Do we have any example of Finetuned GPT3

At a glance

The community member is trying to use a finetuned GPT-3 model to generate a response, but is encountering an error stating that the model name is unknown. The community member is using version 0.6.15 of the llamaIndex library and is wondering if finetuned models can be used in newer versions of the library.

In the comments, another community member provides a solution to the issue by handling finetuned models by splitting the model name on the colon character.

WWhiteFang_Jr

Do we have any example of Finetuned GPT3 model being used to generate response?
As I'm trying with an finetuned model but getting this error

Plain Text

Unknown model: davinci:ft-finetuned-2023-07-08-13-19-29. Please provide a valid OpenAI model name.Known models are: gpt-4, gpt-4-0314, gpt-4-32k, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-0301, text-ada-001, ada, text-babbage-001, babbage, text-curie-001, curie, davinci, text-davinci-003, text-davinci-002, code-davinci-002, code-davinci-001, code-cushman-002, code-cushman-001

This occurs in modelname_to_contextsize from langchain OpenAI class. I'm using 0.6.15 for llamaIndex. I know the version is old. lol 😅
Are we allowing using finetuned model to be used in newer versions of llamaIndex?

1 comment

WWhiteFang_Jr

Okay, Saw that it is now being handled.

Plain Text

    # handling finetuned models
    if "ft-" in modelname:
        modelname = modelname.split(":")[0]

Ignore my question 🙌

Add a reply

Find answers from the community

Do we have any example of Finetuned GPT3