Hello, currently we are limited at 3 concurrent fine tuning jobs, and they take about an hour to complete.
Sometimes we need to just queue more jobs because of client signup spikes.
Is it possible to get this limit lifted?
P.s We are at the highest usage tier already
Thank you!
That is an unusual case; I can see it is not expected that customers would invest large money in a third-party developer that then controls access to AI models they paid for and developed, to the tune of 48 per day.
On your limits page, you’ll have to pick another category of limit in “request an exception”, and hope a human reads the request justification text instead of it automatically being granted or denied.
If you need to just queue more jobs with some smart monitoring, and give a result tomorrow - you just need to queue jobs…you give an impression that more work was being done with the spending than if the model was available within seconds. 
Thank you for a detailed response, will give it a try to request exception.
Also did not knew you could queue them, I was just using the /v1/fine_tuning/jobs endpoint to create a new job : D
You’d have to hold them back yourself if getting an API error because too many are in progress.