Fix preprocessing layers signatures.

PiperOrigin-RevId: 342531890
Change-Id: I543fcf6057e03647bf64c7dae7252194f329a31c

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

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras categorical preprocessing layers."""
+"""Keras category crossing preprocessing layers."""
 # pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division

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

@@ -12,7 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras text CategoryEncoding preprocessing layer."""
+"""Keras CategoryEncoding preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -86,8 +87,7 @@ class CategoryEncoding(base_preprocessing_layer.CombinerPreprocessingLayer):
            [0. , 0.2, 0.3, 0. ],
            [0. , 0.2, 0. , 0.4]])>
 
-
-  Attributes:
+  Arguments:
     max_tokens: The maximum size of the vocabulary for this layer. If None,
       there is no cap on the size of the vocabulary.
     output_mode: Specification for the output of the layer.

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

@@ -12,7 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras preprocessing layers."""
+"""Keras discretization preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function

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

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras categorical preprocessing layers."""
+"""Keras hashing preprocessing layer."""
 # pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division

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

@@ -12,7 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras text vectorization preprocessing layer."""
+"""Keras index lookup preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -57,7 +58,7 @@ class IndexLookup(base_preprocessing_layer.CombinerPreprocessingLayer):
   vocabulary size, the most frequent terms will be used to create the
   vocabulary.
 
-  Attributes:
+  Arguments:
     max_tokens: The maximum size of the vocabulary for this layer. If None,
       there is no cap on the size of the vocabulary. Note that this vocabulary
       includes the OOV and mask tokens, so the effective number of tokens is

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

@@ -13,6 +13,7 @@
 # limitations under the License.
 # ==============================================================================
 """Keras string lookup preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -39,7 +40,7 @@ class IntegerLookup(index_lookup.IndexLookup):
   vocabulary size, the most frequent terms will be used to create the
   vocabulary.
 
-  Attributes:
+  Arguments:
     max_values: The maximum size of the vocabulary for this layer. If None,
       there is no cap on the size of the vocabulary. Note that this vocabulary
       includes the OOV and mask values, so the effective number of values is

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

@@ -12,7 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras preprocessing layers."""
+"""Normalization preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -59,7 +60,7 @@ class Normalization(base_preprocessing_layer.CombinerPreprocessingLayer):
     as the layer's weights. `adapt` should be called before `fit`, `evaluate`,
     or `predict`.
 
-  Attributes:
+  Arguments:
       axis: Integer or tuple of integers, the axis or axes that should be
         "kept". These axes are not be summed over when calculating the
         normalization statistics. By default the last axis, the `features` axis
@@ -102,11 +103,7 @@ class Normalization(base_preprocessing_layer.CombinerPreprocessingLayer):
          [ 0.        ]], dtype=float32)>
   """
 
-  def __init__(self, axis=-1, dtype=None, mean=None, variance=None, **kwargs):
-    # This ensures that if the value of K.floatx() changes after file-loading
-    # time, the dtype value will change to reflect it.
-    dtype = dtype or K.floatx()
-
+  def __init__(self, axis=-1, mean=None, variance=None, **kwargs):
     # Standardize `axis` to a tuple.
     if axis is None:
       axis = ()
@@ -116,7 +113,7 @@ class Normalization(base_preprocessing_layer.CombinerPreprocessingLayer):
       axis = tuple(axis)
 
     super(Normalization, self).__init__(
-        combiner=_NormalizingCombiner(axis), dtype=dtype, **kwargs)
+        combiner=_NormalizingCombiner(axis), **kwargs)
     base_preprocessing_layer.keras_kpl_gauge.get_cell('Normalization').set(True)
 
     if 0 in axis:

tensorflow/python/keras/layers/preprocessing/normalization_v1.py

@@ -27,7 +27,7 @@ from tensorflow.python.util.tf_export import keras_export
 @keras_export(v1=['keras.layers.experimental.preprocessing.Normalization'])
 class Normalization(normalization.Normalization, CombinerPreprocessingLayer):
 
-  def __init__(self, axis=-1, dtype=None, **kwargs):
-    super(Normalization, self).__init__(axis, dtype, **kwargs)
+  def __init__(self, axis=-1, **kwargs):
+    super(Normalization, self).__init__(axis, **kwargs)
     base_preprocessing_layer.keras_kpl_gauge.get_cell(
         'Normalization v1').set(True)

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

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Keras categorical preprocessing layers."""
+"""Keras reduction layer."""
 # pylint: disable=g-classes-have-attributes
 
 from __future__ import absolute_import

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

@@ -13,6 +13,7 @@
 # limitations under the License.
 # ==============================================================================
 """Keras string lookup preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -39,7 +40,7 @@ class StringLookup(index_lookup.IndexLookup):
   vocabulary size, the most frequent terms will be used to create the
   vocabulary.
 
-  Attributes:
+  Arguments:
     max_tokens: The maximum size of the vocabulary for this layer. If None,
       there is no cap on the size of the vocabulary. Note that this vocabulary
       includes the OOV and mask tokens, so the effective number of tokens is

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

@@ -13,6 +13,7 @@
 # limitations under the License.
 # ==============================================================================
 """Keras text vectorization preprocessing layer."""
+# pylint: disable=g-classes-have-attributes
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
@@ -117,7 +118,7 @@ class TextVectorization(base_preprocessing_layer.CombinerPreprocessingLayer):
        ["another", "string", "to", "split"]]`. This makes the callable site
        natively compatible with `tf.strings.split()`.
 
-  Attributes:
+  Arguments:
     max_tokens: The maximum size of the vocabulary for this layer. If None,
       there is no cap on the size of the vocabulary. Note that this vocabulary
       contains 1 OOV token, so the effective number of tokens is `(max_tokens -
@@ -162,6 +163,7 @@ class TextVectorization(base_preprocessing_layer.CombinerPreprocessingLayer):
       times, an error will be thrown.
 
   Example:
+
   This example instantiates a TextVectorization layer that lowercases text,
   splits on whitespace, strips punctuation, and outputs integer vocab indices.
 
@@ -202,6 +204,7 @@ class TextVectorization(base_preprocessing_layer.CombinerPreprocessingLayer):
          [1, 3, 0, 0]])
 
   Example:
+
   This example instantiates a TextVectorization layer by passing a list
   of vocabulary terms to the layer's __init__ method.
 

tensorflow/tools/api/golden/v1/tensorflow.keras.layers.experimental.preprocessing.-normalization.pbtxt

@@ -133,7 +133,7 @@ tf_class {
   }
   member_method {
     name: "__init__"
-    argspec: "args=[\'self\', \'axis\', \'dtype\'], varargs=None, keywords=kwargs, defaults=[\'-1\', \'None\'], "
+    argspec: "args=[\'self\', \'axis\'], varargs=None, keywords=kwargs, defaults=[\'-1\'], "
   }
   member_method {
     name: "adapt"

tensorflow/tools/api/golden/v2/tensorflow.keras.layers.experimental.preprocessing.-normalization.pbtxt

@@ -131,7 +131,7 @@ tf_class {
   }
   member_method {
     name: "__init__"
-    argspec: "args=[\'self\', \'axis\', \'dtype\', \'mean\', \'variance\'], varargs=None, keywords=kwargs, defaults=[\'-1\', \'None\', \'None\', \'None\'], "
+    argspec: "args=[\'self\', \'axis\', \'mean\', \'variance\'], varargs=None, keywords=kwargs, defaults=[\'-1\', \'None\', \'None\'], "
   }
   member_method {
     name: "adapt"