News for hackers. Stuff that matters.

A comment on Exploring GPTs: ChatGPT in a trench coat?.

Re:Exploring GPTs: ChatGPT in a trench coat? by sinuhe69 on Wednesday November 15, 2023 @11:59AM (#38279006)
Does anyone know how they can run it so cheaply? Fine-tuning is fairly cheap for ChatGPT. But if it’s truly a fine-tuned model, then they have to run a copy of the customer’s model, isn’t it? How can they provision it so quickly? Or did they know how to do sort of delta with the last layers?
[ Parent | on HN ]
Re:Exploring GPTs: ChatGPT in a trench coat? by freedomben on Wednesday November 15, 2023 @12:02PM (#38279063)
It's not a different model, it's just a customized system prompt. It costs basically nothing extra for them to run. Actual fine-tuning is $$$
[ Parent | on HN ]
Re:Exploring GPTs: ChatGPT in a trench coat? by spdustin on Wednesday November 15, 2023 @01:32PM (#38280433)
Technically, it is (it’s gpt-4-gizmo). There’s a chance that had some fine-tuning for understanding the presence of the just-in-time function calls for actions.

(Names for Actions in Custom GPTs are tagged with “jit”)

[ Parent | on HN ]
Re:Exploring GPTs: ChatGPT in a trench coat? by siva7 on Wednesday November 15, 2023 @12:21PM (#38279404)
It has nothing to do with fine-tuning - like most of these startups using the openai api. It's basically all clever prompt engineering techniques that emerged over the last years finally combined into ChatGPT.
[ Parent | on HN ]
Re:Exploring GPTs: ChatGPT in a trench coat? by agodfrey on Wednesday November 15, 2023 @12:03PM (#38279089)
It isn’t a finetuned model. Not sure what exactly it is, but it seems like just a bunch of prompt engineering.
[ Parent | on HN ]
Re:Exploring GPTs: ChatGPT in a trench coat? by minimaxir on Wednesday November 15, 2023 @12:03PM (#38279090)
There is likely zero finetuning: everything GPTs can do can also be done with clever system prompt engineering.

GPTs likely use extra prompt engineering to align everything.

[ Parent | on HN ]