Hello, thanks for your excellent paper and code! I want to follow your great work but got some problems.
I download the image set of SIM10K dataset from the official website and found the split is the combination of both the training set and the validation set. I want to know how the authors split the official set into a training set and a validation set?