legacy fine-tuning

[rdlfperezIII]

Hello,
I've noticed that the ai-training module JSONL format being sent is for legacy fine tuning models such as babbage-002 and davinci-002. Will there be an OOB support for the current fine-tuning method ?

Thanks in advance

[joaquin-alfaro]

Thanks for your suggestion, yes we should support both legacy and new models of OpenAI. If you have any proposal of implementation, feel free to open a pull request.

Thanks @rdlfperezIII