Patch/170506 misspell (#9725)

* Fix misspells.

* Fix misspells.

* Fix misspells.

CONTRIBUTING.md

@@ -40,7 +40,7 @@ TensorFlow coding style.
 * Include unit tests when you contribute new features, as they help to
   a) prove that your code works correctly, b) guard against future breaking
   changes to lower the maintenance cost.
-* Bug fixes also generally require unit tests, because the presense of bugs
+* Bug fixes also generally require unit tests, because the presence of bugs
   usually indicates insufficient test coverage.
 * Keep API compatibility in mind when you change code in core TensorFlow,
   e.g., code in [tensorflow/core](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/core) and  [tensorflow/python](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/python).

tensorflow/contrib/distributions/python/ops/operator_pd_identity.py

@@ -115,7 +115,7 @@ class OperatorPDIdentity(operator_pd.OperatorPDBase):
     """Static check that the argument `x` is proper `shape`, `dtype`."""
     # x is a typical argument e.g. to matmul or solve.  In both cases, x should
     # have the same type/shape since this is a square matrix.  These checks are
-    # ususally not needed since we ususally have some tensor backing this
+    # usually not needed since we usually have some tensor backing this
     # distribution, and the calls to tf.matmul do a shape/type check.
     #
     # Static checks only for efficiency, the identity should be fast.

tensorflow/contrib/keras/python/keras/datasets/imdb.py

@@ -41,7 +41,7 @@ def load_data(path='imdb.npz',
       num_words: max number of words to include. Words are ranked
           by how often they occur (in the training set) and only
           the most frequent words are kept
-      skip_top: skip the top N most frequently occuring words
+      skip_top: skip the top N most frequently occurring words
           (which may not be informative).
       maxlen: truncate sequences after this length.
       seed: random seed for sample shuffling.

tensorflow/contrib/keras/python/keras/datasets/reuters.py

@@ -43,7 +43,7 @@ def load_data(path='reuters.npz',
       num_words: max number of words to include. Words are ranked
           by how often they occur (in the training set) and only
           the most frequent words are kept
-      skip_top: skip the top N most frequently occuring words
+      skip_top: skip the top N most frequently occurring words
           (which may not be informative).
       maxlen: truncate sequences after this length.
       test_split: Fraction of the dataset to be used as test data.

tensorflow/contrib/keras/python/keras/engine/topology.py

@@ -649,7 +649,7 @@ class Layer(tf_base_layers.Layer):
                           'but was passed an input_mask: ' + str(mask))
       # masking not explicitly supported: return None as mask
       return None
-    # if masking is explictly supported, by default
+    # if masking is explicitly supported, by default
     # carry over the input mask
     return mask
 

tensorflow/contrib/keras/python/keras/engine/training.py

@@ -245,7 +245,7 @@ def _check_array_lengths(inputs, targets, weights):
 
 
 def _check_loss_and_target_compatibility(targets, loss_fns, output_shapes):
-  """Does validation on the compatiblity of targets and loss functions.
+  """Does validation on the compatibility of targets and loss functions.
 
   This helps prevent users from using loss functions incorrectly.
 

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

@@ -169,7 +169,7 @@ class BatchNormalization(Layer):
 
       def normalize_inference():
         if needs_broadcasting:
-          # In this case we must explictly broadcast all parameters.
+          # In this case we must explicitly broadcast all parameters.
           broadcast_moving_mean = K.reshape(self.moving_mean, broadcast_shape)
           broadcast_moving_variance = K.reshape(self.moving_variance,
                                                 broadcast_shape)

tensorflow/contrib/keras/python/keras/models.py

@@ -221,7 +221,7 @@ def load_model(filepath, custom_objects=None):
         obj: object, dict, or list.
 
