How do I change the Model I use in Azure AI Service

Question

I have deployed gpt-35-turbo and text-embedding-ada-002 and it works great, but now I want to use gpt-4-32k. I tried deleting the gpt-35 model and deploying the gpt-4 model and it deploys successfully but I get an error message in my app that the model doesn't exist. For reference it is a Java Spring app using the new Spring AI framework. My RAG-Pattern prototype was working very well but changing the model causes problems. Do I need to go through the application process again or is there something I'm missing. And yes I changed the name of the model in my application.properties file. Any help would be much appreciated. Cheers!

Answer 1

To change the model you are using in Azure OpenAI Service, you need to create a new deployment with the desired model. In your case, you tried deleting the existing gpt-35-turbo model and deploying the gpt-4-32k model, but encountered an error message in your app that the model doesn't exist. This is because the deployment name is used in your code to call the model by using the client libraries and the REST APIs. Therefore, you need to update the deployment name in your code to match the new deployment name for the gpt-4-32k model.

You do not need to go through the application process again to change the model you are using. Simply follow the steps to create a new deployment with the desired model and update the deployment name in your code to match the new deployment name.

References:

• Create and deploy an Azure OpenAI Service resource
• Edit Spring application