Merge pull request keras-team#15543 from kianmeng:fix-typos

PiperOrigin-RevId: 407106182

CONTRIBUTING.md

@@ -136,7 +136,7 @@ python3 -m venv venv_dir
 ```
 
 You can activate the venv with the following command. You should always run the
-tests with the venv activated. You need to activate the venv everytime you open
+tests with the venv activated. You need to activate the venv every time you open
 a new shell.
 
 ```shell
@@ -185,12 +185,12 @@ in the `BUILD` file.
 ### Run a single test case
 
 The best way to run a single test case is to comment out the rest of the test
-cases in a file before runing the test file.
+cases in a file before running the test file.
 
 ### Run all tests
 
-You can run all the tests locally by running the following commmand
-in the repo root directory.
+You can run all the tests locally by running the following command in the repo
+root directory.
 
 ```
 bazel test --test_timeout 300,450,1200,3600 --test_output=errors --keep_going --define=use_fast_cpp_protos=false --build_tests_only --build_tag_filters=-no_oss --test_tag_filters=-no_oss keras/...

keras/applications/mobilenet_v2.py

@@ -481,7 +481,7 @@ def _inverted_res_block(inputs, expansion, stride, alpha, filters, block_id):
 
   x = layers.ReLU(6., name=prefix + 'depthwise_relu')(x)
 
-  # Project wiht a pointwise 1x1 convolution.
+  # Project with a pointwise 1x1 convolution.
   x = layers.Conv2D(
       pointwise_filters,
       kernel_size=1,

keras/benchmarks/BUILD

@@ -27,7 +27,7 @@ filegroup(
 # to the regular expression is executed.
 # e.g. --benchmarks=".*lstm*." will run all lstm layer related benchmarks.
 
-# Add all benchmarks related utils here for pip testing dependencis.
+# Add all benchmarks related utils here for pip testing dependencies.
 py_library(
     name = "keras_benchmark_lib_pip",
     srcs_version = "PY3",

keras/benchmarks/keras_examples_benchmarks/antirectifier_benchmark_test.py

@@ -128,7 +128,7 @@ def benchmark_antirectifier_bs_512_gpu_2(self):
 
 
 class Antirectifier(tf.keras.layers.Layer):
-  """Build simple custome layer."""
+  """Build simple custom layer."""
 
   def __init__(self, initializer="he_normal", **kwargs):
     super(Antirectifier, self).__init__(**kwargs)

keras/callbacks.py

@@ -599,7 +599,7 @@ class Callback:
 
   1. You should pack all your callbacks into a single `callbacks.CallbackList`
      so they can all be called together.
-  2. You will need to manually call all the `on_*` methods at the apropriate
+  2. You will need to manually call all the `on_*` methods at the appropriate
      locations in your loop. Like this:
 
      ```
@@ -2360,7 +2360,7 @@ def _init_profile_batch(self, profile_batch):
         of positive integers signify a range of batches to profile.
 
     Raises:
-      ValueError: If profile_batch is not an integer or a comma seperated pair
+      ValueError: If profile_batch is not an integer or a comma separated pair
                   of positive integers.
 
     """

keras/distribute/BUILD

@@ -771,7 +771,7 @@ distribute_py_test(
     shard_count = 50,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "nomultivm",  # TODO(b/170502145)
         "notsan",  # TODO(b/184542721)
@@ -795,7 +795,7 @@ distribute_py_test(
     shard_count = 21,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "nomultivm",  # TODO(b/170502145)
         "notsan",  # TODO(b/184542721)

keras/engine/BUILD

@@ -516,7 +516,7 @@ tf_py_test(
     srcs = ["training_utils_v1_test.py"],
     python_version = "PY3",
     tags = [
-        "no_oss",  # TODO(b/135021748) reenable
+        "no_oss",  # TODO(b/135021748) re-enable
         "notsan",
     ],
     deps = [

keras/engine/base_layer.py

@@ -668,7 +668,7 @@ def getter(*args, **kwargs):  # pylint: disable=function-redefined
         return autocast_variable.create_autocast_variable(variable)
       # Also the caching_device does not work with the mixed precision API,
       # disable it if it is specified.
-      # TODO(b/142020079): Reenable it once the bug is fixed.
+      # TODO(b/142020079): Re-enable it once the bug is fixed.
       if caching_device is not None:
         tf_logging.warning(
             '`caching_device` does not work with mixed precision API. Ignoring '
@@ -3410,7 +3410,7 @@ def _apply_name_scope_on_model_declaration(enable):
 
 
 class BaseRandomLayer(Layer):
-  """A layer handle the random nubmer creation and savemodel behavior."""
+  """A layer handle the random number creation and savemodel behavior."""
 
   @tf.__internal__.tracking.no_automatic_dependency_tracking
   def __init__(self, seed=None, force_generator=False, **kwargs):
@@ -3430,7 +3430,7 @@ def __init__(self, seed=None, force_generator=False, **kwargs):