     Returns:
-        The same structure, where occurences
+        The same structure, where occurrences
             of a custom object name have been replaced
             with the custom object.
     """

tensorflow/contrib/keras/python/keras/preprocessing/sequence.py

@@ -156,7 +156,7 @@ def skipgrams(sequence,
           of word indices (integers). If using a `sampling_table`,
           word indices are expected to match the rank
           of the words in a reference dataset (e.g. 10 would encode
-          the 10-th most frequently occuring token).
+          the 10-th most frequently occurring token).
           Note that index 0 is expected to be a non-word and will be skipped.
       vocabulary_size: int. maximum possible word index + 1
       window_size: int. actually half-window.

tensorflow/contrib/labeled_tensor/python/ops/core.py

@@ -810,7 +810,7 @@ def axis_order_scope(axis_order=None):
   Example usage:
 
     with lt.axis_order_scope(['x', 'y', 'z']):
-      # result is guranteed to have the correct axis order
+      # result is guaranteed to have the correct axis order
       result = w + b
 
   You can nest scopes, in which case only the inner-most scope applies, e.g.,

tensorflow/contrib/layers/python/layers/feature_column.py

@@ -451,7 +451,7 @@ class _SparseColumn(
     return input_tensor
 
   def is_compatible(self, other_column):
-    """Check compatability of two sparse columns."""
+    """Check compatibility of two sparse columns."""
     if self.lookup_config and other_column.lookup_config:
       return self.lookup_config == other_column.lookup_config
     compatible = (self.length == other_column.length and
@@ -2069,7 +2069,7 @@ class _CrossedColumn(_FeatureColumn,
                                              "hash_key",
                                              "combiner", "ckpt_to_load_from",
                                              "tensor_name_in_ckpt"])):
-  """Represents a cross transformation also known as conjuction or combination.
+  """Represents a cross transformation also known as conjunction or combination.
 
   Instances of this class are immutable. It crosses given `columns`. Crossed
   column output will be hashed to hash_bucket_size.

tensorflow/contrib/layers/python/layers/feature_column_ops_test.py

@@ -1467,8 +1467,8 @@ class SequenceInputFromFeatureColumnTest(test.TestCase):
     expected_input_shape = [4, 3, embedding_dimension]
     self.assertAllEqual(expected_input_shape, model_input.shape)
 
-    # `ids_tensor` consists of 7 instances of <empty>, 3 occurences of "b",
-    # 2 occurences of "c" and 1 instance of "a".
+    # `ids_tensor` consists of 7 instances of <empty>, 3 occurrences of "b",
+    # 2 occurrences of "c" and 1 instance of "a".
     expected_gradient_values = sorted([0., 3., 2., 1.] * embedding_dimension)
     actual_gradient_values = np.sort(gradients[0].values, axis=None)
     self.assertAllClose(expected_gradient_values, actual_gradient_values)

tensorflow/contrib/layers/python/layers/feature_column_test.py

@@ -173,7 +173,7 @@ class FeatureColumnTest(test.TestCase):
     for i in range(len(b1_value)):
       self.assertAllClose(b1_value[i], b2_value[i])
 
-    # Test the case when a shared_embedding_name is explictly specified.
+    # Test the case when a shared_embedding_name is explicitly specified.
     d = fc.shared_embedding_columns(
         [a1, a2],
         dimension=4,

tensorflow/contrib/layers/python/layers/layers.py

@@ -278,7 +278,7 @@ def _fused_batch_norm(
         trainable=trainable_gamma)
 
     # Create moving_mean and moving_variance variables and add them to the
-    # appropiate collections.
+    # appropriate collections.
     moving_mean_collections = utils.get_variable_collections(
         variables_collections, 'moving_mean')
     moving_mean_initializer = param_initializers.get(
@@ -632,7 +632,7 @@ def batch_norm(inputs,
                                        trainable=trainable)
 
     # Create moving_mean and moving_variance variables and add them to the
-    # appropiate collections. We disable variable partitioning while creating
+    # appropriate collections. We disable variable partitioning while creating
     # them, because assign_moving_average is not yet supported for partitioned
     # variables.
     partitioner = variable_scope.get_variable_scope().partitioner

tensorflow/contrib/learn/python/learn/estimators/dynamic_rnn_estimator.py

@@ -115,7 +115,7 @@ def dict_to_state_tuple(input_dict, cell):
 
 
 def _concatenate_context_input(sequence_input, context_input):
-  """Replicates `context_input` accross all timesteps of `sequence_input`.
+  """Replicates `context_input` across all timesteps of `sequence_input`.
 
