contrib: minor spelling tweaks (#17788)

packages:
  model_pruning
  rnn
  solvers
  tensorrt

12 files changed, 16 insertions, 16 deletions

diff --git a/tensorflow/contrib/model_pruning/python/layers/layers.py b/tensorflow/contrib/model_pruning/python/layers/layers.py
index 988748ad75..466daf204a 100644
--- a/tensorflow/contrib/model_pruning/python/layers/layers.py
+++ b/tensorflow/contrib/model_pruning/python/layers/layers.py
@@ -214,7 +214,7 @@ def masked_convolution(inputs,
     elif data_format == 'NCHW':
       df = 'channels_first'
     else:
-      raise ValueError('Unsupported data fromat', data_format)
+      raise ValueError('Unsupported data format', data_format)
 
     layer = layer_class(
         filters=num_outputs,
diff --git a/tensorflow/contrib/model_pruning/python/pruning.py b/tensorflow/contrib/model_pruning/python/pruning.py
index 86963be4b8..5146a4a2de 100644
--- a/tensorflow/contrib/model_pruning/python/pruning.py
+++ b/tensorflow/contrib/model_pruning/python/pruning.py
@@ -216,7 +216,7 @@ def _partitioned_variable_assign(partitioned_var, new_value):
   """Assign op for partitioned variables.
 
   Args:
-    partitioned_var: A partitioned tensotflow variable
+    partitioned_var: A partitioned tensorflow variable
     new_value: Value to be assigned to the variable var
 
   Returns:
diff --git a/tensorflow/contrib/rnn/ops/gru_ops.cc b/tensorflow/contrib/rnn/ops/gru_ops.cc
index e91d1e8a80..9c8e40851a 100644
--- a/tensorflow/contrib/rnn/ops/gru_ops.cc
+++ b/tensorflow/contrib/rnn/ops/gru_ops.cc
@@ -69,7 +69,7 @@ Element-wise dot product of a and b is represented by ab
 Element-wise dot product is represented by \circ
 Matrix multiplication is represented by *
 
-Baises are initialized with :
+Biases are initialized with :
 `b_ru` - constant_initializer(1.0)
 `b_c` - constant_initializer(0.0)
 
diff --git a/tensorflow/contrib/rnn/python/kernel_tests/lstm_ops_test.py b/tensorflow/contrib/rnn/python/kernel_tests/lstm_ops_test.py
index 7957edf68c..ffd2421894 100644
--- a/tensorflow/contrib/rnn/python/kernel_tests/lstm_ops_test.py
+++ b/tensorflow/contrib/rnn/python/kernel_tests/lstm_ops_test.py
@@ -54,7 +54,7 @@ def blocks_match(sess, use_peephole):
   initializer = init_ops.random_uniform_initializer(-0.01, 0.01, seed=19890212)
 
