Task 02: Create AI model deployments

Introduction

To enable AI-driven customer service and content moderation, Adatum requires robust AI models that can process language effectively. By deploying models like gpt-4o-mini for conversation handling and text-embedding-ada-002 for data representation, Adatum can power its AI chatbot and optimize search results based on user queries.

Description

In this task, you’ll deploy the gpt-4o-mini and text-embedding-ada-002 models to your project. These models are necessary for creating an AI-assisted chat application.

Success Criteria

The gpt-4o-mini model has been deployed.
The text-embedding-ada-002 model has been deployed.

Learning Resources

Deploy Models

Key tasks

01: Deploy gpt-4o-mini and text-embedding-ada-002 models

Expand this section to view the solution

From the Overview tab on the project1 page, select Models + endpoints from the left menu.
From the Manage deployments of your models and services page, select + Deploy model and then select Deploy base model.
From the Select a model page, select gpt-4o-mini, then select Confirm.
On the Deploy model gpt-4o-mini page, set the Deployment type to Standard.
Move the Tokens per Minute Rate Limit slider to around 450K. Leave all other settings as default and select Deploy.
Once deployed, select the back button to return to the Manage deployments of your models and services page and deploy another base model.
For this one, select text-embedding-ada-002, then select Confirm.
Ensure the Deployment type is set to Standard, then select Deploy.

These models will be used later to create a chatbot for querying product data.

You’ve successfully completed this task.