The code does not inhibit the learning process of the discriminator

[fabbrimatteo]

The line discriminator.trainable = False does not stop the discriminator from learning. Replace that types of line with a call at the following function:

def make_trainable(net, val):
     net.trainable = val
     for l in net.layers:
         l.trainable = val

[jacobgil]

Thanks for the issue.
It seems that .trainable = False freezes the entire model.
Also discussed here: keras-team/keras#4510
Can you please elaborate why do you think it should be set for each layer separately?

[li-js]

After setting the flag, do we need to re-compile the model before the parameters can be frozen?
Here is a quotation from https://keras.io/getting-started/faq/

How can I "freeze" Keras layers?

Additionally, you can set the trainable property of a layer to True or False after instantiation. For this to take effect, you will need to call compile() on your model after modifying the trainable property

[fabbrimatteo]

@jacobgil i am confused by the fact that if you run:

model.summary()

before and after the command .trainable=False you can see that the number of trainable parameters does not change.