   with variable_scope.variable_scope("test", initializer=initializer):
-    # magic naming so that the cells pick up these variables and resuse them
+    # magic naming so that the cells pick up these variables and reuse them
     if use_peephole:
       wci = variable_scope.get_variable(
           "rnn/lstm_cell/w_i_diag", shape=[cell_size], dtype=dtypes.float32)
diff --git a/tensorflow/contrib/rnn/python/ops/rnn_cell.py b/tensorflow/contrib/rnn/python/ops/rnn_cell.py
index 358b2eb02b..2f6ae9f367 100644
--- a/tensorflow/contrib/rnn/python/ops/rnn_cell.py
+++ b/tensorflow/contrib/rnn/python/ops/rnn_cell.py
@@ -534,7 +534,7 @@ class GridLSTMCell(rnn_cell_impl.RNNCell):
       initializer: (optional) The initializer to use for the weight and
         projection matrices, default None.
       num_unit_shards: (optional) int, default 1, How to split the weight
-        matrix. If > 1,the weight matrix is stored across num_unit_shards.
+        matrix. If > 1, the weight matrix is stored across num_unit_shards.
       forget_bias: (optional) float, default 1.0, The initial bias of the
         forget gates, used to reduce the scale of forgetting at the beginning
         of the training.
@@ -993,7 +993,7 @@ class BidirectionalGridLSTMCell(GridLSTMCell):
       initializer: (optional) The initializer to use for the weight and
         projection matrices, default None.
       num_unit_shards: (optional) int, default 1, How to split the weight
-        matrix. If > 1,the weight matrix is stored across num_unit_shards.
+        matrix. If > 1, the weight matrix is stored across num_unit_shards.
       forget_bias: (optional) float, default 1.0, The initial bias of the
         forget gates, used to reduce the scale of forgetting at the beginning
         of the training.
diff --git a/tensorflow/contrib/solvers/python/ops/least_squares.py b/tensorflow/contrib/solvers/python/ops/least_squares.py
index fb7c0eb649..6e164f5342 100644
--- a/tensorflow/contrib/solvers/python/ops/least_squares.py
+++ b/tensorflow/contrib/solvers/python/ops/least_squares.py
@@ -33,7 +33,7 @@ def cgls(operator, rhs, tol=1e-6, max_iter=20, name="cgls"):
   r"""Conjugate gradient least squares solver.
 
   Solves a linear least squares problem \\(||A x - rhs||_2\\) for a single
-  righ-hand side, using an iterative, matrix-free algorithm where the action of
+  right-hand side, using an iterative, matrix-free algorithm where the action of
   the matrix A is represented by `operator`. The CGLS algorithm implicitly
   applies the symmetric conjugate gradient algorithm to the normal equations
   \\(A^* A x = A^* rhs\\). The iteration terminates when either
diff --git a/tensorflow/contrib/solvers/python/ops/linear_equations.py b/tensorflow/contrib/solvers/python/ops/linear_equations.py
index d791d46763..9305c6a11c 100644
--- a/tensorflow/contrib/solvers/python/ops/linear_equations.py
+++ b/tensorflow/contrib/solvers/python/ops/linear_equations.py
@@ -41,7 +41,7 @@ def conjugate_gradient(operator,
   r"""Conjugate gradient solver.
 
   Solves a linear system of equations `A*x = rhs` for selfadjoint, positive
-  definite matrix `A` and righ-hand side vector `rhs`, using an iterative,
+  definite matrix `A` and right-hand side vector `rhs`, using an iterative,
   matrix-free algorithm where the action of the matrix A is represented by
   `operator`. The iteration terminates when either the number of iterations
   exceeds `max_iter` or when the residual norm has been reduced to `tol`
diff --git a/tensorflow/contrib/tensorrt/convert/convert_graph.h b/tensorflow/contrib/tensorrt/convert/convert_graph.h
index e1596e89e2..e01e4a5328 100644
--- a/tensorflow/contrib/tensorrt/convert/convert_graph.h
+++ b/tensorflow/contrib/tensorrt/convert/convert_graph.h
@@ -35,7 +35,7 @@ tensorflow::Status ConvertCalibGraphToInferGraph(
 
 // max_batch_size: maximum batch size which can be used for inference for
 //                 optimization targets inference run with max batch size.
-// max_workspace_size_bytes: The upper bound of memory allowence for
+// max_workspace_size_bytes: The upper bound of memory allowance for
 //                 engine building.
 tensorflow::Status ConvertGraphDefToTensorRT(
     const tensorflow::GraphDef& graph_def,
diff --git a/tensorflow/contrib/tensorrt/convert/convert_nodes.cc b/tensorflow/contrib/tensorrt/convert/convert_nodes.cc
index 75a3c3d034..92a692baa7 100644
--- a/tensorflow/contrib/tensorrt/convert/convert_nodes.cc
+++ b/tensorflow/contrib/tensorrt/convert/convert_nodes.cc
@@ -455,7 +455,7 @@ class Converter {
       if (trt_tensors_.count(name)) {
         inputs.push_back(trt_tensors_.at(name));
       } else {
-        LOG(FATAL) << "input: " << name << " not availabled for node at, "
+        LOG(FATAL) << "input: " << name << " not available for node at, "
                    << node_def.name();
       }
     }
@@ -884,7 +884,7 @@ tensorflow::Status BinaryTensorOpWeight(
   // default to element-wise
   auto scale_mode = nvinfer1::ScaleMode::kELEMENTWISE;
 
