How to set up automated notification to remind the OpenAI model's retirement dates 3 months in advance

Zhai, Jim 40 Reputation points
2024-10-22T19:02:24.69+00:00

I wanted to set up automated monitoring and notification functions to track and remind the retirement date of the models that I have deployed in the production environment, and send reminder email to specific people three months in advance. However, from my understanding, these retirement dates of models are stored in the model's metadata and not in the logs? I can't find retirement date from "signals" of the monitor/alerts features. It seems like I cannot simply query the retirement dates from the system or from logs, nor can I find this field in the alerts settings.

I would like to ask how I can configure these retirement dates into my alerts, and what should be the easiest way to get those information of retirement dates and setup into automated alerts? Should I do customized logs? please help me.

Azure OpenAI Service
Azure OpenAI Service
An Azure service that provides access to OpenAI’s GPT-3 models with enterprise capabilities.
3,239 questions
0 comments No comments
{count} votes

Accepted answer
  1. Sina Salam 12,011 Reputation points
    2024-10-22T19:38:06.51+00:00

    Hello Zhai, Jim,

    Welcome to the Microsoft Q&A and thank you for posting your questions here.

    I understand that you would like to set up automated notification to remind the OpenAI model's retirement dates 3 months in advance.

    What I thought might be simplest in three steps, if you are familiar with a little python, Azure API and logic app. This is an example of how you can achieve this:

    Step 1:

    Use Azure’s API to get the metadata of your models by Extract Retirement Dates. For example, you can use the Azure SDK for Python:

    from azure.ai.openai import OpenAIClient
    from azure.identity import DefaultAzureCredential
    client = OpenAIClient(credential=DefaultAzureCredential())
    models = client.list_models()
    retirement_dates = {}
    for model in models:
        retirement_dates[model.name] = model.metadata.get('retirement_date')
    

    Step 2:

    Store the extracted dates in a database or a configuration file. For simplicity, let’s assume you’re storing them in a dictionary.

    Step 3:

    Use Azure Logic Apps to create a workflow that checks the retirement dates and sends email notifications:

    • Create a Logic App: Go to the Azure portal and create a new Logic App.
    • Add a Recurrence Trigger: Set up a trigger to run the workflow daily.
    • Add an HTTP Action: Use an HTTP action to call your API or function that checks the retirement dates.
    • Add a Condition: Check if any model’s retirement date is within the next three months.
    • Send Email: If the condition is met, use the “Send an email” action to notify the relevant people.
    • Your Logic App Workflow will be similar such as:
      • Recurrence Trigger: Runs daily.
      • HTTP Action: Calls an API to get the retirement dates.
      • Condition: Checks if any retirement date is within the next three months.
      • Send Email: Sends an email notification if the condition is met.

    There are many other methods, but with the above steps, you can automate the monitoring and notification process for your model retirement dates.

    I hope this is helpful! Do not hesitate to let me know if you more detail along the way.


    Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful.

    1 person found this answer helpful.

0 additional answers

Sort by: Most helpful

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.