Finetuning needs to be cheaper

[jonathan.sabbah]

You can fine tune old Davinci but not InstructGPT
There is a confirmation here from openai staff :

Error when trying to fine tune divinci-003

I am not finding exact confirmation of this in the documentation, but could have missed it…but I am trying to fine-tune divimci-003 and I received this error: Error: Invalid base model: text-davinci-003 (model must be one of ada, babbage, curie, davinci) or a fine-tuned model created by your organization: org-bAE2T0PRuZJFaj9ZTirpswxF (HTTP status code: 400) So is it correct that divimci-003 CANNOT BE the base model for fine tuning?