       seed: optional integer, used to create RandomGenerator.
       force_generator: boolean, default to False, whether to force the
         RandomGenerator to use the code branch of tf.random.Generator.
-      **kwargs: other keyward arguements that will be passed to the parent class
+      **kwargs: other keyword arguments that will be passed to the parent class
     """
     super().__init__(**kwargs)
     self._random_generator = backend.RandomGenerator(

keras/engine/base_layer_v1.py

@@ -413,7 +413,7 @@ def getter(*args, **kwargs):  # pylint: disable=function-redefined
         return autocast_variable.create_autocast_variable(variable)
       # Also the caching_device does not work with the mixed precision API,
       # disable it if it is specified.
-      # TODO(b/142020079): Reenable it once the bug is fixed.
+      # TODO(b/142020079): Re-enable it once the bug is fixed.
       if caching_device is not None:
         tf_logging.warning(
             '`caching_device` does not work with mixed precision API. Ignoring '

keras/engine/data_adapter.py

@@ -74,7 +74,7 @@ def can_handle(x, y=None):
     """Whether the current DataAdapter could handle the input x and y.
 
     Structure wise, x and y can be single object, or list of objects if there
-    multiple input/output, or dictionary of objects when the intput/output are
+    multiple input/output, or dictionary of objects when the input/output are
     named.
 
     Args:

keras/engine/functional.py

@@ -715,7 +715,7 @@ def from_config(cls, config, custom_objects=None):
         return model
       # The config does not contain all the information necessary to revive a
       # Functional model. This happens when the user creates subclassed models
-      # with a Functional constructor and has overriden the `get_config` method
+      # with a Functional constructor and has overridden the `get_config` method
       # to return a completely new dictionary.
       try:
         return cls(**config)
@@ -769,7 +769,7 @@ def _validate_graph_inputs_and_outputs(self):
         for x in self.inputs])
     input_batch_sizes.discard(None)
     if len(input_batch_sizes) > 1:
-      logging.warning('Found incompatiable static batch sizes among the '
+      logging.warning('Found incompatible static batch sizes among the '
                       f'inputs. Batch sizes: {sorted(input_batch_sizes)}')
 
     for x in self.outputs:

keras/engine/functional_utils_test.py

@@ -138,7 +138,7 @@ def test_build_model_from_intermediate_tensor(self):
     loaded_model = models.load_model(output_path)
     self.assertEqual(model.summary(), loaded_model.summary())
 
-    # Also make sure the orignal inputs and y can still be used to build model
+    # Also make sure the original inputs and y can still be used to build model
     new_model = models.Model(inputs, y)
     # Make sure no new node is attached to layer2
     self.assertLen(layer2.inbound_nodes, 2)

keras/engine/training_test.py

@@ -1125,7 +1125,7 @@ def call(self, inputs):
       training_module.Model([input1, input2], outputs)
       self.assertEqual(
           mock_warn.call_args_list[0][0][0],
-          'Found incompatiable static batch sizes among the inputs. '
+          'Found incompatible static batch sizes among the inputs. '
           'Batch sizes: [2, 3]')
 
   @combinations.generate(combinations.combine(mode=['graph', 'eager']))
@@ -3511,7 +3511,7 @@ def call(self, x):
 
   @keras_parameterized.run_all_keras_modes(always_skip_v1=True)
   def DISABLED_test_add_metric_invalid_aggregation(self):
-    # TODO(psv): Reenable test once it is fixed.
+    # TODO(psv): Re-enable test once it is fixed.
     x = layers_module.Input(shape=(1,))
     y = layers_module.Dense(1, kernel_initializer='ones')(x)
     model = training_module.Model(x, y)

keras/initializers/initializers_v1.py

@@ -77,7 +77,7 @@ class RandomNormal(tf.compat.v1.random_normal_initializer):
   Random seed behavior:
   Also be aware that if you pass a seed to the TF2 initializer
   API it will reuse that same seed for every single initialization
-  (unlike the TF1 intializer)
+  (unlike the TF1 initializer)
 
   #### Structural Mapping to Native TF2
 
@@ -192,7 +192,7 @@ class RandomUniform(tf.compat.v1.random_uniform_initializer):
 
   Also be aware that if you pass a seed to the TF2 initializer
   API it will reuse that same seed for every single initialization
-  (unlike the TF1 intializer)
+  (unlike the TF1 initializer)
 
   #### Structural Mapping to Native TF2
 
@@ -313,7 +313,7 @@ class TruncatedNormal(tf.compat.v1.truncated_normal_initializer):
   Random seed behavior:
