Various spelling and grammar fixes (#7793)

* Various spelling and grammar fixes

* changed Nvidia to NVIDIA as per review

* spelling fix

docker/README.md

@@ -16,7 +16,7 @@ quick links here:
 
 We are using `Makefile` to simplify docker commands within make commands.
 
-Build the container and start a jupyter notebook
+Build the container and start a Jupyter Notebook
 
     $ make notebook
 
@@ -28,7 +28,7 @@ Build the container and start a bash
 
     $ make bash
 
-For GPU support install NVidia drivers (ideally latest) and
+For GPU support install NVIDIA drivers (ideally latest) and
 [nvidia-docker](https://github.com/NVIDIA/nvidia-docker). Run using
 
     $ make notebook GPU=0 # or [ipython, bash]

docs/templates/regularizers.md

@@ -39,7 +39,7 @@ def l1_reg(weight_matrix):
     return 0.01 * K.sum(K.abs(weight_matrix))
 
 model.add(Dense(64, input_dim=64,
-                kernel_regularizer=l1_reg)
+                kernel_regularizer=l1_reg))
 ```
 
 Alternatively, you can write your regularizers in an object-oriented way;

examples/README.md

@@ -40,7 +40,7 @@ Trains a convolutional stack followed by a recurrent stack network on the IMDB s
 Trains a FastText model on the IMDB sentiment classification task.
 
 [imdb_lstm.py](imdb_lstm.py)
-Trains a LSTM on the IMDB sentiment classification task.
+Trains an LSTM model on the IMDB sentiment classification task.
 
 [lstm_benchmark.py](lstm_benchmark.py)
 Compares different LSTM implementations on the IMDB sentiment classification task.

examples/conv_lstm.py

@@ -125,7 +125,7 @@ def generate_movies(n_samples=1200, n_frames=15):
     if i >= 7:
         ax.text(1, 3, 'Predictions !', fontsize=20, color='w')
     else:
-        ax.text(1, 3, 'Inital trajectory', fontsize=20)
+        ax.text(1, 3, 'Initial trajectory', fontsize=20)
 
     toplot = track[i, ::, ::, 0]
 

examples/image_ocr.py

@@ -80,7 +80,7 @@ def paint_text(text, w, h, rotate=False, ud=False, multi_fonts=False):
     with cairo.Context(surface) as context:
         context.set_source_rgb(1, 1, 1)  # White
         context.paint()
-        # this font list works in Centos 7
+        # this font list works in CentOS 7
         if multi_fonts:
             fonts = ['Century Schoolbook', 'Courier', 'STIX', 'URW Chancery L', 'FreeMono']
             context.select_font_face(np.random.choice(fonts), cairo.FONT_SLANT_NORMAL,

examples/imdb_lstm.py

@@ -1,4 +1,4 @@
-'''Trains a LSTM on the IMDB sentiment classification task.
+'''Trains an LSTM model on the IMDB sentiment classification task.
 The dataset is actually too small for LSTM to be of any advantage
 compared to simpler, much faster methods such as TF-IDF + LogReg.
 Notes:

examples/mnist_hierarchical_rnn.py

@@ -1,6 +1,6 @@
 """This is an example of using Hierarchical RNN (HRNN) to classify MNIST digits.
 
-HRNNs can learn across multiple levels of temporal hiearchy over a complex sequence.
+HRNNs can learn across multiple levels of temporal hierarchy over a complex sequence.
 Usually, the first recurrent layer of an HRNN encodes a sentence (e.g. of word vectors)
 into a  sentence vector. The second recurrent layer then encodes a sequence of
 such vectors (encoded by the first layer) into a document vector. This

examples/mnist_swwae.py

@@ -105,7 +105,7 @@ def getwhere(x):
                        'Theano backend for the time being, '
                        'because it requires taking the gradient '
                        'of a gradient, which isn\'t '
-                       'supported for all TF ops.')
+                       'supported for all TensorFlow ops.')
 
 # This example assume 'channels_first' data format.
 K.set_image_data_format('channels_first')

examples/pretrained_word_embeddings.py

@@ -1,7 +1,7 @@
 '''This script loads pre-trained word embeddings (GloVe embeddings)
 into a frozen Keras Embedding layer, and uses it to
 train a text classification model on the 20 Newsgroup dataset
-(classication of newsgroup messages into 20 different categories).
+(classification of newsgroup messages into 20 different categories).
 
 GloVe embedding data can be found at:
 http://nlp.stanford.edu/data/glove.6B.zip

keras/applications/resnet50.py

@@ -43,7 +43,7 @@ def identity_block(input_tensor, kernel_size, filters, stage, block):
     # Arguments
         input_tensor: input tensor
         kernel_size: default 3, the kernel size of middle conv layer at main path
-        filters: list of integers, the filterss of 3 conv layer at main path
+        filters: list of integers, the filters of 3 conv layer at main path
