Azure OpenAI Service integrates powerful generative AI models into the Azure platform, offering robust security, scalability, and seamless integration with other Azure services.
This guide provides an overview of using Azure OpenAI Studio to create AI solutions.
Accessing Azure OpenAI To start, you need Azure subscription that has been approved for access to the Azure OpenAI service. Note that Azure OpenAI Service is in limited access, and you must apply for it at this link.
Sign into Azure portal and Create a resource. When creating the Azure OpenAI resource, you'll provide details like; subscription name, resource group, region, instance name, and pricing tier.
Create an Azure OpenAI resource with the following settings:
Subscription: Select an Azure subscription that has been approved for access to the Azure OpenAI service
Resource group: Choose or create a resource group Region: Make a random choice from any of the following regions* Australia East, Canada East, East US, East US 2, France Central, Japan East, North Central US, Sweden Central, Switzerland North, UK South.
Name: A unique name of your choice
Pricing tier: Standard S0
- Azure OpenAI resources are constrained by regional quotas. The listed regions include default quota for the model type(s) used in this exercise. Randomly choosing a region reduces the risk of a single region reaching its quota limit in scenarios where you are sharing a subscription with other users. In the event of a quota limit being reached later in the exercise, there's a possibility you may need to create another resource in a different region.
- Wait for deployment to complete. Then go to the deployed Azure OpenAI resource in the Azure portal.
Using Azure OpenAI Studio After setting up your resource, access Azure OpenAI Studio via the Azure portal. This studio offers tools for model management, deployment, experimentation, customization, and learning.
Model Deployment In Azure OpenAI Studio, initiate your first model deployment through the call-to-action button on the main screen. This opens the Deployments page, where you can deploy and experiment with a base model.
Types of Models Azure OpenAI Service offers several model types:
GPT-4 Models: Latest generation, capable of generating natural language and code completions.
GPT-3.5 Models: Similar to GPT-4 but optimized for chat interactions, especially the GPT-35-turbo variant.
Embeddings Models: Convert text into numeric vectors for language analytics.
DALL-E Models: Generate images from natural language prompts (currently in preview and not listed in the studio interface).
- Deploy a Model Deploy a model to make API calls for prompt completions. You can create multiple deployments within one or several Azure OpenAI resources, adhering to the regional TPM limit of 240K.
Experimenting with Prompts Once deployed, test your model's responses (completions) to prompts. A prompt is a text request sent to the model, and completions are the model's text, code, or other formatted responses.
Playgrounds Azure OpenAI Studio's Playgrounds are interactive interfaces for experimenting with deployed models without building a client application. They offer various parameter tuning options to refine your AI solutions.
In Azure OpenAI Studio, in the Playground section, select the Chat page. The Chat playground page consists of three main sections:
Setup - used to set the context for the model's responses. Chat session - used to submit chat messages and view responses. Configuration - used to configure settings for the model deployment.
In the Chat session, submit a query e.g. what is azure
By leveraging Azure OpenAI Service, you can build sophisticated AI applications that benefit from Azure's robust infrastructure and integration capabilities.