Minor improvement in AdversarialRegularization.

PiperOrigin-RevId: 294286252

Author: csferng

neural_structured_learning/keras/adversarial_regularization.py

@@ -503,7 +503,7 @@ def compile(self,
         loss_weights=self._compile_arg_loss_weights,
         **kwargs)
 
-    if hasattr(self.base_model, 'output_names'):
+    if getattr(self.base_model, 'output_names', None):
       # Organizes losses after the base model is fully compiled. The output
       # names from the base model is needed in case the loss (and/or
       # loss_weights) is specified in a dict().