SOP-001: Initial Setup

DOCUMENT CONTROL

Field	Value
SOP ID	SOP-001
Version	1.0
Status	Active

Here is a detailed SOP document for "Initial Setup: Steps to set up Vertex AI on your system":

INFO

SOP Document: Initial Setup - Vertex AI

Document Version: 1.0 Last Updated: 2023-04-12 Author: John Doe Owner: Acme Inc. Cloud Team

Purpose

This SOP document outlines the steps required to set up Vertex AI on your system. Vertex AI is a fully-managed machine learning platform that enables you to build, deploy, and scale ML models quickly. This document provides a step-by-step guide to get you started with Vertex AI.

Procedure Flowchart

┌───────────────────────────┐
│ Start Initial Vertex AI  │
│        Setup             │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│  Create Google Cloud     │
│     Platform Project     │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│ Enable Vertex AI API and │
│    related services      │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│  Set up authentication   │
│      credentials         │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│   Configure Vertex AI    │
│       environment        │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│     Verify Vertex AI     │
│      installation        │
└───────────────────────────┘
     ▼
┌───────────────────────────┐
│        End Initial       │
│       Vertex AI Setup    │
└───────────────────────────┘

Step-by-step Procedure

1. Create Google Cloud Platform Project

INFO

Step 1: Create a new Google Cloud Platform project or use an existing one.

You can create a new project by following these steps:

1. Go to the Google Cloud Console.
2. Click on the project dropdown at the top of the page and select "New Project".
3. Provide a project name and optionally, a project ID. Then, click "Create".

2. Enable Vertex AI API and related services

INFO

Step 2: Enable the Vertex AI API and any other related services required for your use case.

1. In the Google Cloud Console, navigate to the "APIs & Services" section.
2. Click on the "Library" tab and search for "Vertex AI".
3. Click on the "Vertex AI API" and then click the "Enable" button to enable the API.
4. Repeat the process for any other related services you need, such as Cloud Storage, Cloud Dataflow, etc.

3. Set up authentication credentials

INFO

Step 3: Set up the necessary authentication credentials to access Vertex AI.

1. In the Google Cloud Console, navigate to the "IAM & Admin" section.
2. Click on "Service Accounts" and create a new service account or use an existing one.
3. Grant the necessary roles and permissions to the service account, such as "Vertex AI Administrator" or "Vertex AI User".
4. Download the JSON key file for the service account.

4. Configure Vertex AI environment

INFO

Step 4: Configure your local development environment to work with Vertex AI.

1. Install the Google Cloud SDK (if not already installed) from the official website.
2. Initialize the Google Cloud SDK by running gcloud init and follow the prompts to authenticate with your Google account.
3. Set the default project for the Google Cloud SDK by running gcloud config set project [PROJECT_ID], replacing [PROJECT_ID] with your project ID.
4. (Optional) Set up a virtual environment for your Python project and install the necessary Vertex AI SDK dependencies by running pip install google-cloud-aiplatform.

5. Verify Vertex AI installation

INFO

Step 5: Verify that the Vertex AI setup is complete and working correctly.

1. Create a simple Vertex AI model training pipeline using the following code:

   from google.cloud import aiplatform
   
   # Set the project, location, and other settings
   aiplatform.init(project="[PROJECT_ID]", location="us-central1")
   
   # Define the training pipeline
   pipeline = aiplatform.pipeline.TabularDatasetTrainingPipeline(
       display_name="my-pipeline",
       # Add your pipeline details here
   )
   
   # Run the pipeline
   pipeline.run()

2. Verify that the pipeline run was successful in the Vertex AI console.

Verification Checklist

• [ ] Google Cloud Platform project created or selected
• [ ] Vertex AI API and related services enabled
• [ ] Authentication credentials set up
• [ ] Local development environment configured
• [ ] Vertex AI model training pipeline successfully created and run

Troubleshooting

Issue	Possible Cause	Solution
Unable to enable Vertex AI API	API not available in the selected region	Check the Vertex AI availability and select a supported region.
Authentication credentials not working	Incorrect service account permissions	Ensure the service account has the necessary roles and permissions to access Vertex AI.
Pipeline run fails	Incorrect pipeline configuration	Review the pipeline code and ensure all required parameters are set correctly.
Unable to find Vertex AI SDK	Missing Python dependencies	Verify that the Google Cloud SDK and Vertex AI SDK are installed correctly in your Python environment.

DANGER

If you encounter any issues that you cannot resolve, please reach out to the Acme Inc. Cloud Team for assistance.

See Also

• SOP-002: Configuration
• SOP-003: Model Creation
• SOP-004: Model Deployment