Keras how to reshape MNIST images after dimensionality reduction

Question

I have been running a keras sequential model on mnist, and the first layer is as follows:

model.add(Reshape((28, 28, 1), input_shape=(X_train.shape[1],)))
mod