Can someone explain to me how the training process works?

API
, ,
1

We are using the yaia WordPress plugin, and I’m trying to train this ChatGPT. I have tried fine-tuning locally, but I’m not quite sure how it works. Fine-tuning created a new model for me. Should I specify that model in this chat, or does training happen automatically after I submit a request via OpenAI using the OPENAI_API_KEY?

1 Like
2

Welcome to the Forums!

Sounds like there may be a misunderstanding here about Plugins and the API, the two things are not compatible ecosystems, so if you are using a WordPress Plugin with ChatGPT then you cannot use that with the API, you can fine tune base models but you cannot fine tune the 3.5-Turbo or GPT-4 models.

1 Like
3

It’s quite expensive to finetune a model, few-shot prompt tuning can work well already. If you still wanna go with finetuning path, you gonna need lots of data.

1 Like