   Expands dimension 1 of `context_input` then tiles it `sequence_length` times.
   This value is appended to `sequence_input` on dimension 2 and the result is
@@ -177,7 +177,7 @@ def build_sequence_input(features,
       describing sequence features. All items in the set should be instances
       of classes derived from `FeatureColumn`.
     context_feature_columns: An iterable containing all the feature columns
-      describing context features i.e. features that apply accross all time
+      describing context features i.e. features that apply across all time
       steps. All items in the set should be instances of classes derived from
       `FeatureColumn`.
     weight_collections: List of graph collections to which weights are added.
@@ -419,7 +419,7 @@ def _get_dynamic_rnn_model_fn(
       describing sequence features. All items in the set should be instances
       of classes derived from `FeatureColumn`.
     context_feature_columns: An iterable containing all the feature columns
-      describing context features, i.e., features that apply accross all time
+      describing context features, i.e., features that apply across all time
       steps. All items in the set should be instances of classes derived from
       `FeatureColumn`.
     predict_probabilities: A boolean indicating whether to predict probabilities
@@ -603,7 +603,7 @@ class DynamicRnnEstimator(estimator.Estimator):
         describing sequence features. All items in the iterable should be
         instances of classes derived from `FeatureColumn`.
       context_feature_columns: An iterable containing all the feature columns
-        describing context features, i.e., features that apply accross all time
+        describing context features, i.e., features that apply across all time
         steps. All items in the set should be instances of classes derived from
         `FeatureColumn`.
       num_classes: the number of classes for a classification problem. Only

tensorflow/contrib/learn/python/learn/estimators/head.py

@@ -163,7 +163,7 @@ class Head(object):
           ModeFnOps.loss to compute and apply gradients.
       logits: logits `Tensor` to be used by the head.
       logits_input: `Tensor` from which to build logits, often needed when you
-        don't want to compute the logits. Typicaly this is the activation of the
+        don't want to compute the logits. Typically this is the activation of the
         last hidden layer in a DNN. Some heads (like the ones responsible for
         candidate sampling) intrinsically avoid computing full logits and only
         accepts logits_input.

tensorflow/contrib/learn/python/learn/estimators/head_test.py

@@ -895,7 +895,7 @@ class BinaryClassificationHeadTest(test.TestCase):
       _assert_summary_tags(self, ["loss"])
       # logloss: z:label, x:logit
       # z * -log(sigmoid(x)) + (1 - z) * -log(1 - sigmoid(x))
-      # expected_loss is (total_weighted_loss)/1 since htere is 1 nonzero
+      # expected_loss is (total_weighted_loss)/1 since there is 1 nonzero
       # weight.
       expected_loss = 0.062652342
       _assert_metrics(

tensorflow/contrib/learn/python/learn/estimators/state_saving_rnn_estimator.py

@@ -144,7 +144,7 @@ def _prepare_features_for_sqss(features, labels, mode,
       describing sequence features. All items in the set should be instances
       of classes derived from `FeatureColumn`.
     context_feature_columns: An iterable containing all the feature columns
-      describing context features, i.e., features that apply accross all time
+      describing context features, i.e., features that apply across all time
       steps. All items in the set should be instances of classes derived from
       `FeatureColumn`.
 
@@ -261,7 +261,7 @@ def _read_batch(cell,
       describing sequence features. All items in the set should be instances
       of classes derived from `FeatureColumn`.
     context_feature_columns: An iterable containing all the feature columns
-      describing context features, i.e., features that apply accross all time
+      describing context features, i.e., features that apply across all time
       steps. All items in the set should be instances of classes derived from
       `FeatureColumn`.
     num_threads: The Python integer number of threads enqueuing input examples
@@ -420,7 +420,7 @@ def _get_rnn_model_fn(cell_type,
       describing sequence features. All items in the set should be instances
       of classes derived from `FeatureColumn`.
     context_feature_columns: An iterable containing all the feature columns
-      describing context features, i.e., features that apply accross all time
+      describing context features, i.e., features that apply across all time
       steps. All items in the set should be instances of classes derived from
       `FeatureColumn`.
     predict_probabilities: A boolean indicating whether to predict probabilities
@@ -563,7 +563,7 @@ class StateSavingRnnEstimator(estimator.Estimator):
         describing sequence features. All items in the set should be instances
         of classes derived from `FeatureColumn`.
       context_feature_columns: An iterable containing all the feature columns
-        describing context features, i.e., features that apply accross all time
+        describing context features, i.e., features that apply across all time
         steps. All items in the set should be instances of classes derived from
         `FeatureColumn`.
       num_classes: The number of classes for categorization. Used only and

