Remove references to private name_scope endpoint.

PiperOrigin-RevId: 322623654
Change-Id: Id82f764b7768fdf40db43bea1e32404cc220a850

tensorflow/python/keras/backend_test.py

@@ -169,7 +169,7 @@ class BackendUtilsTest(test.TestCase):
           sess.run(y, feed_dict={x: np.random.random((2, 3))})
 
   def test_learning_phase_name(self):
-    with ops.name_scope('test_scope'):
+    with backend.name_scope('test_scope'):
       # Test that outer name scopes do not affect the learning phase's name.
       lp = backend.symbolic_learning_phase()
     self.assertEqual(lp.name, 'keras_learning_phase:0')

tensorflow/python/keras/feature_column/dense_features.py

@@ -21,7 +21,6 @@ from __future__ import print_function
 import json
 
 from tensorflow.python.feature_column import feature_column_v2 as fc
-from tensorflow.python.framework import ops
 from tensorflow.python.keras import backend
 from tensorflow.python.keras.feature_column import base_feature_layer as kfc
 from tensorflow.python.keras.saving.saved_model import json_utils
@@ -161,7 +160,7 @@ class DenseFeatures(kfc._BaseFeaturesLayer):  # pylint: disable=protected-access
     transformation_cache = fc.FeatureTransformationCache(features)
     output_tensors = []
     for column in self._feature_columns:
-      with ops.name_scope(column.name):
+      with backend.name_scope(column.name):
         try:
           tensor = column.get_dense_tensor(
               transformation_cache, self._state_manager, training=training)

tensorflow/python/keras/feature_column/dense_features_v2.py

@@ -89,7 +89,7 @@ class DenseFeatures(dense_features.DenseFeatures):
 
   def build(self, _):
     for column in self._feature_columns:
-      with ops.name_scope(column.name):
+      with ops.name_scope_v2(column.name):
         column.create_state(self._state_manager)
     # We would like to call Layer.build and not _DenseFeaturesHelper.build.
     # pylint: disable=protected-access

tensorflow/python/keras/feature_column/sequence_feature_column.py

@@ -143,7 +143,7 @@ class SequenceFeatures(kfc._BaseFeaturesLayer):
     sequence_lengths = []
 
     for column in self._feature_columns:
-      with ops.name_scope(column.name):
+      with backend.name_scope(column.name):
         try:
           dense_tensor, sequence_length = column.get_sequence_dense_tensor(
               transformation_cache, self._state_manager, training=training)
@@ -164,7 +164,7 @@ class SequenceFeatures(kfc._BaseFeaturesLayer):
 
 def _assert_all_equal_and_return(tensors, name=None):
   """Asserts that all tensors are equal and returns the first one."""
-  with ops.name_scope(name, 'assert_all_equal', values=tensors):
+  with backend.name_scope(name or 'assert_all_equal'):
     if len(tensors) == 1:
       return tensors[0]
     assert_equal_ops = []

tensorflow/python/keras/layers/legacy_rnn/rnn_cell_impl.py

@@ -33,6 +33,7 @@ from tensorflow.python.framework import ops
 from tensorflow.python.framework import tensor_shape
 from tensorflow.python.framework import tensor_util
 from tensorflow.python.keras import activations
+from tensorflow.python.keras import backend
 from tensorflow.python.keras import initializers
 from tensorflow.python.keras.engine import base_layer_utils
 from tensorflow.python.keras.engine import input_spec
@@ -334,7 +335,7 @@ class RNNCell(base_layer.Layer):
       if (last_batch_size == batch_size and last_dtype == dtype and
           last_state_size == state_size):
         return last_output
-    with ops.name_scope(type(self).__name__ + "ZeroState", values=[batch_size]):
+    with backend.name_scope(type(self).__name__ + "ZeroState"):
       output = _zero_state_tensors(state_size, batch_size, dtype)
     if is_eager:
       self._last_zero_state = (state_size, batch_size, dtype, output)
@@ -1269,7 +1270,7 @@ class MultiRNNCell(RNNCell):
     return self._cells[-1].output_size
 
   def zero_state(self, batch_size, dtype):
-    with ops.name_scope(type(self).__name__ + "ZeroState", values=[batch_size]):
+    with backend.name_scope(type(self).__name__ + "ZeroState"):
       if self._state_is_tuple:
         return tuple(cell.zero_state(batch_size, dtype) for cell in self._cells)
       else:

