Re:Gorilla: LLM Connected with APIs
by elexhobby on Friday May 26, 2023 @12:46AM
(#36080055)
Not sure if this is obvious. But its incorrect to ditch on GPT-4. The paper uses self-instruct on GPT-4 to generate the training data on which it is fine-tuned. This paper would not exist without GPT-4. Although they claim GPT-4 can be replaced by any LLM, I'm sure the results would be nowhere as good, and so they stuck with GPT-4.
[ Parent
| on HN ]
Re:Gorilla: LLM Connected with APIs
by fnordpiglet on Friday May 26, 2023 @12:55AM
(#36080127)
It does make me wonder what the converged fixed point on this technique is. If I fine tune with GPT4 to make model A, which then performs better than GPT4, then fine tune model B with A, at what point does either artifacting or diminishing returns set in?
GPT-4 is powerful over a diverse set of tasks. They use it to build a model which is better for a narrow sub-task. Pretty sure the model is sub-optimal to GPT-4 for everything else.
Re:Gorilla: LLM Connected with APIs
by fnordpiglet on Friday May 26, 2023 @09:49AM
(#36084095)
Yeah, but there have been papers being published in general LLMs as well finetuned off of GPT4 instead of humans. Even in the narrow space the question remains. If I build a superior model for task X using gpt4 and it’s superior at X, can I think use my new model to train another model in X and continue to see benefits ?
For another possibility, see https://arxiv.org/abs/2305.15717. The new models may not actually be better - the evaluation may be broken.