tensorflow/contrib/learn/python/learn/monitors.py

@@ -473,7 +473,7 @@ class LoggingTrainable(EveryN):
 
   def every_n_step_begin(self, step):
     super(LoggingTrainable, self).every_n_step_begin(step)
-    # Get a list of trainable variables at the begining of every N steps.
+    # Get a list of trainable variables at the beginning of every N steps.
     # We cannot get this in __init__ because train_op has not been generated.
     trainables = ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES,
                                     scope=self._scope)

tensorflow/contrib/makefile/README.md

@@ -295,7 +295,7 @@ itself, you'll see it's broken up into host and target sections. If you are
 cross-compiling, you should look at customizing the target settings to match
 what you need for your desired system.
 
-## Dependency Managment
+## Dependency Management
 
 The Makefile loads in a list of dependencies stored in text files. These files
 are generated from the main Bazel build by running 

tensorflow/contrib/session_bundle/example/export_half_plus_two.py

@@ -97,7 +97,7 @@ def Export(export_dir, use_checkpoint_v2):
     }
 
     # Create two filename assets and corresponding tensors.
-    # TODO(b/26254158) Consider adding validation of file existance as well as
+    # TODO(b/26254158) Consider adding validation of file existence as well as
     # hashes (e.g. sha1) for consistency.
     original_filename1 = tf.constant("hello1.txt")
     tf.add_to_collection(tf.GraphKeys.ASSET_FILEPATHS, original_filename1)

tensorflow/contrib/slim/python/slim/learning_test.py

@@ -840,7 +840,7 @@ class TrainTest(test.TestCase):
         # Initialize the variables.
         sess.run(variables_lib.global_variables_initializer())
 
-        # Get the intial weights and biases values.
+        # Get the initial weights and biases values.
         weights_values, biases_values = sess.run([weights, biases])
         self.assertGreater(np.linalg.norm(weights_values), 0)
         self.assertAlmostEqual(np.linalg.norm(biases_values), 0)

tensorflow/contrib/sparsemax/python/kernel_tests/sparsemax_loss_test.py

@@ -159,7 +159,7 @@ class SparsemaxLossTest(test.TestCase):
     self.assertShapeEqual(q, tf_sparsemax_op)
 
   def _test_gradient_against_estimate(self, dtype, random, use_gpu):
-    """check sparsemax-loss Rop, aginst estimated-loss Rop"""
+    """check sparsemax-loss Rop, against estimated-loss Rop"""
     z = random.uniform(low=-3, high=3, size=(test_obs, 10)).astype(dtype)
     q = np.zeros((test_obs, 10)).astype(dtype)
     q[np.arange(0, test_obs), np.random.randint(0, 10, size=test_obs)] = 1
@@ -178,7 +178,7 @@ class SparsemaxLossTest(test.TestCase):
     self.assertLess(err, 1e-4)
 
   def _test_gradient_against_numpy(self, dtype, random, use_gpu):
-    """check sparsemax-loss Rop, aginst numpy Rop"""
+    """check sparsemax-loss Rop, against numpy Rop"""
     z = random.uniform(low=-3, high=3, size=(test_obs, 10))
     q = np.zeros((test_obs, 10))
     q[np.arange(0, test_obs), np.random.randint(0, 10, size=test_obs)] = 1

tensorflow/contrib/sparsemax/python/kernel_tests/sparsemax_test.py

@@ -188,7 +188,7 @@ class SparsemaxTest(test.TestCase):
     self.assertShapeEqual(z, tf_sparsemax_op)
 
   def _test_gradient_against_estimate(self, dtype, random, use_gpu):
-    """check sparsemax Rop, aginst estimated Rop"""
+    """check sparsemax Rop, against estimated Rop"""
     z = random.uniform(low=-3, high=3, size=(test_obs, 10)).astype(dtype)
 
