adding hints on how to use conv.py

Author: mnielsen

src/conv.py

@@ -106,6 +106,9 @@ def dbl_conv_relu():
                 SoftmaxLayer(n_in=100, n_out=10)], mini_batch_size)
             net.SGD(training_data, 60, mini_batch_size, 0.03, validation_data, test_data, lmbda=lmbda)
 
+#### Some subsequent functions may make use of the expanded MNIST
+#### data.  That can be generated by running expand_mnist.py.
+
 def expanded_data():
     expanded_training_data, _, _ = network3.load_data_shared(
         "../data/mnist_expanded.pkl.gz")