Get error 500 when sending requet to cohere-rerank v4.0-fast

Question

We deployed the model cohere-rerank v4.0-fast on Azure Foundry in France Central.

But any request returns an error 500.

What did we miss to make it work?

Here is a sample python progam using the cohere client

import cohere

co = cohere.ClientV2(
    api_key="XXXXX",
    base_url="https://XXXXX.services.ai.azure.com/providers/cohere/v2/rerank",
)

documents = [
    {
        "Title": "Incorrect Password",
        "Content": "Hello, I have been trying to access my account for the past hour and it keeps saying my password is incorrect. Can you please help me?",
    },
    {
        "Title": "Confirmation Email Missed",
        "Content": "Hi, I recently purchased a product from your website but I never received a confirmation email. Can you please look into this for me?",
    },
]


query = "What emails have been about refunds?"

results = co.rerank(
    model="model",
    documents=documents,
    query=query,
    top_n=3,
)

Answer 1

The error 500 you are encountering when sending requests to the cohere-rerank v4.0-fast model may be due to several factors. Here are some troubleshooting steps you can take:

1. Check the Model Name: Ensure that the model name you are using in the request (model="model") is correct. It should match the deployment name of the model you have set up in Azure.
2. Verify API Key and Endpoint: Confirm that the API key and the base URL you are using are correct and that they correspond to the Azure deployment of the model.
3. Inspect Document Format: Make sure that the documents you are sending in the request are formatted correctly and meet any requirements specified by the API.
4. Reduce Query Complexity: Sometimes, simplifying the query can help in identifying if the issue is related to the complexity of the input.
5. Check for Service Availability: Ensure that the Azure service is operational and not experiencing any outages or issues that could affect the model's performance.
6. Implement Retry Logic: Since 500 errors can be transient, implementing retry logic in your code might help. You can retry the request after a short delay.

If you continue to face issues, consider reaching out to Azure support for more specific assistance regarding your deployment.

Answer 2

Changing the name of the model fixed the error 500.
Unfortunately, it still does not work but now with an error 404