Fine-tuning Support for OpenAI Model

anarchy-ai / LLM-VM

irresponsible innovation. Try now at https://chat.dev/

https://anarchy.ai/

MIT License

477 stars 149 forks source link

Fine-tuning Support for OpenAI Model #269

Closed sushmanthreddy closed 12 months ago

sushmanthreddy commented 12 months ago

In the finetune method of the code, it appears that fine-tuning is not currently supported for the OpenAI model. The method raises an exception indicating this limitation. However, there's commented-out code that suggests there might have been an attempt or a plan to support fine-tuning in the future.


def finetune(self, dataset, optimizer, c_id):
    print("fine tuning isn't supported by OpenAI on this model", file=sys.stderr)
    raise Exception("fine tuning isn't supported by OpenAI on this model")
    # ... (rest of the method)

sushmanthreddy commented 12 months ago

@mmirman I would like to work on this issue..

mmirman commented 12 months ago

@VictorOdede @abhigya-sodani Why isn't there finetuning support for OAI here? I recall when I first made the optimizer endpoint there was finetuning for OAI

abhigya-sodani commented 12 months ago

Last I remember, there was finetuning for OpenAI too.

sushmanthreddy commented 12 months ago

@mmirman @abhigya-sodani it has been commented out as per the present version anarchy the code has been commented out and exception has been raised over there. No fine tuning support for it

mmirman commented 12 months ago

@sushmanthreddy https://github.com/anarchy-ai/LLM-VM/blob/main/src/llm_vm/completion/optimize.py#L27 There is finetuning support for OAI just not in this file. I'm asking @VictorOdede and @cartazio why not in this file