         stage: integer, current stage label, used for generating layer names
         block: 'a','b'..., current block label, used for generating layer names
 
@@ -81,7 +81,7 @@ def conv_block(input_tensor, kernel_size, filters, stage, block, strides=(2, 2))
     # Arguments
         input_tensor: input tensor
         kernel_size: default 3, the kernel size of middle conv layer at main path
-        filters: list of integers, the filterss of 3 conv layer at main path
+        filters: list of integers, the filters of 3 conv layer at main path
         stage: integer, current stage label, used for generating layer names
         block: 'a','b'..., current block label, used for generating layer names
 

keras/backend/cntk_backend.py

@@ -88,7 +88,7 @@ def in_train_phase(x, alt, training=None):
 
 def in_test_phase(x, alt):
     global _LEARNING_PHASE
-    # Similiar as in_train_phase, use element_select as workaround.
+    # Similar as in_train_phase, use element_select as workaround.
     if callable(x) and isinstance(x, C.cntk_py.Function) is False:
         x = x()
     if callable(alt) and isinstance(alt, C.cntk_py.Function) is False:
@@ -700,7 +700,7 @@ def _normalize_axis(axis, x):
 
     if nones > ndim:
         raise ValueError('CNTK Backend: tensor with keras shape: `%s` has '
-                         '%d cntk dynamic axis, this is not expected, plesae '
+                         '%d cntk dynamic axis, this is not expected, please '
                          'double check the keras shape history.' % (str(shape), nones))
 
     # Current cntk does not support shape like (1, batch). so using the workaround
@@ -1079,7 +1079,7 @@ def reshape(x, shape):
                 return x
             return C.user_function(ReshapeBatch(x, shape[1:]))
         else:
-            # no collaps, then first need to padding the shape
+            # no collapse, then first need to padding the shape
             if num_dynamic_axis >= len(shape):
                 i = 0
                 while i < len(shape):
@@ -1194,7 +1194,7 @@ def _static_rnn(step_function, inputs, initial_states,
         raise ValueError('CNTK Backend: the input of static rnn '
                          'has shape `%s`, the second axis '
                          'is not static. If you want to run '
-                         'rnn with non-static axis, plesae try '
+                         'rnn with non-static axis, please try '
                          'dynamic rnn with sequence axis.' % shape)
     if constants is None:
         constants = []

keras/backend/tensorflow_backend.py

@@ -35,7 +35,7 @@
 
 # This dictionary holds a mapping {graph: UID_DICT}.
 # each UID_DICT is a dictionary mapping name prefixes to a current index,
-# used for generatic graph-specific string UIDs
+# used for generic graph-specific string UIDs
 # for various names (e.g. layer names).
 _GRAPH_UID_DICTS = {}
 
@@ -3591,10 +3591,10 @@ def truncated_normal(shape, mean=0.0, stddev=1.0, dtype=None, seed=None):
 
 
 # CTC
-# tensorflow has a native implemenation, but it uses sparse tensors
+# TensorFlow has a native implementation, but it uses sparse tensors
 # and therefore requires a wrapper for Keras. The functions below convert
 # dense to sparse tensors and also wraps up the beam search code that is
-# in tensorflow's CTC implementation
+# in TensorFlow's CTC implementation
 
 
 def ctc_label_dense_to_sparse(labels, label_lengths):

keras/backend/theano_backend.py

@@ -1159,8 +1159,8 @@ def reverse(x, axes):
     return x[slices]
 
 
-def pattern_broadcast(x, broatcastable):
-    return T.patternbroadcast(x, broatcastable)
+def pattern_broadcast(x, broadcastable):
+    return T.patternbroadcast(x, broadcastable)
 
 # VALUE MANIPULATION
 
@@ -2249,7 +2249,7 @@ def truncated_normal(shape, mean=0.0, stddev=1.0, dtype=None, seed=None):
 # Theano implementation of CTC
 # Used with permission from Shawn Tan
 # https://github.com/shawntan/
-# Note that tensorflow's native CTC code is significantly
+# Note that TensorFlow's native CTC code is significantly
 # faster than this
 
 

keras/callbacks.py

@@ -248,7 +248,7 @@ class ProgbarLogger(Callback):
     # Arguments
         count_mode: One of "steps" or "samples".
             Whether the progress bar should
-            count samples seens or steps (batches) seen.
+            count samples seen or steps (batches) seen.
 
     # Raises
         ValueError: In case of invalid `count_mode`.

keras/engine/topology.py

@@ -616,7 +616,7 @@ def __call__(self, inputs, **kwargs):
             else:
                 output = output_ls_copy
 
