What is the difference between instruct and fine-tuning?

can you explain to me what changes between fine-tuning and instruct?
I noticed this difference on the chat-gpt3 pricing page.

Both 2 alternatives use the same 4 pre-trained models.

Also having noticed that there is a limit to the maximum number of tokens per month on models without fine-tuning (e.g. davinci has a monthly limit of 2,049 tokens), I wonder if this also applies to models with fine-tuning.

Thank you

Hi @gaetanocaira

The following page:

Model index for researchers



1 Like