     logits = array_ops.placeholder(dtype, name='z')
@@ -204,7 +204,7 @@ class SparsemaxTest(test.TestCase):
     self.assertLess(err, 1e-4)
 
   def _test_gradient_against_numpy(self, dtype, random, use_gpu):
-    """check sparsemax Rop, aginst numpy Rop"""
+    """check sparsemax Rop, against numpy Rop"""
     z = random.uniform(low=-3, high=3, size=(test_obs, 10)).astype(dtype)
 
     logits = constant_op.constant(z, name='z')

tensorflow/contrib/tensor_forest/client/random_forest.py

@@ -79,7 +79,7 @@ class TensorForestLossHook(session_run_hook.SessionRunHook):
     current_loss = run_values.results['current_loss']
     current_step = run_values.results['global_step']
     self.steps += 1
-    # Gaurd against the global step going backwards, which might happen
+    # Guard against the global step going backwards, which might happen
     # if we recover from something.
     if self.last_step == -1 or self.last_step > current_step:
       logging.info('TensorForestLossHook resetting last_step.')

tensorflow/contrib/training/python/training/evaluation.py

@@ -254,7 +254,7 @@ def checkpoints_iterator(checkpoint_dir,
         logging.info('Timed-out waiting for a checkpoint.')
         return
       if timeout_fn():
-        # The timeout_fn indicated that we are truely done.
+        # The timeout_fn indicated that we are truly done.
         return
       else:
         # The timeout_fn indicated that more checkpoints may come.

tensorflow/contrib/training/python/training/feeder.py

@@ -18,7 +18,7 @@
 
 This helper handles the plumbing in order to set up a feeder task to
 push generated inputs to a pool of remote consumers; or to run an
-identical feeding mechanism in a seperate thread in the same process.
+identical feeding mechanism in a separate thread in the same process.
 
 Example usage for distributed feeding:
 
@@ -331,7 +331,7 @@ class Feeder(object):
     they never close their queue. Second, they are added to the
     `Feeder.REMOTE_QUEUE_RUNNERS` collection, rather than
     `ops.GraphKeys.QUEUE_RUNNERS`, so they can be started/stopped
-    seperately.
+    separately.
 
     Args:
       queue: The queue.

tensorflow/contrib/training/python/training/feeder_test.py

@@ -156,7 +156,7 @@ class FeederTest(test.TestCase):
     coord.join()
 
   def testFeederSeparateThread(self):
-    # Start a feeder on a seperate thread, but with a shared local queue
+    # Start a feeder on a separate thread, but with a shared local queue
     servers = self._create_local_cluster(worker=1)
     coord = coordinator.Coordinator()
     feed_thread = FeederThread(self, coord, servers, 'worker', 0)

tensorflow/contrib/training/python/training/hparam.py

@@ -164,7 +164,7 @@ class HParams(object):
   import argparse
   parser = argparse.ArgumentParser(description='Train my model.')
   parser.add_argument('--hparams', type=str,
-                      help='Comma seperated list of "name=value" pairs.')
+                      help='Comma separated list of "name=value" pairs.')
   args = parser.parse_args()
   ...
   def my_program():

tensorflow/contrib/training/python/training/training_test.py

@@ -508,7 +508,7 @@ class TrainTest(test.TestCase):
         # Initialize the variables.
         session.run(variables_lib2.global_variables_initializer())
 
-        # Get the intial weights and biases values.
+        # Get the initial weights and biases values.
         weights_values, biases_values = session.run([weights, biases])
         self.assertGreater(np.linalg.norm(weights_values), 0)
         self.assertAlmostEqual(np.linalg.norm(biases_values), 0)

tensorflow/core/common_runtime/visitable_allocator.h

@@ -44,7 +44,7 @@ class VisitableAllocator : public Allocator {
 };
 
 // Needed for cases when a VisitableAllocator gets wrapped for tracking.
-// Multiple-inheritance is considered acceptible in this case because
+// Multiple-inheritance is considered acceptable in this case because
 // VisitableAllocator is a pure virtual interface and only TrackingAllocator
 // has default implementation.
 class TrackingVisitableAllocator : public TrackingAllocator,

tensorflow/docs_src/install/install_java.md

@@ -106,7 +106,7 @@ As an example, these steps will create a Maven project that uses TensorFlow:
 
