remove pydot depndency

ShowerGAN_signal.py

@@ -80,9 +80,9 @@ def build_critic():
     return critic
 
 generator = build_generator(latent_size)
-plot_model(generator, to_file=log_dir + '/generator.png', show_shapes=True)
+#plot_model(generator, to_file=log_dir + '/generator.png', show_shapes=True)
 critic = build_critic()
-plot_model(critic, to_file=log_dir + '/critic.png', show_shapes=True)
+#plot_model(critic, to_file=log_dir + '/critic.png', show_shapes=True)
 
 # make trainings model for generator
 utils.make_trainable(critic, False)
@@ -93,7 +93,7 @@ generator_out = generator(generator_in)
 critic_out = critic(generator_out)
 generator_training = Model(inputs=generator_in, outputs=critic_out)
 generator_training.compile(optimizer=Adam(0.0001, beta_1=0.5, beta_2=0.9, decay=0.0), loss=[utils.wasserstein_loss])
-plot_model(generator_training, to_file=log_dir + '/generator_training.png', show_shapes=True)
+#plot_model(generator_training, to_file=log_dir + '/generator_training.png', show_shapes=True)
 
 # make trainings model for critic
 utils.make_trainable(critic, True)
@@ -110,7 +110,7 @@ critic_training = Model(inputs=[generator_in_critic_training, shower_in_critic_t
 gradient_penalty = partial(utils.gradient_penalty_loss, averaged_batch=averaged_batch, penalty_weight=GRADIENT_PENALTY_WEIGHT)
 gradient_penalty.__name__ = 'gradient_penalty'
 critic_training.compile(optimizer=Adam(0.0001, beta_1=0.5, beta_2=0.9, decay=0.0), loss=[utils.wasserstein_loss, utils.wasserstein_loss, gradient_penalty])
-plot_model(critic_training, to_file=log_dir + '/critic_training.png', show_shapes=True)
+#plot_model(critic_training, to_file=log_dir + '/critic_training.png', show_shapes=True)
 
 # For Wassersteinloss
 positive_y = np.ones(BATCH_SIZE)