Fine-tuning API 'gpt-4-0613' not existed

[fychen]

Hi, I already have GPT-4 access and have tried latest ‘gpt-4-1106-preview’, which is no problem. But when I tried Fine-tuning API with model ‘gpt-4-0613’, it showed ‘InvalidRequestError: Model gpt-4-0613 is not available for fine-tuning or does not exist.’
Does anyone has the same problem and know how to fix it?

[jochenschultz]

Login to platform.openai.com and select Fine-tuning.

Create a new and select a model. If there is no GPT-4 model shown it means you have not been selected for the alpha.

I don’t have it either yet. I would expect a couple of weeks of waiting time.

[daehyeonkong]

how many days did you wait for GPT-4 finetuning access?

[HouseBunting]

It’s been almost ten months now. The eligibility to train on GPT-4 is based on what?

[turbolucius]

I don’t have access either but according to several users on this forum who do, you get the option to apply after doing a LOT of fine-tuning with the current models (3.5 or older). You then have to fill out a form explaining why your use-case didn’t work with the current models and why you need GPT-4 instead. Really selective stuff.

How do I request fine-tuning for GPT-4o?

Here’s a link. If you don’t have a reasonable presentation of why your application failed with fine-tuning gpt-3.5-turbo, and why you will provide even more resources to gpt-4 and expect success that could not be accomplished otherwise (and are a company willing to fund development of thousands of examples), you’ll likely never hear back.