Azure OpenAI Service Deployments

Overview

Azure OpenAI Service allows you to leverage powerful language models like GPT-4, GPT-3.5 Turbo, and DALL-E for a wide range of natural language processing tasks. Deploying a model is the first step to integrating these capabilities into your applications.

Prerequisites

• An Azure Subscription.
• Access to Azure OpenAI Service.
• An Azure OpenAI Service resource created in your subscription.

If you haven't already, please visit the Azure OpenAI Service access request page to apply for access.

Steps to Deploy a Model

1. Navigate to your Azure OpenAI Resource: Go to the Azure portal and find your Azure OpenAI Service resource.
2. Open the Azure OpenAI Studio: Within your resource, click on the "Go to Azure OpenAI Studio" button.
3. Select "Deployments": In the Studio, navigate to the "Deployments" section in the left-hand menu.
4. Create a New Deployment: Click the "+ Create new deployment" button.
5. Choose a Model: Select the desired model from the dropdown list (e.g., `gpt-35-turbo`, `text-davinci-003`).
6. Configure Deployment Settings:
   • Deployment name: A unique name for your deployment (e.g., `my-chat-deployment`).
   • Content filter: Configure content filtering settings as needed.
   • Advanced options: Set capacity (tokens per minute) based on your needs and the model's limits.
7. Create: Click the "Create" button to deploy the model.

Example Deployment (CLI)

You can also deploy models using the Azure CLI:

Replace the placeholders like <deployment-name>, <resource-group-name>, <your-azure-openai-resource-name>, <model-name>, and <tokens-per-minute> with your specific values.

Next Steps

Once your model is deployed, you can start interacting with it:

• Explore the playground in Azure OpenAI Studio.
• Integrate your deployment into applications using the REST API or SDKs.
• Learn about prompt engineering to get the best results.