"The specified base model does not support fine-tuning" error when trying to fine tune gpt-35-turbo using rest API

Sunanda Varma 20

Hi,

We are trying to fine tune the "gpt-35-turbo" model programmatically using rest API with the resource in Sweden Central region. API url: https://xxxx.openai.azure.com/openai/fine-tunes?api-version=2023-07-01-preview Model details: { model = "gpt-35-turbo", training_file = "file-fileId" };

response = httpClient.PostAsync(url, httpContent).GetAwaiter().GetResult();

Encountered the below error.

"error": { "code": "invalidPayload", "message": "The specified base model does not support fine-tuning." }

when we try to create a fine tune job with model “gpt-35-turbo-0613”, got the below error

“The given base model is not supported on fine-tunes api route”.

We tried with gpt-35-turbo-0613 and gpt-35-turbo-1106 and gpt-35-turbo models and encountered the same issue.

please note, we are able to create a custom fine tune model and also a deployment using this fine-tuned model from portal directly(not through rest API).

Please advise on how to fine tune "gpt-35-turbo" model using rest API

Thanks,

Sunanda

Vinodh247-1375 11,396 Reputation points

2024-04-04T07:10:21.2333333+00:00

I see you have already raised the same question a few weeks ago... were you able to implement the suggestions provided by the experts as answer?

https://learn.microsoft.com/en-us/answers/questions/1599638/the-specified-base-model-does-not-support-fine-tun
Sunanda Varma 20 Reputation points

2024-04-04T08:34:18.84+00:00

Yes, the role is granted by default when the OpenAI instance is created. Please check below. Also, when calling REST API doesn't it use API key? why is the need for user to have the role assigned?
Vinodh247-1375 11,396 Reputation points

2024-04-04T10:37:07.64+00:00

though the API key handles authenticatoin and authorization at a basic level, assigning roles helps to manage access and permissions more effectively. This is important in environments where multiple users or applications might be accessing the api & you want to control what particular user or application can perform.
Sunanda Varma 20 Reputation points

2024-04-04T10:48:45.1733333+00:00

As the Cognitive Services OpenAI Contributor role is already assigned, why is the fine tuning not working using rest API? is there anything that we are missing.
AshokPeddakotla-MSFT 28,226 Reputation points

2024-05-01T02:07:02.0833333+00:00

Hello Sunanda Varma,

If you are still blocked, For a deeper investigation and immediate assistance on this issue, please file a support request @ https://aka.ms/azsupt?

Share via

"The specified base model does not support fine-tuning" error when trying to fine tune gpt-35-turbo using rest API