   Also be aware that if you pass a seed to the TF2 initializer
   API it will reuse that same seed for every single initialization
-  (unlike the TF1 intializer)
+  (unlike the TF1 initializer)
 
   #### Structural Mapping to Native TF2
 

keras/integration_test/BUILD

@@ -170,7 +170,7 @@ distribute_py_test(
     shard_count = 50,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "noasan",  # TODO(b/184542721)
         "nomsan",  # TODO(b/184542721)
@@ -216,7 +216,7 @@ distribute_py_test(
     shard_count = 50,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "noasan",  # TODO(b/184542721)
         "nomsan",  # TODO(b/184542721)
@@ -237,7 +237,7 @@ distribute_py_test(
     shard_count = 50,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "noasan",  # TODO(b/184542721)
         "nomsan",  # TODO(b/184542721)
@@ -258,7 +258,7 @@ distribute_py_test(
     shard_count = 50,
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "noasan",  # TODO(b/184542721)
         "nomsan",  # TODO(b/184542721)
@@ -278,7 +278,7 @@ distribute_py_test(
     python_version = "PY3",
     tags = [
         "multi_gpu",
-        "no_oss",  # TODO(b/183640564): Reenable
+        "no_oss",  # TODO(b/183640564): Re-enable
         "no_rocm",
         "noasan",  # TODO(b/184542721)
         "nomsan",  # TODO(b/184542721)

keras/integration_test/parameter_server_keras_preprocessing_test.py

@@ -258,7 +258,7 @@ def serve_fn(raw_features):
       self.assertIn(prediction0, ("yes", "no"))
 
       prediction1 = loaded_serving_fn(
-          tf.constant(["ironman", "ironman", "unkonwn"]))["output_0"]
+          tf.constant(["ironman", "ironman", "unknown"]))["output_0"]
       self.assertIn(prediction1, ("yes", "no"))
 
 

keras/integration_test/preprocessing_test_utils.py

@@ -66,7 +66,7 @@ def make_preprocessing_model(file_dir):
   normalization = preprocessing.Normalization()
   normalization.adapt(ds.map(lambda features, labels: features["float_col"]))
   float_out = normalization(float_in)
-  # Lookup ints by adapting a vocab of interger IDs.
+  # Lookup ints by adapting a vocab of integer IDs.
   int_lookup = preprocessing.IntegerLookup()
   int_lookup.adapt(ds.map(lambda features, labels: features["int_col"]))
   int_out = int_lookup(int_in)

keras/integration_test/tpu_strategy_test.py

@@ -232,7 +232,7 @@ def serve_fn(raw_features):
     self.assertIn(prediction1, ("yes", "no"))
 
     prediction2 = loaded_serving_fn(
-        tf.constant(["ironman", "ironman", "unkonwn"]))["output_0"]
+        tf.constant(["ironman", "ironman", "unknown"]))["output_0"]
     self.assertIn(prediction2, ("yes", "no"))
 
 

keras/layers/convolutional.py

@@ -195,7 +195,7 @@ def build(self, input_shape):
                                        self.filters)
 
     # compute_output_shape contains some validation logic for the input shape,
-    # and make sure the output shape has all positive dimentions.
+    # and make sure the output shape has all positive dimensions.
     self.compute_output_shape(input_shape)
 
     self.kernel = self.add_weight(

keras/layers/core/core_test.py

@@ -114,7 +114,7 @@ def test_dropout_with_savemodel(self):
     predict2 = loaded_model(np.ones((20, 5, 10)))
 
     self.assertAllClose(predict, predict2)
-    # Make sure the model droput different value after loading
+    # Make sure the model dropout different value after loading
     train2 = loaded_model(np.ones((20, 5, 10)), training=True)
     self.assertNotAllClose(train, train2)
     self.assertIsNotNone(loaded_model.layers[1]._random_generator)

keras/layers/core/lambda_layer.py

@@ -215,7 +215,7 @@ def _check_variables(self, created_variables, accessed_variables):
           but are not tracked by said layer:
           {variable_str}
           The layer cannot safely ensure proper Variable reuse across multiple
-          calls, and consquently this behavior is disallowed for safety. Lambda
+          calls, and consequently this behavior is disallowed for safety. Lambda
           layers are not well suited to stateful computation; instead, writing a
           subclassed Layer is the recommend way to define layers with
           Variables.""").format(

keras/layers/core/tf_op_layer.py

@@ -279,7 +279,7 @@ def _check_variables(self, created_variables, accessed_variables):
           'The following Variables were created within a Lambda layer '
           f'({self.name}) but are not tracked by said layer: {variable_str}\n'
           'The layer cannot safely ensure proper Variable reuse '
-          'across multiple calls, and consquently this behavior is disallowed '
+          'across multiple calls, and consequently this behavior is disallowed '
