I want to take the EMNIST dataset of letters to train a model to read these letters.
In order to give the model more training input I want to do augmentation on the t