 
 The preceding command should output <tt>Hello from <i>version</i></tt>. If it
-does, you've succesfully set up TensorFlow for Java and are ready to use it in
+does, you've successfully set up TensorFlow for Java and are ready to use it in
 Maven projects. If not, check
 [Stack Overflow](http://stackoverflow.com/questions/tagged/tensorflow)
 for possible solutions.  You can skip reading the rest of this document.

tensorflow/python/debug/cli/debugger_cli_common.py

@@ -648,7 +648,7 @@ class CommandHandlerRegistry(object):
         3) the handler is found for the prefix, but it fails to return a
           RichTextLines or raise any exception.
       CommandLineExit:
-        If the command handler raises this type of exception, tihs method will
+        If the command handler raises this type of exception, this method will
         simply pass it along.
     """
     if not prefix:

tensorflow/python/debug/cli/profile_analyzer_cli.py

@@ -55,7 +55,7 @@ class ProfileDatum(object):
 
   @property
   def exec_time(self):
-    """Measures compute function exection time plus pre- and post-processing."""
+    """Measures compute function execution time plus pre- and post-processing."""
     return self.node_exec_stats.all_end_rel_micros
 
 

tensorflow/python/debug/lib/debug_data.py

@@ -975,7 +975,7 @@ class DebugDumpDir(object):
       slot = datum.output_slot
       # In some cases (e.g., system clocks with insufficient precision),
       # the upstream and downstream tensors may have identical timestamps, the
-      # following check examines this possibilty and avoids raising an error if
+      # following check examines this possibility and avoids raising an error if
       # that is the case.
       if not self._satisfied_at_timestamp(
           pending_inputs[node], datum.timestamp, start_i=i + 1):

tensorflow/python/debug/wrappers/hooks.py

@@ -66,7 +66,7 @@ class LocalCLIDebugHook(session_run_hook.SessionRunHook,
     """Add a tensor filter.
 
     See doc of `LocalCLIDebugWrapperSession.add_tensor_filter()` for details.
-    Override default behavior to accomodate the possibility of this method being
+    Override default behavior to accommodate the possibility of this method being
     called prior to the initialization of the underlying
     `LocalCLIDebugWrapperSession` object.
 

tensorflow/python/framework/subscribe.py

@@ -276,7 +276,7 @@ def subscribe(tensors, side_effects):
     Subscribed tensors, which are identity copies of the passed in tensors
       in the same passed in structure, but the graph has been modified
       such that these are downstream of the control dependencies for
-      the side effect graphs. Use these functionally equivelant tensors
+      the side effect graphs. Use these functionally equivalent tensors
       instead of the passed in tensors for further construction or running.
   """
   if not hasattr(side_effects, '__iter__'):

tensorflow/python/framework/test_util.py

@@ -250,7 +250,7 @@ class TensorFlowTestCase(googletest.TestCase):
     """Returns a unique temporary directory for the test to use.
 
     If you call this method multiple times during in a test, it will return the
-    same folder. However, accross different runs the directories will be
+    same folder. However, across different runs the directories will be
     different. This will ensure that across different runs tests will not be
     able to pollute each others environment.
     If you need multiple unique directories within a single test, you should

tensorflow/python/kernel_tests/linalg_ops_test.py

@@ -28,7 +28,7 @@ from tensorflow.python.platform import test
 
 
 def _random_pd_matrix(n, rng):
-  """Random postive definite matrix."""
+  """Random positive definite matrix."""
   temp = rng.randn(n, n)
   return temp.dot(temp.T)
 

tensorflow/python/layers/normalization.py

@@ -322,7 +322,7 @@ class BatchNormalization(base.Layer):
 
     def _broadcast(v):
       if needs_broadcasting and v is not None:
-        # In this case we must explictly broadcast all parameters.
+        # In this case we must explicitly broadcast all parameters.
         return array_ops.reshape(v, broadcast_shape)
       return v
 

tensorflow/python/ops/math_ops.py

@@ -2331,7 +2331,7 @@ def tensordot(a, b, axes, name=None):
     using `array_ops.transpose` and `array_ops.reshape`. The method takes a
     tensor and performs the correct transpose and reshape operation for a given
     set of indices. It returns the reshaped tensor as well as a list of indices
-    necesary to reshape the tensor again after matrix multiplication.
+    necessary to reshape the tensor again after matrix multiplication.
 