           'for safety reasons. Lambda layers are not well suited for stateful '
           'computation; instead, writing a subclassed Layer is the recommend '
           'way to define layers with Variables.')

keras/layers/normalization/layer_normalization_test.py

@@ -197,7 +197,7 @@ def _test_forward_pass(self, batch_input_shape, axis, fp64_tol=1e-14,
     Args:
       batch_input_shape: The input shape that will be used to test, including
         the batch dimension.
-      axis: A list of axises to normalize. Will be passed to the `axis` argument
+      axis: A list of axes to normalize. Will be passed to the `axis` argument
         of Layerlayer_normalization.
       fp64_tol: The relative and absolute tolerance for float64.
       fp32_tol: The relative and absolute tolerance for float32.
@@ -255,7 +255,7 @@ def _test_backward_pass(self, batch_input_shape, axis, fp64_tol=1e-5,
     Args:
       batch_input_shape: The input shape that will be used to test, including
         the batch dimension.
-      axis: A list of axises to normalize. Will be passed to the `axis` argument
+      axis: A list of axes to normalize. Will be passed to the `axis` argument
         of Layerlayer_normalization.
       fp64_tol: The relative and absolute tolerance for float64.
       fp32_tol: The relative and absolute tolerance for float32.

keras/layers/preprocessing/discretization.py

@@ -218,7 +218,7 @@ def __init__(self,
     if "dtype" not in kwargs or kwargs["dtype"] is None:
       kwargs["dtype"] = tf.int64 if output_mode == INT else backend.floatx()
     elif output_mode == "int" and not tf.as_dtype(kwargs["dtype"]).is_integer:
-      # Compat for when dtype was alwyas floating and ingored by the layer.
+      # Compat for when dtype was always floating and ignored by the layer.
       kwargs["dtype"] = tf.int64
 
     super().__init__(**kwargs)

keras/layers/preprocessing/hashing.py

@@ -168,7 +168,7 @@ def __init__(self,
     if 'dtype' not in kwargs or kwargs['dtype'] is None:
       kwargs['dtype'] = tf.int64 if output_mode == INT else backend.floatx()
     elif output_mode == 'int' and not tf.as_dtype(kwargs['dtype']).is_integer:
-      # Compat for when dtype was alwyas floating and ingored by the layer.
+      # Compat for when dtype was always floating and ignored by the layer.
       kwargs['dtype'] = tf.int64
 
     super().__init__(**kwargs)

keras/layers/preprocessing/index_lookup.py

@@ -268,7 +268,7 @@ def __init__(self,
         # value of the index_lookup table.
         self._default_value = self._oov_start_index()
       else:
-        # If we hav multiple OOV values, we need to do a further hashing step;
+        # If we have multiple OOV values, we need to do a further hashing step;
         # to make this easier, we set the OOV value to -1. (This lets us do a
         # vectorized add and cast to boolean to determine locations where we
         # need to do extra hashing.)
@@ -293,7 +293,7 @@ def __init__(self,
       # to be uninitialized as a StaticHashTable cannot be initialized twice.
       self.lookup_table = self._uninitialized_lookup_table()
 
-    # Only set up adapt state if we did not recieve a vocab on construction.
+    # Only set up adapt state if we did not receive a vocab on construction.
     if not self._has_input_vocabulary:
       # Add a custom weight handler to return the layers vocab as it's weight.
       self._add_trackable(VocabWeightHandler(self), False)
@@ -648,8 +648,9 @@ def _lookup_dense(self, inputs):
     """Lookup table values for a dense Tensor, handling masking and OOV."""
     # When executing eagerly and tracing keras.Inputs, do not call lookup. This
     # is critical for restoring SavedModel, which will first trace layer.call
-    # and then attempt to restore the table. We need the table to be unitialized
-    # for the restore to work, but calling the table unitialized would error.
+    # and then attempt to restore the table. We need the table to be
+    # uninitialized for the restore to work, but calling the table uninitialized
+    # would error.
     if tf.executing_eagerly() and backend.is_keras_tensor(inputs):
       lookups = tf.zeros_like(inputs, dtype=self._value_dtype)
     else:

keras/layers/preprocessing/preprocessing_stage.py

@@ -123,7 +123,7 @@ class FunctionalPreprocessingStage(functional.Functional,
 
   Args:
     inputs: An input tensor (must be created via `tf.keras.Input()`), or a list,
-      a dict, or a nested strcture of input tensors.
+      a dict, or a nested structure of input tensors.
     outputs: An output tensor, or a list, a dict or a nested structure of output
       tensors.
     name: String, optional. Name of the preprocessing stage.