Clarify how to use the dataset(include external)

Thanks for holding this competition.
This competition is very exciting for me.

this competition data rule is very complex, I would like to clarify.

  1. finetune is ok? or not?(e.g. LoRA)
  2. LLM(e.g. chatgpt) based new problems generation is ok? or not?(not re labeling)
    I understand public dataset will become public after competiton.
1 Like

Hi, thanks for your interest in the challenge.

  1. Fine-tuning, of course yes, as long as the base model is publicly available.
  2. ChatGPT based new problem generation, of course yes, as long as the re-labeling is based on existing datasets.

In short, both of them are fine, but you should make both resources available after the challenge, and make sure you are developing based on publicly available resources.

Hi! Thank you for reply!
I understand your response.