     Args:
       a: `Tensor`.

tensorflow/python/ops/metrics_impl.py

@@ -1735,7 +1735,7 @@ def _streaming_sparse_true_positive_at_k(labels,
     A tuple of `Variable` and update `Operation`.
 
   Raises:
-    ValueError: If `weights` is not `None` and has an incomptable shape.
+    ValueError: If `weights` is not `None` and has an incompatible shape.
   """
   with ops.name_scope(
       name, _at_k_name('true_positive', k, class_id=class_id),
@@ -1831,7 +1831,7 @@ def _streaming_sparse_false_negative_at_k(labels,
     A tuple of `Variable` and update `Operation`.
 
   Raises:
-    ValueError: If `weights` is not `None` and has an incomptable shape.
+    ValueError: If `weights` is not `None` and has an incompatible shape.
   """
   with ops.name_scope(
       name, _at_k_name('false_negative', k, class_id=class_id),
@@ -2653,7 +2653,7 @@ def _streaming_sparse_false_positive_at_k(labels,
     A tuple of `Variable` and update `Operation`.
 
   Raises:
-    ValueError: If `weights` is not `None` and has an incomptable shape.
+    ValueError: If `weights` is not `None` and has an incompatible shape.
   """
   with ops.name_scope(
       name, _at_k_name('false_positive', k, class_id=class_id),

tensorflow/python/ops/parsing_ops.py

@@ -58,7 +58,7 @@ class SparseFeature(
         ["index_key", "value_key", "dtype", "size", "already_sorted"])):
   """Configuration for parsing a sparse input feature from an `Example`.
 
-  Note, preferrably use `VarLenFeature` (possibly in combination with a
+  Note, preferably use `VarLenFeature` (possibly in combination with a
   `SequenceExample`) in order to parse out `SparseTensor`s instead of
   `SparseFeature` due to its simplicity.
 

tensorflow/python/training/coordinator.py

@@ -366,7 +366,7 @@ class Coordinator(object):
     # If any thread is still alive, wait for the grace period to expire.
     # By the time this check is executed, threads may still be shutting down,
     # so we add a sleep of increasing duration to give them a chance to shut
-    # down without loosing too many cycles.
+    # down without losing too many cycles.
     # The sleep duration is limited to the remaining grace duration.
     stop_wait_secs = 0.001
     while any(t.is_alive() for t in threads) and stop_grace_period_secs >= 0.0:

tensorflow/python/util/deprecation.py

@@ -182,7 +182,7 @@ def deprecated_args(date, instructions, *deprecated_arg_names_or_tuples):
     return d
 
   def _get_deprecated_positional_arguments(names_to_ok_vals, arg_spec):
-    """Builds a dictionary from deprecated arguments to thier spec.
+    """Builds a dictionary from deprecated arguments to their spec.
 
     Returned dict is keyed by argument name.
     Each value is a DeprecatedArgSpec with the following fields:

tensorflow/stream_executor/lib/statusor.h

@@ -135,7 +135,7 @@ class StatusOr {
   // operators, to support move-only types and avoid unnecessary copying.
   StatusOr(T&& value);  // NOLINT
 
-  // Move conversion operator to avoid unecessary copy.
+  // Move conversion operator to avoid unnecessary copy.
   // T must be assignable from U.
   // Not marked with explicit so the implicit conversion can happen.
   template <typename U>

tensorflow/tools/docs/py_guide_parser.py

@@ -34,7 +34,7 @@ def md_files_in_dir(py_guide_src_dir):
 class PyGuideParser(object):
   """Simple parsing of a guide .md file.
 
-  Decendents can override the process_*() functions (called by process())
+  Descendants can override the process_*() functions (called by process())
   to either record infromation from the guide, or call replace_line()
   to affect the return value of process().
   """