-            # Infering the output shape is only relevant for Theano.
+            # Inferring the output shape is only relevant for Theano.
             if all([s is not None for s in _to_list(input_shape)]):
                 output_shape = self.compute_output_shape(input_shape)
             else:

keras/layers/convolutional_recurrent.py

@@ -503,7 +503,7 @@ def input_conv(self, x, w, b=None, padding='valid'):
                                   data_format=self.data_format)
         return conv_out
 
-    def reccurent_conv(self, x, w):
+    def recurrent_conv(self, x, w):
         conv_out = K.conv2d(x, w, strides=(1, 1),
                             padding='same',
                             data_format=self.data_format)
@@ -524,13 +524,13 @@ def step(self, inputs, states):
                               padding=self.padding)
         x_o = self.input_conv(inputs * dp_mask[3], self.kernel_o, self.bias_o,
                               padding=self.padding)
-        h_i = self.reccurent_conv(h_tm1 * rec_dp_mask[0],
+        h_i = self.recurrent_conv(h_tm1 * rec_dp_mask[0],
                                   self.recurrent_kernel_i)
-        h_f = self.reccurent_conv(h_tm1 * rec_dp_mask[1],
+        h_f = self.recurrent_conv(h_tm1 * rec_dp_mask[1],
                                   self.recurrent_kernel_f)
-        h_c = self.reccurent_conv(h_tm1 * rec_dp_mask[2],
+        h_c = self.recurrent_conv(h_tm1 * rec_dp_mask[2],
                                   self.recurrent_kernel_c)
-        h_o = self.reccurent_conv(h_tm1 * rec_dp_mask[3],
+        h_o = self.recurrent_conv(h_tm1 * rec_dp_mask[3],
                                   self.recurrent_kernel_o)
 
         i = self.recurrent_activation(x_i + h_i)

keras/layers/core.py

@@ -36,7 +36,7 @@ class Masking(Layer):
     # Example
 
     Consider a Numpy data array `x` of shape `(samples, timesteps, features)`,
-    to be fed to a LSTM layer.
+    to be fed to an LSTM layer.
     You want to mask timestep #3 and #5 because you lack data for
     these timesteps. You can:
 

keras/losses.py

@@ -4,6 +4,7 @@
 from .utils.generic_utils import deserialize_keras_object
 
 
+# noinspection SpellCheckingInspection
 def mean_squared_error(y_true, y_pred):
     return K.mean(K.square(y_pred - y_true), axis=-1)
 

keras/preprocessing/image.py

@@ -908,7 +908,7 @@ class DirectoryIterator(Iterator):
             to use for random transformations and normalization.
         target_size: tuple of integers, dimensions to resize input images to.
         color_mode: One of `"rgb"`, `"grayscale"`. Color mode to read images.
-        classes: Optional list of strings, names of sudirectories
+        classes: Optional list of strings, names of subdirectories
             containing images from each class (e.g. `["dogs", "cats"]`).
             It will be computed automatically if not set.
         class_mode: Mode for yielding the targets:

keras/preprocessing/sequence.py

@@ -139,7 +139,7 @@ def skipgrams(sequence, vocabulary_size,
             integers (eg. [0, 1, 1 .. ]),
             if True labels will be categorical eg. [[1,0],[0,1],[0,1] .. ]
         sampling_table: 1D array of size `vocabulary_size` where the entry i
-            encodes the probabibily to sample a word of rank i.
+            encodes the probability to sample a word of rank i.
         seed: random seed.
 
     # Returns

tests/integration_tests/test_temporal_data_tasks.py

@@ -206,7 +206,7 @@ def test_masked_temporal():
     assert(np.abs(history.history['loss'][-1] - ground_truth) < 0.06)
 
 
-@pytest.mark.skipif(K.backend() != 'tensorflow', reason='Requires TF backend')
+@pytest.mark.skipif(K.backend() != 'tensorflow', reason='Requires TensorFlow backend')
 @keras_test
 def test_embedding_with_clipnorm():
     model = Sequential()

tests/keras/applications/applications_test.py

@@ -164,31 +164,31 @@ def test_vgg19_pooling_specified_input_shape():
 
 @keras_test
 @pytest.mark.skipif((K.backend() != 'tensorflow'),
-                    reason='Requires tensorflow backend')
+                    reason='Requires TensorFlow backend')
 def test_xception():
     model = applications.Xception(weights=None)
     assert model.output_shape == (None, 1000)
 
 
 @keras_test
 @pytest.mark.skipif((K.backend() != 'tensorflow'),
-                    reason='Requires tensorflow backend')
+                    reason='Requires TensorFlow backend')
 def test_xception_notop():
     model = applications.Xception(weights=None, include_top=False)
     assert model.output_shape == (None, None, None, 2048)
 
 
