The Omniglot dataset has about 963 characters in the training set, but in your work you claimed to use 1200 characters from training set for your task-incremental continual learning setup.
Can you please provide the omni_train_rot.npy file referred in the Omniglot dataset? The number of classes used in your experiments do not match up to the available training set of Omniglot.
The Omniglot dataset has about 963 characters in the training set, but in your work you claimed to use 1200 characters from training set for your task-incremental continual learning setup.
Can you please provide the omni_train_rot.npy file referred in the Omniglot dataset? The number of classes used in your experiments do not match up to the available training set of Omniglot.