Fix RNN layers dynamic trainable attr

keras/layers/recurrent.py

@@ -731,13 +731,17 @@ def from_config(cls, config, custom_objects=None):
 
     @property
     def trainable_weights(self):
+        if not self.trainable:
+            return []
         if isinstance(self.cell, Layer):
             return self.cell.trainable_weights
         return []
 
     @property
     def non_trainable_weights(self):
         if isinstance(self.cell, Layer):
+            if not self.trainable:
+                return self.cell.weights
             return self.cell.non_trainable_weights
         return []
 

tests/keras/layers/recurrent_test.py

@@ -212,6 +212,23 @@ def test_regularizer(layer_class):
     assert len(layer.get_losses_for(x)) == 1
 
 
+@rnn_test
+def test_trainability(layer_class):
+    layer = layer_class(units)
+    layer.build((None, None, embedding_dim))
+    assert len(layer.weights) == 3
+    assert len(layer.trainable_weights) == 3
+    assert len(layer.non_trainable_weights) == 0
+    layer.trainable = False
+    assert len(layer.weights) == 3
+    assert len(layer.trainable_weights) == 0
+    assert len(layer.non_trainable_weights) == 3
+    layer.trainable = True
+    assert len(layer.weights) == 3
+    assert len(layer.trainable_weights) == 3
+    assert len(layer.non_trainable_weights) == 0
+
+
 @keras_test
 def test_masking_layer():
     ''' This test based on a previously failing issue here: