating Procedure (SOP) document provides the steps to set up Vertex AI Studio for the first time. It covers the necessary configurations and settings to get Vertex AI Studio up and running in your environment.

Procedure Flow ​

Procedure ​

Step 1: Sign in to Google Cloud Console ​

WARNING

Make sure you have the necessary permissions to create a new Google Cloud project and enable the Vertex AI API.

  1. Go to the Google Cloud Console and sign in with your Google account.

Step 2: Create a new Google Cloud Project ​

  1. In the Google Cloud Console, click on the project selector in the top-left corner.
  2. Click on "New Project" and provide a name for your project.
  3. Select the appropriate organization (if applicable) and location, then click "Create".
  4. Wait for the project to be created, then select it from the project selector.

Step 3: Enable the Vertex AI API ​

  1. In the Google Cloud Console, navigate to the "APIs & Services" section.
  2. Click on "Library" in the left-hand menu.
  3. Search for "Vertex AI" and click on the "Vertex AI" service.
  4. Click on the "Enable" button to enable the Vertex AI API for your project.

Step 4: Set up Vertex AI Studio ​

  1. In the Google Cloud Console, navigate to the "Vertex AI" section.
  2. Click on "Vertex AI Studio" in the left-hand menu.
  3. Click on the "Open Vertex AI Studio" button to launch the Vertex AI Studio interface.
  4. Follow the on-screen instructions to complete the setup process, which may include:
    • Configuring your compute resources (e.g., selecting a runtime version, instance type, and disk size)
    • Connecting your data sources (e.g., Google Cloud Storage, BigQuery, or other integrations)
    • Customizing your Vertex AI Studio workspace (e.g., creating notebooks, importing datasets, or setting up pipelines)

Step 5: Verify the Vertex AI Studio setup ​

DANGER

Ensure you have completed all the necessary steps before proceeding with further tasks in Vertex AI Studio.

[ ] Verify that you can access the Vertex AI Studio interface from the Google Cloud Console. [ ] Confirm that you can create a new notebook and run a simple test code snippet. [ ] Check that you can access your data sources (e.g., Google Cloud Storage, BigQuery) from within Vertex AI Studio. [ ] Ensure that you can create and run a basic machine learning pipeline in Vertex AI Studio.

Troubleshooting ​

IssuePossible CauseResolution
Unable to access Vertex AI StudioIncorrect permissions or project selection1. Verify that you have the necessary permissions to access Vertex AI Studio.
2. Ensure that you have selected the correct Google Cloud project in the console.
Error when creating a notebookIncorrect runtime configuration1. Check the available runtime versions and instance types in your Vertex AI Studio.
2. Ensure that you have selected the appropriate runtime for your use case.
Data source integration issuesIncorrect data source credentials or permissions1. Verify the connection details and permissions for your data sources.
2. Check that you have the necessary IAM roles and access to the data sources.
Pipeline execution failuresIncorrect pipeline configuration or dependencies1. Review the pipeline definition and ensure that all components are properly configured.
2. Check the pipeline logs for any errors or missing dependencies.

See Also ​