Preparing the Dataset

Load the train and test data using keras.datasets.fashion_mnist.load_data().

(X_train_full, y_train_full), (X_test, y_test) = << your code comes here >>

Scale the X_train_full and X_test data sets with 255.

X_train_full = X_train_full.astype(np.float32) / 255
X_test = X_test.astype(np.float32) / 255

Slice the X_train_full, such that the last 5000 samples form the validation data X_valid and the remaining samples form the train data X_train.

X_train, X_valid = << your code comes here >>[:-5000], << your code comes here >>[-5000:]

Similarly, slice the y_train_full, such that the last 5000 samples form the validation data labels y_valid and the remaining samples form the train data labels y_train.

y_train, y_valid = << your code comes here >>[:-5000], << your code comes here >>[-5000:]