Question of RandAugment

[likanchuan]

Hi! Thanks for your awesome work! I have question of producing results with RandAugment in the ablation study.

I have used the torch.transform.RandAugment as the only data augmentation to train on CIFAR10-LT with num_ops=3,magnitude=4 and gotten worse performance than before. I suppose I use it wrongly and I want to know what is the parameters of RandAugment used in your experiments and should I use it as the only data augmentation.