Azure OpenAI Service provides REST API access to OpenAI's powerful language models including the GPT-3, Codex and Embeddings model series. In addition, the new GPT-4 and ChatGPT (gpt-35-turbo) model series have now reached general availability. These models can be easily adapted to your specific task including but not limited to content generation, summarization, semantic search, and natural language to code translation.
Functionality for Azure OpenAI Service is split into two nodes:
Azure OpenAI Authoring includes APIs for fine-tuning and managing deployments of OpenAI models.
Azure OpenAI Inference includes APIs for completions and search.
You can find more details and the supported operation on Azure OpenAI | qibb.
How to get started with Azure OpenAI Service integration
Before you begin, ensure that you have the necessary permissions to perform the integration. When you have your Azure OpenAI Service, you can follow the steps below to get started:
Obtain your Client ID, Client Password, and Tenant ID for authentication purposes.
Use the search bar on your Azure Dashboard to search for Azure Active Directory and click on the Service.
Your Tenant ID is displayed on the starting page of the Azure Active Directory.
In the left toolbar select App registrations.
Locate your application. Your Application (client) ID is displayed here as well. Click on your application name.
In the left toolbar select Certificates & secrets.
Click the link for your Client credentials.
Your Client Password is the Value associated with your Secret ID. Secret values can only be viewed immediately after creation. Be sure to store your secret value in a secure location for use in your project.
Locate your Host URL.
Make note of your Api-Version, which you will include as a query parameter. Microsoft provides this documentation: Azure OpenAI Service REST API reference.
Note: qibb makes the Api-Version visible when editing the node settings. See the blue boxes in the screenshot for step 7.
Open the qibb Workflow Editor and click on the Node Catalog button in the right sidebar. Install the Azure OpenAI nodes by clicking on the + button.
Drag and drop the Azure OpenAI Auth, Azure OpenAI Authoring and Azure OpenAI Inference nodes from the left sidebar to your workflow and connect them to build the workflow you need.
Enter your credentials in the Client ID, Client Password, and Tenant ID fields in the Azure OpenAI Auth node.
Enter your Api-Version parameter and any other necessary parameters into the Azure OpenAI Authoring and/or Azure OpenAI Inference nodes and start using the integration!
Importing flow examples of a node
Once you've successfully installed a node from the qibb catalog, you can easily import example flows that showcase how the nodes can be used. These example flows are directly accessible within the Flow Editor and are designed to guide users with helpful hints, including how to configure mandatory parameters and credentials.
Here's how to import examples:
In the top right corner of the Flow Editor, click on the menu icon.
Select Import to open the Import Dialog.
Click on the Examples tab.
Choose an example from the list.
Confirm your selection by clicking on the Import button.
By following these steps, you are now ready to make basic API calls using the provided examples. For further exploration of what’s possible, you can find more complex examples in the Flow Catalog.