NotFoundError: Error code: 404 - {'error': {'code': 'DeploymentNotFound', 'message': 'The API deployment for this resource does not exist. If you created the deployment within the last 5 minutes, please wait a moment and try again.'}}

[pranavvr-lumiq]

I am getting the following error:

NotFoundError: Error code: 404 - {'error': {'code': 'DeploymentNotFound', 'message': 'The API deployment for this resource does not exist. If you created the deployment within the last 5 minutes, please wait a moment and try again.'}}

trying to recreate this notebook locally:
https://github.com/microsoft/autogen/blob/main/notebook/agentchat_auto_feedback_from_code_execution.ipynb

[julianakiseleva]

@pranavvr-lumiq it's looks to me as a problem with your endpoint for OpenAI api.

you can see more here: https://learn.microsoft.com/en-us/answers/questions/1308161/the-api-deployment-for-this-resource-does-not-exis

Can you please check if you configuration file is setup properly?
Have you just started and it's your first try? Or have you used autogen successfully, and it's start happening suddenly?

[pranavvr-lumiq]

@pranavvr-lumiq it's looks to me as a problem with your endpoint for OpenAI api.

you can see more here: https://learn.microsoft.com/en-us/answers/questions/1308161/the-api-deployment-for-this-resource-does-not-exis

Can you please check if you configuration file is setup properly? Have you just started and it's your first try? Or have you used autogen successfully, and it's start happening suddenly?

Thanks for the quick reply @julianakiseleva. This is not my first time using autogen. i was using it succesfully with a conventional openai API key, until my company temporarily shut them down. Right now i am trying to use an azure openai key, and this issue started. I do believe the config json file is set up correctly. I followed the instructions closely.

[tengda89757]

I also encountered with the same problem. Sad :-(

[olivMertens]

Yeah same thing here

[sonichi]

Could you share an example of the config that failed?

[pranavvr-lumiq]

Could you share an example of the config that failed?

Sure. here it is:

{
"model": "gpt-4",
"base_url": "https://my_link.openai.azure.com/",
"api_type": "azure",
"api_version": "2023-07-01-preview",
"api_key": "my_apa_key"
}

[winstxnhdw]

The Azure API endpoint requires a deployment_name parameter which autogen does not expose..

EDIT: Nevermind, the docs are so terrible that apparently the model field is supposed to be your deployment_name. Nice.

[pranavvr-lumiq]

Finally figured it out. The deployment name my coworker gave me was incomplete. Filled in the full deployment name in the model section and it worked.

[gregrobison]

What is the proper format for Azure OpenAI deployment name that works in autogen?

[rickyloynd-microsoft]

Normally the "model" key should have your model deployment name as its value.

[mrmechazilla]

In spring AI, the application.properties should look like this:

spring.ai.azure.openai.api-key=
spring.ai.azure.openai.endpoint=
spring.ai.azure.openai.model=
spring.ai.azure.openai.api-version=

this will solve the problem

[sanjeevtrivedi]

In spring AI, the application.properties should look like this:

spring.ai.azure.openai.api-key= spring.ai.azure.openai.endpoint= spring.ai.azure.openai.model= spring.ai.azure.openai.api-version=

this will solve the problem

Sorry, as per documentation and with spring-ai version 0.8.0-SNAPSHOT, this does not works.