diff --git a/examples/mnist_acgan.py b/examples/mnist_acgan.py
index b28e8ac1c..20d949b02 100644
--- a/examples/mnist_acgan.py
+++ b/examples/mnist_acgan.py
@@ -160,8 +160,6 @@ if __name__ == '__main__':
         loss=['binary_crossentropy', 'sparse_categorical_crossentropy']
     )
 
-    discriminator.trainable = True
-
     # get our mnist data, and force it to be of shape (..., 1, 28, 28) with
     # range [-1, 1]
     (X_train, y_train), (X_test, y_test) = mnist.load_data()
@@ -217,10 +215,7 @@ if __name__ == '__main__':
             noise = np.random.uniform(-1, 1, (2 * batch_size, latent_size))
             sampled_labels = np.random.randint(0, 10, 2 * batch_size)
 
-            # we want to fix the discriminator and let the generator train to
-            # trick it
-            discriminator.trainable = False
-
+            # we want to train the genrator to trick the discriminator
             # For the generator, we want all the {fake, not-fake} labels to say
             # not-fake
             trick = np.ones(2 * batch_size)
@@ -228,8 +223,6 @@ if __name__ == '__main__':
             epoch_gen_loss.append(combined.train_on_batch(
                 [noise, sampled_labels.reshape((-1, 1))], [trick, sampled_labels]))
 
-            discriminator.trainable = True
-
         print('\nTesting for epoch {}:'.format(epoch + 1))
 
         # evaluate the testing loss here