WizardMath training problem:

nlpxucan / WizardLM

LLMs build upon Evol Insturct: WizardLM, WizardCoder, WizardMath

9.22k stars 715 forks source link

WizardMath training problem: #168

Open ahong007007 opened 1 year ago

ahong007007 commented 1 year ago

Thanks for your work and open source spirit.

WizardMath training problem: The paper claims to use Reinforced Evol-Instruct, but there is no relevant content for code training, which is similar to WizardLM/WizardCoder.

How about the training process Reinforced Evol-Instruct method?

thank you very much!