Hi guys, I’m trying to fine tune for a project. I used the UI to create the job and uploaded my jsonl file but got the error “We’re having trouble accessing your files right now. Please try again later.” the whole time.
The input file is pretty straighforward and I have tried to use a smaller file of just 15 sample and still failed. Does anyone know this error and how to solve it?
I also experienced same issue. No matter what method I used for solving, I can’t go forward with the same issue.
not working for me as well after 4 hours. might be a tomorrow thing.
Same here. Any solutions?
Same here. I’ve been trying it for the last couple of days, and no matter what I do, it throws the same error: “We’re having trouble accessing your files right now. Please try again later.”
Hi guys, thanks for the feedback!
I think it has to do with 4o-mini. I tried to finetune with gpt-4.1-mini-2025-04-14 as base and it worked!
Just got mine working again.
The issue was the model gpt-4o-mini-2024-07-18 fine-tuning.
Switching to gpt-4.1-mini-2025-04-14 got it working.
I encountered the same issue. Is it still possible to use 4o normally? Has the official team shut down the use of this model?
Same here - we had much better results with 4o mini and we used it for long time. Now we want to check, if fine tuning will improve results.
I have the same problem . Do you all have the same problem with different models, including gpt-3.5-turbo-0125, gpt-3.5-turbo-1106, gpt-4o-mini-2024-07-18 .
gpt-4o-mini-2024-07-18 as a base model appears to be working now.
I’ve been consistently getting this error for the past couple of days, including 5 minutes ago, so it has NOT been solved.
I’m fine-tuning with model gpt-4o-mini-2024-07-18 using a process that has succeeded hundreds of times.
I’d appreciate an acknowledgement from OpenAI (https://status.openai.com/ says fully operational) and an ETA to a fix. We have a dependency on fine-tuning here that is increasingly seen by my management as a liability. We were unable to fine-tune for a number of days last month too ( Fine-tuning jobs failing with "internal error" ).
Thanks very much for discovering that the problem is limited to 4o-mini.
OpenAI – please don’t consider this to be a solution. It is at best a workaround, since it is not necessarily possible to shift to a different model.
Did anyone have any updates yet? I am still waiting for GPT 3.5 Turbo to actually work on its.
Hi everyone,
This should now be resolved. The team investigated and identified the underlying issue, and a fix has been deployed.
Appreciate everyone flagging this. Feel free to share if anything similar comes up again.
I am also facing the same issue where the fine-tuning is failing for the gpt-4o-mini-2024-07-18 model. Please fix this.
did they fixed it or no ?
This topic was automatically closed 5 days after the last reply. New replies are no longer allowed.
