How do we testing on different dataset?

[ShuuTsubaki]

For example, if we want to test it on imagenet-12, are we going to create a similiar py script likes cifar 10 and make change on the cmd_args and get_data_loders function. Moreover, I received following output after I want to run !python train.py --label-corrupt-prob=1.0, Traceback (most recent call last): File "train.py", line 340, in main() File "train.py", line 333, in main train_loader, val_loader = get_data_loaders(args, shuffle_train=True) File "train.py", line 81, in get_data_loaders corrupt_prob=args.label_corrupt_prob), File "/content/cifar10_data.py", line 25, in init self.corrupt_labels(corrupt_prob) File "/content/cifar10_data.py", line 28, in corrupt_labels labels = np.array(self.train_labels if self.train else self.test_labels) AttributeError: 'CIFAR10RandomLabels' object has no attribute 'train_labels'