Copy loss and metric to prevent side effect

keras/engine/compile_utils.py

@@ -107,7 +107,7 @@ def __init__(self, losses, loss_weights=None, output_names=None):
     self._user_losses = losses
     self._user_loss_weights = loss_weights
 
-    self._losses = losses
+    self._losses = copy.copy(losses)
     self._loss_weights = loss_weights
     self._per_output_metrics = None  # Per-output losses become metrics.
     self._loss_metric = metrics_mod.Mean(name='loss')  # Total loss.
@@ -309,7 +309,7 @@ def __init__(self, metrics=None, weighted_metrics=None, output_names=None,
     self._user_metrics = metrics
     self._user_weighted_metrics = weighted_metrics
 
-    self._metrics = metrics
+    self._metrics = copy.copy(metrics)
     self._weighted_metrics = weighted_metrics
     self._built = False
 

keras/engine/compile_utils_test.py

@@ -195,6 +195,19 @@ def test_nested_structure(self):
     self.assertEqual(b_1_metric.name, 'b_1_loss')
     self.assertEqual(b_1_metric.result().numpy(), 0.5)
 
+  def test_no_input_mutation(self):
+    loss = {'a': 'mae'}
+    loss_container = compile_utils.LossesContainer(loss)
+
+    y_t = {'a': tf.zeros((10, 1))}
+    y_p = {'a': tf.ones((10, 1)), 'b': tf.zeros((10, 1))}
+    sw = tf.convert_to_tensor([0, 0, 0, 0, 0, 1, 1, 1, 1, 1])
+
+    total_loss = loss_container(y_t, y_p, sample_weight=sw)
+    self.assertIsInstance(total_loss, tf.Tensor)
+    self.assertEqual(total_loss.numpy(), 0.5)
+    self.assertLen(loss, 1)
+
   def test_broadcast_single_loss(self):
     loss_container = compile_utils.LossesContainer('mse')
 
@@ -585,6 +598,18 @@ def test_nested_structure(self):
     self.assertEqual(b_1_mse_metric.name, 'b_1_mse')
     self.assertEqual(b_1_mse_metric.result().numpy(), 4.)
 
+  def test_no_input_mutation(self):
+    metric = {'a': 'mae'}
+    metric_container = compile_utils.MetricsContainer(metric)
+
+    y_t = {'a': tf.zeros((10, 1))}
+    y_p = {'a': tf.ones((10, 1)), 'b': tf.zeros((10, 1))}
+
+    metric_container.update_state(y_t, y_p)
+    self.assertLen(metric, 1)
+    mae_metric = metric_container.metrics[0]
+    self.assertEqual(mae_metric.result().numpy(), 1.)
+
   def test_crossentropy(self):
     metric_container = compile_utils.MetricsContainer('crossentropy')
     y_t, y_p = tf.ones((10, 1)), tf.ones((10, 1))