tensorflow/python/keras/layers/legacy_rnn/rnn_cell_wrapper_impl.py

@@ -113,7 +113,7 @@ class DropoutWrapperBase(object):
       raise TypeError("dropout_state_filter_visitor must be callable")
     self._dropout_state_filter = (
         dropout_state_filter_visitor or _default_dropout_state_filter_visitor)
-    with ops.name_scope("DropoutWrapperInit"):
+    with ops.name_scope_v2("DropoutWrapperInit"):
 
       def tensor_and_const_value(v):
         tensor_value = ops.convert_to_tensor(v)
@@ -199,7 +199,7 @@ class DropoutWrapperBase(object):
     self.built = True
 
   def zero_state(self, batch_size, dtype):
-    with ops.name_scope(type(self).__name__ + "ZeroState", values=[batch_size]):
+    with ops.name_scope_v2(type(self).__name__ + "ZeroState"):
       return self.cell.zero_state(batch_size, dtype)
 
   def _variational_recurrent_dropout_value(
@@ -346,7 +346,7 @@ class ResidualWrapperBase(object):
     return self.cell.output_size
 
   def zero_state(self, batch_size, dtype):
-    with ops.name_scope(type(self).__name__ + "ZeroState", values=[batch_size]):
+    with ops.name_scope_v2(type(self).__name__ + "ZeroState"):
       return self.cell.zero_state(batch_size, dtype)
 
   def _call_wrapped_cell(self, inputs, state, cell_call_fn, **kwargs):
@@ -433,7 +433,7 @@ class DeviceWrapperBase(object):
     return self.cell.output_size
 
   def zero_state(self, batch_size, dtype):
-    with ops.name_scope(type(self).__name__ + "ZeroState", values=[batch_size]):
+    with ops.name_scope_v2(type(self).__name__ + "ZeroState"):
       with ops.device(self._device):
         return self.cell.zero_state(batch_size, dtype)
 

tensorflow/python/keras/layers/normalization_v2.py

@@ -21,7 +21,7 @@ from __future__ import print_function
 from tensorflow.python.distribute import distribution_strategy_context as ds
 from tensorflow.python.distribute import reduce_util
 from tensorflow.python.framework import dtypes
-from tensorflow.python.framework import ops
+from tensorflow.python.keras import backend
 from tensorflow.python.keras.layers import normalization
 from tensorflow.python.ops import array_ops
 from tensorflow.python.ops import math_ops
@@ -158,7 +158,7 @@ class SyncBatchNormalization(normalization.BatchNormalizationBase):
 
   def _calculate_mean_and_var(self, x, axes, keep_dims):
 
-    with ops.name_scope('moments', values=[x, axes]):
+    with backend.name_scope('moments'):
       # The dynamic range of fp16 is too limited to support the collection of
       # sufficient statistics. As a workaround we simply perform the operations
       # on 32-bit floats before converting the mean and variance back to fp16

tensorflow/python/keras/layers/preprocessing/image_preprocessing.py

@@ -585,7 +585,7 @@ def get_translation_matrix(translations, name=None):
     A tensor of shape (num_images, 8) projective transforms which can be given
       to `transform`.
   """
-  with ops.name_scope(name, 'translation_matrix'):
+  with K.name_scope(name or 'translation_matrix'):
     num_translations = array_ops.shape(translations)[0]
     # The translation matrix looks like:
     #     [[1 0 -dx]
@@ -665,7 +665,7 @@ def transform(images,
     TypeError: If `image` is an invalid type.
     ValueError: If output shape is not 1-D int32 Tensor.
   """
-  with ops.name_scope(name, 'transform'):
+  with K.name_scope(name or 'transform'):
     if output_shape is None:
       output_shape = array_ops.shape(images)[1:3]
       if not context.executing_eagerly():
@@ -708,7 +708,7 @@ def get_rotation_matrix(angles, image_height, image_width, name=None):
        `(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
        where `k = c0 x + c1 y + 1`.
   """
-  with ops.name_scope(name, 'rotation_matrix'):
+  with K.name_scope(name or 'rotation_matrix'):
     x_offset = ((image_width - 1) - (math_ops.cos(angles) *
                                      (image_width - 1) - math_ops.sin(angles) *
                                      (image_height - 1))) / 2.0
@@ -1015,7 +1015,7 @@ def get_zoom_matrix(zooms, image_height, image_width, name=None):
        `(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
        where `k = c0 x + c1 y + 1`.
   """
-  with ops.name_scope(name, 'zoom_matrix'):
+  with K.name_scope(name or 'zoom_matrix'):
     num_zooms = array_ops.shape(zooms)[0]
     # The zoom matrix looks like:
     #     [[zx 0 0]

