Fix doc errors and typos across the board (#8139)

* Fix doc errors and typos across the board

* Fix a typo

* Fix the CI

* Fix more typos

* Fix CI

* More fixes

* Fix CI

* More fixes

* More fixes

src/transformers/tokenization_utils_fast.py

@@ -53,7 +53,7 @@ TOKENIZER_FILE = "tokenizer.json"
 SPECIAL_TOKENS_MAP_FILE = "special_tokens_map.json"
 TOKENIZER_CONFIG_FILE = "tokenizer_config.json"
 
-# Slow tokenizers have an additional addedd tokens files
+# Slow tokenizers have an additional added tokens files
 ADDED_TOKENS_FILE = "added_tokens.json"
 
 
@@ -211,7 +211,7 @@ class PreTrainedTokenizerFast(PreTrainedTokenizerBase):
         vocabulary.
 
         Args:
-            token (:obj:`str` or :obj:`List[str]`): One or several token(s) to convert to token id(s).
+            tokens (:obj:`str` or :obj:`List[str]`): One or several token(s) to convert to token id(s).
 
         Returns:
             :obj:`int` or :obj:`List[int]`: The token id or list of token ids.
@@ -473,7 +473,7 @@ class PreTrainedTokenizerFast(PreTrainedTokenizerBase):
         )
 
         # Return tensor is None, then we can remove the leading batch axis
-        # Overfolwing tokens are returned as a batch of output so we keep them in this case
+        # Overflowing tokens are returned as a batch of output so we keep them in this case
         if return_tensors is None and not return_overflowing_tokens:
             batched_output = BatchEncoding(
                 {

src/transformers/tokenization_xlm.py

@@ -909,7 +909,7 @@ class XLMTokenizer(PreTrainedTokenizer):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(
                 map(

src/transformers/tokenization_xlm_prophetnet.py

@@ -202,7 +202,7 @@ class XLMProphetNetTokenizer(PreTrainedTokenizer):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

src/transformers/tokenization_xlm_roberta.py

@@ -205,7 +205,7 @@ class XLMRobertaTokenizer(PreTrainedTokenizer):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

src/transformers/tokenization_xlm_roberta_fast.py

@@ -194,7 +194,7 @@ class XLMRobertaTokenizerFast(PreTrainedTokenizerFast):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

src/transformers/tokenization_xlnet.py

@@ -270,7 +270,7 @@ class XLNetTokenizer(PreTrainedTokenizer):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

src/transformers/tokenization_xlnet_fast.py

@@ -209,7 +209,7 @@ class XLNetTokenizerFast(PreTrainedTokenizerFast):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

src/transformers/trainer.py

@@ -1524,8 +1524,6 @@ class Trainer:
         method in the model or subclass and override this method.
 
         Args:
-            model (:obj:`nn.Module`):
-                The model to evaluate.
             inputs (:obj:`Dict[str, Union[torch.Tensor, Any]]`):
                 The inputs and targets of the model.
 

src/transformers/trainer_pt_utils.py

@@ -135,7 +135,7 @@ def torch_distributed_zero_first(local_rank: int):
 
 class SequentialDistributedSampler(Sampler):
     """
-    Distributed Sampler that subsamples indicies sequentially, making it easier to collate all results at the end.
+    Distributed Sampler that subsamples indices sequentially, making it easier to collate all results at the end.
 
     Even though we only use this sampler for eval and predict (no training), which means that the model params won't
     have to be synced (i.e. will not hang for synchronization even if varied number of forward passes), we still add

src/transformers/training_args.py

@@ -402,7 +402,7 @@ class TrainingArguments:
             n_gpu = torch.cuda.device_count()
         else:
             # Here, we'll use torch.distributed.
-            # Initializes the distributed backend which will take care of sychronizing nodes/GPUs
+            # Initializes the distributed backend which will take care of synchronizing nodes/GPUs
             torch.distributed.init_process_group(backend="nccl")
             device = torch.device("cuda", self.local_rank)
             n_gpu = 1

src/transformers/utils/notebook.py

@@ -205,7 +205,7 @@ class NotebookTrainingTracker(NotebookProgressBar):
 
         num_steps (:obj:`int`): The number of steps during training.
         column_names (:obj:`List[str]`, `optional`):
-            The list of column names for the metrics table (will be infered from the first call to
+            The list of column names for the metrics table (will be inferred from the first call to
             :meth:`~transformers.utils.notebook.NotebookTrainingTracker.write_line` if not set).
     """
 
@@ -246,7 +246,7 @@ class NotebookTrainingTracker(NotebookProgressBar):
 
     def add_child(self, total, prefix=None, width=300):
         """
-        Add a child progress bar disaplyed under the table of metrics. The child progress bar is returned (so it can be
+        Add a child progress bar displayed under the table of metrics. The child progress bar is returned (so it can be
         easily updated).
 
