Allow API_KEY to finetune GPT3 model

API
1

Hello, genius developers!
I have openai API_KEY but this is not allowed to fine-tune model.
I run this code in python.

import openai
OPENAI_API_KEY = “”
openai.api_key = OPENAI_API_KEY
openai.Model.list()

Here, I found the model: “allow_fine_tuning”: true.
But there’s no model with this property.
How can I finetune GPT3 models using my own openai API_KEY?

2

Hi @carollhwrd

Currently only base models (davinci , curie , babbage , and ada) can be fine-tuned. Your code is just listing every model available to you - not all support fine-tuning.

Here’s the fine-tuning guide

2 Likes
3

Yeah, but here I see, the base models are not allowed to fine-tune, also.
Could you tell me why I can’t fine-tune GPT3 base models(davinci, curie, babbage, ada)?

1 Like
4

You can fine-tune base models. LMK if you hit any errors while fine-tuning.

5

Thank you for your help.
But here’s a question.
You mean, with any openai API_KEY, we can fine-tune base models?

6

Absolutely, @carollhwrd you can fine-tune with your OpenAI API Key if you have a payment method set up for your org.

7

But you should run the code that I reported above.
You can check the state of models available for your API_KEY.

8

I have called list models endpoint earlier and I know you are referring to the model permissions object, but the docs do not mention whether or how it’s meant for user consumption.

I have the same permissions as yours, on the base models.

1 Like
9

Thank you for your help.
Would you kindly explain what the model permission object mean?