Must provide an 'engine' parameter to create a <class 'openai.api_resources.completion.Completion'>

Prompting
1

Hello everyone,

I have the following code to access my fine tuned GPT-3 Model:

    prompt = f"{chat_log}Human: {question}\nAI-Doctor:"
    response = openai.Completion.create(
        prompt = prompt,
        #engine="curie",
        model="curie:ft-personal-2022-##-##-##,
        temperature = 0.70,
        top_p=1,
        frequency_penalty=0,
        presence_penalty=0.7,
        n=1,
        best_of=1,
        max_tokens=200,
        stop = "Human:")

which has worked fine until 5 minutes ago.
Now i get the error:

Must provide an 'engine' parameter to create a <class 'openai.api_resources.completion.Completion'>

If i provide an engine i get another error, which is:

You requested a model that is not compatible with this engine. Please contact support@openai.com if you need assistance

Any idea what could have happened? I just want to access my own fine tuned model, and it worked for about a week now.

2

Solving my own problem here:

Just run

pip install --upgrade openai

or try a fresh environment

This solved it for me

2 Likes
4

The docs for using fine-tuned engine state:

Note that no engine is specified on these requests. This is the intended design and something we plan on standardizing across other API endpoints in the future.

1 Like