stanford_alpaca icon indicating copy to clipboard operation
stanford_alpaca copied to clipboard

Published 20 hours ago • verified publisher icon tatsu-lab

Any plans for using GPT-4 for self-instruct? Or using larger llama models?

Open JBX2060 opened this issue 2 years ago • 2 comments

I'm just curious

JBX2060 avatar Mar 16 '23 22:03 JBX2060

Hi, I have tried converting the self-instruction code and prompts runnable with ChatCompletion API (support turbo and gpt4), and does not affect the original Davinci generation. It looks work fine. I'm happy to open a pull request if authors also looking into this @Tiiiger

ArvinZhuang avatar Apr 01 '23 01:04 ArvinZhuang

嗨，我尝试将可使用 ChatCompletion API（支持 turbo 和 gpt4）运行的自指令代码和提示转换为自指令代码和提示，并且不会影响原始的 Davinci 生成。它看起来工作正常。如果作者也在研究这个问题，我很高兴打开一个拉取请求

could you tell me how to use gpt4o for self-instruct?

lichao1237 avatar Aug 05 '24 10:08 lichao1237