         Args:

templates/adding_a_new_example_script/run_xxx.py

@@ -45,7 +45,7 @@ from utils_squad import (
     write_predictions_extended,
 )
 
-# The follwing import is the official SQuAD evaluation script (2.0).
+# The following import is the official SQuAD evaluation script (2.0).
 # You can remove it from the dependencies if you are using this script outside of the library
 # We've added it here for automated tests (see examples/test_examples.py file)
 from utils_squad_evaluate import EVAL_OPTS

templates/adding_a_new_example_script/utils_xxx.py

@@ -426,8 +426,8 @@ def _improve_answer_span(doc_tokens, input_start, input_end, tokenizer, orig_ans
     #
     # However, this is not always possible. Consider the following:
     #
-    #   Question: What country is the top exporter of electornics?
-    #   Context: The Japanese electronics industry is the lagest in the world.
+    #   Question: What country is the top exporter of electronics?
+    #   Context: The Japanese electronics industry is the largest in the world.
     #   Answer: Japan
     #
     # In this case, the annotator chose "Japan" as a character sub-span of

templates/adding_a_new_model/configuration_xxx.py

@@ -57,7 +57,7 @@ class XxxConfig(PretrainedConfig):
 
             If string, :obj:`"gelu"`, :obj:`"relu"`, :obj:`"swish"` and :obj:`"gelu_new"` are supported.
         hidden_dropout_prob (:obj:`float`, `optional`, defaults to 0.1):
-            The dropout probabilitiy for all fully connected layers in the embeddings, encoder, and pooler.
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
         attention_probs_dropout_prob (:obj:`float`, `optional`, defaults to 0.1):
             The dropout ratio for the attention probabilities.
         max_position_embeddings (:obj:`int`, `optional`, defaults to 512):

templates/adding_a_new_model/modeling_tf_xxx.py

@@ -352,7 +352,7 @@ XXX_INPUTS_DOCSTRING = r"""
 
 
 @add_start_docstrings(
-    "The bare XXX Model transformer outputing raw hidden-states without any specific head on top.",
+    "The bare XXX Model transformer outputting raw hidden-states without any specific head on top.",
     XXX_START_DOCSTRING,
 )
 class TFXxxModel(TFXxxPreTrainedModel):

templates/adding_a_new_model/tokenization_xxx.py

@@ -237,7 +237,7 @@ class XxxTokenizer(PreTrainedTokenizer):
             if token_ids_1 is not None:
                 raise ValueError(
                     "You should not supply a second sequence if the provided sequence of "
-                    "ids is already formated with special tokens for the model."
+                    "ids is already formatted with special tokens for the model."
                 )
             return list(map(lambda x: 1 if x in [self.sep_token_id, self.cls_token_id] else 0, token_ids_0))
 

tests/test_logging.py

@@ -34,7 +34,7 @@ class HfArgumentParserTest(unittest.TestCase):
         logger = logging.get_logger("transformers.tokenization_bart")
         msg = "Testing 1, 2, 3"
 
-        # should be able to log warnings (if default settings weren't overriden by `pytest --log-level-all`)
+        # should be able to log warnings (if default settings weren't overridden by `pytest --log-level-all`)
         if level_origin <= logging.WARNING:
             with CaptureLogger(logger) as cl:
                 logger.warn(msg)

tests/test_modeling_common.py

@@ -1130,7 +1130,7 @@ class UtilsFunctionsTest(unittest.TestCase):
                     2.12662941,
                     -9.32562038,
                     2.35652522,
-                ],  # cummulative prob of 5 highest values <= 0.6
+                ],  # cumulative prob of 5 highest values <= 0.6
                 [
                     0.58425518,
                     4.53139238,
@@ -1162,7 +1162,7 @@ class UtilsFunctionsTest(unittest.TestCase):
                     9.67702323,  # 1st highest value; idx. 27
                     -5.89478553,
                     1.85370467,
-                ],  # cummulative prob of 5 highest values <= 0.6
+                ],  # cumulative prob of 5 highest values <= 0.6
             ],
             dtype=torch.float,
             device=torch_device,

tests/test_modeling_tf_lxmert.py

@@ -497,7 +497,7 @@ class TFLxmertModelTest(TFModelTesterMixin, unittest.TestCase):
                 return_obj_labels="PreTraining" in model_class.__name__
             )
 
-            pt_model_class_name = model_class.__name__[2:]  # Skip the "TF" at the beggining
+            pt_model_class_name = model_class.__name__[2:]  # Skip the "TF" at the beginning
             pt_model_class = getattr(transformers, pt_model_class_name)
 
             config.output_hidden_states = True

utils/check_copies.py

@@ -100,7 +100,7 @@ def is_copy_consistent(filename, overwrite=False):
         lines = f.readlines()
     diffs = []
     line_index = 0
-    # Not a foor loop cause `lines` is going to change (if `overwrite=True`).
+    # Not a for loop cause `lines` is going to change (if `overwrite=True`).
     while line_index < len(lines):
         search = _re_copy_warning.search(lines[line_index])
         if search is None:
@@ -164,9 +164,9 @@ def check_copies(overwrite: bool = False):
     if not overwrite and len(diffs) > 0:
         diff = "\n".join(diffs)
         raise Exception(
-            "Found the follwing copy inconsistencies:\n"
+            "Found the following copy inconsistencies:\n"
             + diff
-            + "\nRun `make fix-copies` or `python utils/check_copies --fix_and_overwrite` to fix them."
+            + "\nRun `make fix-copies` or `python utils/check_copies.py --fix_and_overwrite` to fix them."
         )
     check_model_list_copy(overwrite=overwrite)