-  // TODO(jie): maybe use a permuatation instead to support more cases;
+  // TODO(jie): maybe use a permutation instead to support more cases;
   bool permutation_flag = false;
 
   if (weights.count() == 1) {
@@ -1498,7 +1498,7 @@ tensorflow::Status ConvertConst(Converter& ctx,
           weights_tensor.int_val().begin(),
           weights_tensor.int_val()
               .end());  //  make a local copy first to flatten
-                        //  doesn't have to be contigous
+                        //  doesn't have to be contiguous
       memcpy(dst, tensor_data.data(), len_tensor);  // store into weight store
       weights = TRT_ShapedWeights(dtype, dst, scalar_shape);
     }
@@ -2212,7 +2212,7 @@ tensorflow::Status InjectCalibrationNode(tensorrt::convert::SubGraphParams& s) {
   std::list<tensorflow::Node*> order;
   for (tensorflow::Node* node : order_vec) {
     if (s.subgraph_node_ids.count(node->id())) {
-      order.push_front(node);  // we want topological order to contstruct the
+      order.push_front(node);  // we want topological order to construct the
       // network layer by layer
     }
   }
diff --git a/tensorflow/contrib/tensorrt/python/trt_convert.py b/tensorflow/contrib/tensorrt/python/trt_convert.py
index 666220d78c..338475d90e 100644
--- a/tensorflow/contrib/tensorrt/python/trt_convert.py
+++ b/tensorflow/contrib/tensorrt/python/trt_convert.py
@@ -41,7 +41,7 @@ def create_inference_graph(input_graph_def,
                            max_workspace_size_bytes=2 << 20,
                            precision_mode="FP32",
                            minimum_segment_size=3):
-  """Python wrapper for the TRT transormation.
+  """Python wrapper for the TRT transformation.
 
   Args:
     input_graph_def: GraphDef object containing a model to be transformed.
diff --git a/tensorflow/contrib/tensorrt/resources/trt_int8_calibrator.cc b/tensorflow/contrib/tensorrt/resources/trt_int8_calibrator.cc
index 74df75902e..dc7c93f869 100644
--- a/tensorflow/contrib/tensorrt/resources/trt_int8_calibrator.cc
+++ b/tensorflow/contrib/tensorrt/resources/trt_int8_calibrator.cc
@@ -61,7 +61,7 @@ bool TRTInt8Calibrator::setBatch(const std::unordered_map<string, void*>& data,
 
     // TODO(aaroey): we should not use sync copy on default stream. Make sure
     // stream->ThenMemcpy() is used in future PRs.
-    // TODO(sami,aaroey): Need to figureout a way to ensure synchronization
+    // TODO(sami,aaroey): Need to figure out a way to ensure synchronization
     // between stream, perhaps using a tensor?
     auto status = cudaMemcpyAsync(d.first, it.second, d.second,
                                   cudaMemcpyDeviceToDevice, stream);
diff --git a/tensorflow/contrib/tensorrt/test/test_tftrt.py b/tensorflow/contrib/tensorrt/test/test_tftrt.py
index 0b661bd536..ad01bedd8f 100644
--- a/tensorflow/contrib/tensorrt/test/test_tftrt.py
+++ b/tensorflow/contrib/tensorrt/test/test_tftrt.py
@@ -75,7 +75,7 @@ def run_graph(gdef, dumm_inp):
   return val
 
 
-# Use real data that is representatitive of the inference dataset
+# Use real data that is representative of the inference dataset
 # for calibration. For this test script it is random data.
 def run_calibration(gdef, dumm_inp):
   """Run given calibration graph multiple times."""