WizardLM icon indicating copy to clipboard operation
WizardLM copied to clipboard
verified publisher icon nlpxucan

How to reproduce Evol-Instruct datasets?

Open imoneoi opened this issue 2 years ago • 5 comments

I've seen your open-source Evol-Instruct generation scripts. Good job!

Additionally, can you provide instructions on how to reproduce the WizardLM dataset and WizardCoder dataset using the scripts provided?

imoneoi avatar Sep 24 '23 08:09 imoneoi

You can just modify the file path you want to evol in WizardLM/Evol-Instruct/main.py and then run "python main.py". The default file is the alpaca data.

nlpxucan avatar Sep 24 '23 08:09 nlpxucan

What is the seed file for the WizardLM and WizardCoder datasets?

imoneoi avatar Sep 24 '23 10:09 imoneoi

BTW, the scripts seem to be missing the error checker and iterative evolution described in the paper. Are these parts necessary?

imoneoi avatar Sep 24 '23 10:09 imoneoi

Any updates?

imoneoi avatar Sep 30 '23 05:09 imoneoi

BTW, the scripts seem to be missing the error checker and iterative evolution described in the paper. Are these parts necessary?

+1

gantuo avatar Jan 17 '24 03:01 gantuo