I am fine tuning with base model gpt-4.1-mini-2025-04-14. however the fine tuned model is failing on safety evals done by open ai. I also tried fine tuning using an earlier dataset ( seed and all params same) on which the fine tuned model had passed safety evals. However , even on that dataset the model now is failing safety evals. Has something changed at open ai end ? My dataset contains user query for analyzing data as input and steps / plan as output.
1 Like
Hi, and thanks for flagging this issue!
OpenAI has rolled out an update to prevent this from happening again.
You may need to wait a day or two before retrying a previously rejected job, but new jobs should work fine.
If the issue persists, please let us know in this topic: