[error]: GPTAssistantAgent doesn't work with gpt model from Microsoft Azure

[YuanZongjian]

I am trying to apply the gpt-4-1106 model from Microsoft Azure to run the GPTAssistantAgent. Then I got error:

    _Error code: 404 - {'error': {'code': '404', 'message': 'Resource not found'}}
    httpx.HTTPStatusError: Client error '404 Resource Not Found' for url 'https://my azure openai base url/openai/deployments/gpt-4-1106/assistants'
    For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404
    
    During handling of the above exception, another exception occurred:
    
      File "/home/yzj/LLM_project/writer_v0.0/writer.py", line 60, in <module>
        planner = GPTAssistantAgent(
    openai.NotFoundError: Error code: 404 - {'error': {'code': '404', 'message': 'Resource not found'}}_

my oai_config_list is as below:
{ "model":"gpt-4-1106", "api_type":"azure", "base_url": "****", "api_key":"****" },

Is this because the GPTAssistantAgent only run with OpenAI api?

[rickyloynd-microsoft]

@gagb fyi

[gagb]

@YuanZongjian , GPTAssistantAgent uses the OpenAssistant API which is not available on Azure yet.

[YuanZongjian]

@YuanZongjian , GPTAssistantAgent uses the OpenAssistant API which is not available on Azure yet.

Okay, I see. Thank you

[KAIYI-HSU]

May I ask if I cannot use the base_url "localhost:8000/v1" when using the FastChat API on the local end? I encountered an openai.notfounderror with the GPTAssistantAgent, too. The error code is 404.

I have another question:
Is there any difference between running a Python program with GPTAssistantAgent and UserProxyAgent?
It seems like the code interpreter function and the "run Python" function in UserProxyAgent are very similar.
Thank you for your answer.