HI! I'm currently seeking some guidance regarding an issue I've encountered with Azure. I've successfully created a deployment with the Deployment Name "123", using the GPT-4 model and the 1106-Preview version. However, I've run into a roadblock while trying to create an assistant. Every attempt results in the following error: "Error code: 404 - {'error': {'code': '404', 'message': 'Resource not found'}}".

I'm wondering if this issue might be related to incorrect entries in the api_version and model parameters. Could anyone kindly advise on where I can find detailed information about these parameters, or suggest potential solutions to this issue? Your expertise and insights would be greatly appreciated.

Thank you in advance for your time and assistance!

Hello Li, Yuming,

According to the Microsoft Learn document, the Assistants API is currently available in Australia East, East US 2, and Sweden Central. Could you please provide the region where you are currently deployed?

Hello Charlie, thank you for your response. Currently I'm deploying in the Sweden Central area.

Hello Li, Yuming,

Could you please inform me whether you were created using an API or OpenAI Studio? If it was through an API, could you provide the method of your invocation?

I believe I created it using an API approach. I followed the example provided in the official documentation and executed it in VS Code. Here is an outline of the code I used:

python

Copy code

import io

import os

import time

from datetime import datetime

from typing import Iterable

Importing necessary libraries and modules

from dotenv import load_dotenv

from openai import AzureOpenAI

from openai.types.beta.threads.message_content_image_file import MessageContentImageFile

from openai.types.beta.threads.message_content_text import MessageContentText

from openai.types.beta.threads.messages import MessageFile

from PIL import Image

Setting up API credentials and endpoint

api_endpoint = 'https://lotXXXXXXXXXXXXXXXXX2.openai.azure.com/'  # Partially redacted for security

api_key = '433XXXXXXXXXXXXXXXXXXXXXXXXXX5a'  # Partially redacted for security

api_version = '2023-11-15'  # I'm uncertain about this value; I've tried various options like '1106-Preview'

Initializing the AzureOpenAI client

client = AzureOpenAI(api_key=api_key, api_version=api_version, azure_endpoint=api_endpoint)

Creating an assistant with specific instructions and tools

assistant = client.beta.assistants.create(

    name="Data Visualization",

    instructions=(

        "You are a helpful AI assistant who makes interesting visualizations based on data. "

        "You have access to a sandboxed environment for writing and testing code. "

        "When asked to create a visualization, you should follow these steps: "

        "1. Write the code. "

        "2. Display a preview of any new code. "

        "3. Run the code to confirm it works. "

        "4. If successful, display the visualization. "

        "5. If unsuccessful, show the error message, revise the code, and repeat the steps."

    ),

    tools=[{"type": "code_interpreter"}],

    model="1106-Preview"  # I'm also unsure about this; I've tried various options like 'GPT-4'

)

Regarding the api_version and model parameters, I'm not entirely sure what values to input. I've experimented with different content, such as '1106-Preview', and various other methods for the model, including 'GPT-4', but I'm still facing challenges. Any guidance or suggestions on how to correctly configure these parameters would be greatly appreciated.

Hello Li, Yuming,

You are currently using the Python SDK, and I recommend checking out this Quickstart. For the api_version, it is advisable to use the latest version, 2024-02-15-preview. The model parameter is the name of your model deployment, which I believe is 123.

Please try the following modified program.

import os
from openai import AzureOpenAI

client = AzureOpenAI(
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),  
    api_version="2024-02-15-preview",
    azure_endpoint = os.getenv("AZURE_OPENAI_ENDPOINT")
)

assistant = client.beta.assistants.create(
    name="Data Visualization",
    instructions="...",
    tools=[{"type": "code_interpreter"}],
    model="123" #You must replace this value with the deployment name for your model.
)

Best regards,
Charlie

---

If you find my response helpful, please consider accepting this answer and voting 'yes' to support the community. Thank you!

Thank you so much, your solution worked！

Thank you Charlie, i can confirm it was an issue with the api version in use, worked after changing it!