champ
champ copied to clipboard
Train & Test Data Split
Hi, I want to know the train/test split on the TikTok and UBCFashion dataset. Could you please provide how you split the academic public dataset into training and testing? According to https://github.com/Wangt-CN/DisCo/issues/27#issuecomment-1665173783 , they use additional collected data for evaluation.