 @keras_test
 @pytest.mark.skipif((K.backend() != 'tensorflow'),
-                    reason='Requires tensorflow backend')
+                    reason='Requires TensorFlow backend')
 def test_xception_pooling():
     model = applications.Xception(weights=None, include_top=False, pooling='avg')
     assert model.output_shape == (None, 2048)
 
 
 @keras_test
 @pytest.mark.skipif((K.backend() != 'tensorflow'),
-                    reason='Requires tensorflow backend')
+                    reason='Requires TensorFlow backend')
 def test_xception_variable_input_channels():
     input_shape = (1, None, None) if K.image_data_format() == 'channels_first' else (None, None, 1)
     model = applications.Xception(weights=None, include_top=False, input_shape=input_shape)

tests/keras/backend/backend_test.py

@@ -22,7 +22,7 @@
     BACKENDS.append(KTF)
 except ImportError:
     KTF = None
-    warnings.warn('Could not import the Tensorflow backend.')
+    warnings.warn('Could not import the TensorFlow backend.')
 
 try:
     from keras.backend import theano_backend as KTH
@@ -694,7 +694,7 @@ def test_nn_operations(self):
         # cross_entropy call require the label is a valid probability distribution,
         # otherwise it is garbage in garbage out...
         # due to the algo difference, we can't guarantee CNTK has the same result on the garbage input.
-        # so create a seperate test case for valid lable input
+        # so create a separate test case for valid label input
         check_two_tensor_operation('categorical_crossentropy', (4, 2), (4, 2), [KTH, KTF], from_logits=True)
         xval = np.asarray([[0.26157712, 0.0432167], [-0.43380741, 0.30559841],
                            [0.20225059, -0.38956559], [-0.13805378, 0.08506755]], dtype=np.float32)
@@ -1028,7 +1028,7 @@ def test_bias_add(self):
                                        BACKENDS, cntk_dynamicity=True,
                                        data_format=data_format)
 
-        # Test invalid use casess
+        # Test invalid use cases
         for k in BACKENDS:
             x = k.variable(np.random.random(x_shape))
             b = k.variable(np.random.random(bias_shape))

tests/keras/engine/test_topology.py

@@ -689,7 +689,7 @@ def test_shared_layer_depth_is_correct():
 
 
 @keras_test
-def test_layer_sharing_at_heterogenous_depth():
+def test_layer_sharing_at_heterogeneous_depth():
     x_val = np.random.random((10, 5))
 
     x = Input(shape=(5,))
@@ -711,7 +711,7 @@ def test_layer_sharing_at_heterogenous_depth():
 
 
 @keras_test
-def test_layer_sharing_at_heterogenous_depth_with_concat():
+def test_layer_sharing_at_heterogeneous_depth_with_concat():
     input_shape = (16, 9, 3)
     input_layer = Input(shape=input_shape)
 

tests/keras/engine/test_training.py

@@ -417,7 +417,7 @@ def gen_data(batch_sz):
     assert all(['Sequence' not in str(w_.message) for w_ in w]), 'A warning was raised for Sequence.'
 
 
-@pytest.mark.skipif(K.backend() != 'tensorflow', reason='sparse operations supported only by TF')
+@pytest.mark.skipif(K.backend() != 'tensorflow', reason='sparse operations supported only by TensorFlow')
 @keras_test
 def test_sparse_input_validation_split():
     test_input = sparse.random(6, 3, density=0.25).tocsr()
@@ -478,7 +478,7 @@ def test_check_bad_shape():
     assert 'targets to have the same shape' in str(exc)
 
 
-@pytest.mark.skipif(K.backend() != 'tensorflow', reason='Requires TF backend')
+@pytest.mark.skipif(K.backend() != 'tensorflow', reason='Requires TensorFlow backend')
 @keras_test
 def test_model_with_input_feed_tensor():
     """We test building a model with a TF variable as input.

tests/keras/legacy/layers_test.py

@@ -275,7 +275,7 @@ def test_sequential_regression():
 
     branch_1_2 = models.Sequential([legacy_layers.Merge([branch_1, branch_2], mode='concat')], name='branch_1_2')
     branch_1_2.add(layers.Dense(16, name='dense_1_2-0'))
-    # test whether impromtu input_shape breaks the model
+    # test whether impromptu input_shape breaks the model
     branch_1_2.add(layers.Dense(16, input_shape=(16,), name='dense_1_2-1'))
 
     model = models.Sequential([legacy_layers.Merge([branch_1_2, branch_3], mode='concat')], name='final')

tests/keras/optimizers_test.py

@@ -117,7 +117,7 @@ def test_clipvalue():
 
 @keras_test
 @pytest.mark.skipif((K.backend() != 'tensorflow'),
-                    reason='Requires tensorflow backend')
+                    reason='Requires TensorFlow backend')
 def test_tfoptimizer():
     from keras import constraints
     from tensorflow import train