Fix some more typos. (#757)

tensorpack/input_source/input_source.py

@@ -104,7 +104,7 @@ class FeedfreeInput(InputSource):
         pass
 
 
-# TODO enqueu_many? https://github.com/tensorflow/tensorflow/issues/7817#issuecomment-282053155
+# TODO enqueue_many? https://github.com/tensorflow/tensorflow/issues/7817#issuecomment-282053155
 class EnqueueThread(ShareSessionThread):
     def __init__(self, queue, ds, placehdrs):
         super(EnqueueThread, self).__init__()

tensorpack/input_source/input_source_base.py

@@ -110,7 +110,7 @@ class InputSource(object):
     @memoized
     def get_callbacks(self):
         """
-        An InputSource might need some extra maintainance during training,
+        An InputSource might need some extra maintenance during training,
         which is done also through the Callback interface.
         This method returns the callbacks and the return value will be memoized.
 

tensorpack/models/pool.py

@@ -204,7 +204,7 @@ class TestPool(TestModel):
         res = self.run_variable(output)
         self.assertEqual(res.shape, (1, scale * h, scale * w, 3))
 
-        # mat is on cornser
+        # mat is on corner
         ele = res[0, ::scale, ::scale, 0]
         self.assertTrue((ele == mat[:, :, 0]).all())
         # the rest are zeros

tensorpack/models/tflayer.py

@@ -34,7 +34,7 @@ def convert_to_tflayer_args(args_names, name_mapping):
     1. data_format becomes tf.layers style
     2. nl becomes activation
     3. initializers are renamed
-    4. positional args are transformed to correspoding kwargs, according to args_names
+    4. positional args are transformed to corresponding kwargs, according to args_names
     5. kwargs are mapped to tf.layers names if needed, by name_mapping
     """
 

tensorpack/predict/base.py

@@ -25,7 +25,7 @@ class PredictorBase(object):
 
     Attributes:
         return_input (bool): whether the call will also return (inputs, outputs)
-            or just outpus
+            or just outputs
     """
 
     def __call__(self, *args):

tensorpack/predict/concurrency.py

@@ -130,14 +130,14 @@ class PredictorWorkerThread(StoppableThread, ShareSessionThread):
 
 class MultiThreadAsyncPredictor(AsyncPredictorBase):
     """
-    An multithread online async predictor which runs a list of OnlinePredictor.
+    An multithreaded online async predictor which runs a list of OnlinePredictor.
     It would do an extra batching internally.
     """
 
     def __init__(self, predictors, batch_size=5):
         """
         Args:
-            predictors (list): a list of OnlinePredictor avaiable to use.
+            predictors (list): a list of OnlinePredictor available to use.
             batch_size (int): the maximum of an internal batch.
         """
         assert len(predictors)

tensorpack/predict/dataset.py

@@ -79,7 +79,7 @@ class SimpleDatasetPredictor(DatasetPredictorBase):
 
 class MultiProcessDatasetPredictor(DatasetPredictorBase):
     """
-    Run prediction in multiprocesses, on either CPU or GPU.
+    Run prediction in multiple processes, on either CPU or GPU.
     Each process fetch datapoints as tasks and run predictions independently.
     """
     # TODO allow unordered

tensorpack/tfutils/collection.py

@@ -170,5 +170,5 @@ class CollectionGuard(object):
         """
         new = tf.get_collection(key)
         old = set(self.original.get(key, []))
-        # presist the order in new
+        # persist the order in new
         return [x for x in new if x not in old]

tensorpack/tfutils/export.py

@@ -66,7 +66,7 @@ class ModelExport(object):
                tags=[tf.saved_model.tag_constants.SERVING],
                signature_name='prediction_pipeline'):
         """
-        Use SavedModelBuilder to export a trained model without tensorpack depency.
+        Use SavedModelBuilder to export a trained model without tensorpack dependency.
 
         Remarks:
             This produces

tensorpack/tfutils/symbolic_functions.py

@@ -10,7 +10,7 @@ from ..utils.develop import deprecated
 # __all__ = ['get_scalar_var']
 
 
-# this function exists for backwards-compatibilty
+# this function exists for backwards-compatibility
 def prediction_incorrect(logits, label, topk=1, name='incorrect_vector'):
     return tf.cast(tf.logical_not(tf.nn.in_top_k(logits, label, topk)),
                    tf.float32, name=name)

tensorpack/train/config.py

@@ -201,7 +201,7 @@ class AutoResumeTrainConfig(TrainConfig):
             without changing any line of code or command line arguments.
             So it's useful to let resume take priority over user-provided arguments sometimes:
 
-            If your training starts from a pretrained model,
+            If your training starts from a pre-trained model,
             you would want it to use user-provided model loader at the
             beginning, but a "resume" model loader when the job was
             interrupted and restarted.

tensorpack/utils/argtools.py

@@ -16,7 +16,7 @@ __all__ = ['map_arg', 'memoized', 'graph_memoized', 'shape2d', 'shape4d',
 
 def map_arg(**maps):
     """
-    Apply a mapping on certains argument before calling the original function.
+    Apply a mapping on certain argument before calling the original function.
 
     Args:
         maps (dict): {argument_name: map_func}

tensorpack/utils/logger.py

@@ -25,7 +25,7 @@ class _MyFormatter(logging.Formatter):
         else:
             fmt = date + ' ' + msg
         if hasattr(self, '_style'):
-            # Python3 compatibilty
+            # Python3 compatibility
             self._style._fmt = fmt
         self._fmt = fmt
         return super(_MyFormatter, self).format(record)