tensorflow/python/keras/legacy_tf_layers/base.py

@@ -440,7 +440,7 @@ class Layer(base_layer.Layer):
     with vs.variable_scope(
         self._scope, reuse=reuse, auxiliary_name_scope=False) as scope:
       self._current_scope = scope
-      with ops.name_scope(self._name_scope(), skip_on_eager=False):
+      with backend.name_scope(self._name_scope()):
         use_resource = (use_resource or
                         self._use_resource_variables or
                         scope.use_resource)

tensorflow/python/keras/legacy_tf_layers/base_test.py

@@ -28,6 +28,7 @@ from tensorflow.python.eager import def_function
 from tensorflow.python.framework import constant_op
 from tensorflow.python.framework import dtypes
 from tensorflow.python.framework import ops
+from tensorflow.python.keras import backend
 from tensorflow.python.keras import combinations
 from tensorflow.python.keras.engine import base_layer as keras_base_layer
 from tensorflow.python.keras.engine import input_spec
@@ -68,12 +69,12 @@ class BaseLayerTest(test.TestCase, parameterized.TestCase):
   @combinations.generate(combinations.combine(mode=['graph', 'eager']))
   def testKerasStyleAddWeight(self):
     keras_layer = keras_base_layer.Layer(name='keras_layer')
-    with ops.name_scope('foo', skip_on_eager=False):
+    with backend.name_scope('foo'):
       keras_variable = keras_layer.add_variable(
           'my_var', [2, 2], initializer=init_ops.zeros_initializer())
     self.assertEqual(keras_variable.name, 'foo/my_var:0')
 
-    with ops.name_scope('baz', skip_on_eager=False):
+    with backend.name_scope('baz'):
       old_style_layer = base_layers.Layer(name='my_layer')
       # Test basic variable creation.
       variable = old_style_layer.add_variable(
@@ -83,7 +84,7 @@ class BaseLayerTest(test.TestCase, parameterized.TestCase):
     with base_layers.keras_style_scope():
       layer = base_layers.Layer(name='my_layer')
     # Test basic variable creation.
-    with ops.name_scope('bar', skip_on_eager=False):
+    with backend.name_scope('bar'):
       variable = layer.add_variable(
           'my_var', [2, 2], initializer=init_ops.zeros_initializer())
     self.assertEqual(variable.name, 'bar/my_var:0')
@@ -241,7 +242,7 @@ class BaseLayerTest(test.TestCase, parameterized.TestCase):
     my_layer2.apply(inputs)
     self.assertEqual(my_layer2._scope.name, 'my_layer_2')
     # Name scope shouldn't affect names.
-    with ops.name_scope('some_name_scope'):
+    with backend.name_scope('some_name_scope'):
       default_layer2 = PrivateLayer()
       default_layer2.apply(inputs)
       self.assertEqual(default_layer2._scope.name, 'private_layer_2')

tensorflow/python/keras/tests/tracking_util_with_v1_optimizers_test.py

@@ -465,7 +465,7 @@ class CheckpointingTests(keras_parameterized.TestCase):
         root, name=name, shape=[1, 2], dtype=dtypes.float64)
     (named_variable,), _, _ = trackable_utils._serialize_object_graph(
         root, saveables_cache=None)
-    with ops.name_scope("root/" + named_variable.name):
+    with ops.name_scope_v2("root/" + named_variable.name):
       pass  # Make sure we can use this as an op name if we prefix it.
     return named_variable.name
 

tensorflow/python/keras/utils/losses_utils.py

@@ -61,8 +61,7 @@ def remove_squeezable_dimensions(
   Returns:
     Tuple of `labels` and `predictions`, possibly with last dim squeezed.
   """
-  with ops.name_scope(name, 'remove_squeezable_dimensions',
+  with K.name_scope(name or 'remove_squeezable_dimensions'):
-                      [labels, predictions]):
     predictions = ops.convert_to_tensor_v2_with_dispatch(predictions)
     labels = ops.convert_to_tensor_v2_with_dispatch(labels)
     predictions_shape = predictions.get_shape()