Distribute sp (tensorflow#18138)

brettkoonce · caisq · commit 10878ede0688 · 2018-03-30T21:51:21.000-07:00
* contrib/distribute: minor spelling tweaks

* contrib/distributions: minor spelling tweaks
diff --git a/tensorflow/contrib/distribute/python/cross_tower_ops.py b/tensorflow/contrib/distribute/python/cross_tower_ops.py
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-"""Classes for different algortihms of reduction and broadcasting."""
+"""Classes for different algorithms of reduction and broadcasting."""
 
 from __future__ import absolute_import
 from __future__ import division
@@ -155,7 +155,7 @@ def batch_reduce(self, method_string, value_destination_pairs):
     Args:
       method_string: either 'sum' or 'mean' specifying the reduction method.
       value_destination_pairs: a list or a tuple of tuples of PerDevice objects
-        and destinations. If a destionation is None, then the destinations
+        and destinations. If a destination is None, then the destinations
         are set to match the devices of the input PerDevice object.
 
     Returns:
diff --git a/tensorflow/contrib/distribute/python/cross_tower_utils.py b/tensorflow/contrib/distribute/python/cross_tower_utils.py
@@ -316,7 +316,7 @@ def unpack_small_tensors(tower_grads, packing):
       it made to tower_grads.
 
   Returns:
-    new_tower_grads: identical to tower_grads except that concatentations
+    new_tower_grads: identical to tower_grads except that concatenations
       of small tensors have been split apart and returned to their original
       positions, paired with their original variables.
   """
diff --git a/tensorflow/contrib/distribute/python/shared_variable_creator.py b/tensorflow/contrib/distribute/python/shared_variable_creator.py
@@ -46,7 +46,7 @@ def make_fn(shared_variable_store, device_id):
   error.
   Additionally, we de-uniquify variable names before checking for matches. This
   helps re-use variables which are intended to be the same but have different
-  names due to variable uniquificaton happening upstream. Since this might
+  names due to variable uniquification happening upstream. Since this might
   mean we may have multiple variables with the same canonical name, we store
   them in a list per canonical name and return them in the same order as well.
 
diff --git a/tensorflow/contrib/distributions/python/kernel_tests/bijectors/kumaraswamy_bijector_test.py b/tensorflow/contrib/distributions/python/kernel_tests/bijectors/kumaraswamy_bijector_test.py
@@ -69,7 +69,7 @@ def testBijectiveAndFinite(self):
       bijector = Kumaraswamy(
           concentration1=concentration1,
           concentration0=concentration0, validate_args=True)
-      # Omitting the endpoints 0 and 1, since idlj will be inifinity at these
+      # Omitting the endpoints 0 and 1, since idlj will be infinity at these
       # endpoints.
       y = np.linspace(.01, 0.99, num=10).astype(np.float32)
       x = 1 - (1 - y ** concentration1) ** concentration0
diff --git a/tensorflow/contrib/distributions/python/ops/estimator.py b/tensorflow/contrib/distributions/python/ops/estimator.py
@@ -75,7 +75,7 @@ def estimator_head_distribution_regression(make_distribution_fn,
 
 
 class _DistributionRegressionHead(_RegressionHead):
-  """Creates a _RegressionHead instance from an arbitray `Distribution`."""
+  """Creates a _RegressionHead instance from an arbitrary `Distribution`."""
 
   def __init__(self,
                make_distribution_fn,
diff --git a/tensorflow/contrib/distributions/python/ops/independent.py b/tensorflow/contrib/distributions/python/ops/independent.py
@@ -36,7 +36,7 @@ class Independent(distribution_lib.Distribution):
 
   This distribution is useful for regarding a collection of independent,
   non-identical distributions as a single random variable. For example, the
-  `Indpendent` distribution composed of a collection of `Bernoulli`
+  `Independent` distribution composed of a collection of `Bernoulli`
   distributions might define a distribution over an image (where each
   `Bernoulli` is a distribution over each pixel).
 
diff --git a/tensorflow/contrib/distributions/python/ops/onehot_categorical.py b/tensorflow/contrib/distributions/python/ops/onehot_categorical.py
@@ -52,15 +52,15 @@ class OneHotCategorical(distribution.Distribution):
 
   #### Examples
 
-  Creates a 3-class distiribution, with the 2nd class, the most likely to be
+  Creates a 3-class distribution, with the 2nd class, the most likely to be
   drawn from.
 
   ```python
   p = [0.1, 0.5, 0.4]
   dist = OneHotCategorical(probs=p)
   ```
 
-  Creates a 3-class distiribution, with the 2nd class the most likely to be
+  Creates a 3-class distribution, with the 2nd class the most likely to be
   drawn from, using logits.
 
   ```python
diff --git a/tensorflow/contrib/distributions/python/ops/relaxed_bernoulli.py b/tensorflow/contrib/distributions/python/ops/relaxed_bernoulli.py
@@ -35,10 +35,10 @@ class RelaxedBernoulli(transformed_distribution.TransformedDistribution):
 
   The RelaxedBernoulli is a distribution over the unit interval (0,1), which
   continuously approximates a Bernoulli. The degree of approximation is
-  controlled by a temperature: as the temperaturegoes to 0 the RelaxedBernoulli
-  becomes discrete with a distribution described by the `logits` or `probs`
-  parameters, as the temperature goes to infinity the RelaxedBernoulli
-  becomes the constant distribution that is identically 0.5.
+  controlled by a temperature: as the temperature goes to 0 the
+  RelaxedBernoulli becomes discrete with a distribution described by the
+  `logits` or `probs` parameters, as the temperature goes to infinity the
+  RelaxedBernoulli becomes the constant distribution that is identically 0.5.
 
   The RelaxedBernoulli distribution is a reparameterized continuous
   distribution that is the binary special case of the RelaxedOneHotCategorical
diff --git a/tensorflow/contrib/distributions/python/ops/relaxed_onehot_categorical.py b/tensorflow/contrib/distributions/python/ops/relaxed_onehot_categorical.py
@@ -303,7 +303,7 @@ class RelaxedOneHotCategorical(
   The RelaxedOneHotCategorical is a distribution over random probability
   vectors, vectors of positive real values that sum to one, which continuously
   approximates a OneHotCategorical. The degree of approximation is controlled by
-  a temperature: as the temperaturegoes to 0 the RelaxedOneHotCategorical
+  a temperature: as the temperature goes to 0 the RelaxedOneHotCategorical
   becomes discrete with a distribution described by the `logits` or `probs`
   parameters, as the temperature goes to infinity the RelaxedOneHotCategorical
   becomes the constant distribution that is identically the constant vector of
diff --git a/tensorflow/contrib/distributions/python/ops/vector_student_t.py b/tensorflow/contrib/distributions/python/ops/vector_student_t.py
@@ -66,7 +66,7 @@ class _VectorStudentT(transformed_distribution.TransformedDistribution):
   This distribution is an Affine transformation of iid
   [Student's t-distributions](
   https://en.wikipedia.org/wiki/Student%27s_t-distribution)
-  and should not be confused with the [Multivate Student's t-distribution](
+  and should not be confused with the [Multivariate Student's t-distribution](
   https://en.wikipedia.org/wiki/Multivariate_t-distribution). The
   traditional Multivariate Student's t-distribution is type of
   [elliptical distribution](
