status : Fine tune failed during postprocessing

Question

When there, when I tried to use AI foundry for reinforcement learning fine tuning, i got status : Fine tune failed during postprocessing. The experiment failed after 3 retries. Could u guide us on how to see the real issue please?

Answer 1

Hi Oliver,

Make sure of the following points:

• Data is of the same format as provided in the documentation:
  https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/fine-tuning?tabs=azure-openai&pivots=programming-language-studio#example-file-format
• Data doesn't contain any Non-ASCII characters.
• If you are using Python SDK check fine-tuning job status using the following code:

response = client.fine_tuning.jobs.retrieve(job_id)
print("Job ID:", response.id)
print("Status:", response.status)
print(response.model_dump_json(indent=2))

• For creating fine-tuning job using python. Here is relevant documentation:
  https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/fine-tuning?tabs=azure-openai&pivots=programming-language-python#create-a-customized-model
• Make sure your data contains more than 10 data points.

Let me know if the issue persists.
Thankyou for reaching out to The Microsoft QNA Portal.