Make DefaultDataCollator importable from root (#14588)

* Make DefaultDataCollator importable from root

* Add documentation for DefaultDataCollator and add return_tensors argument to all class docstrings

* make style

* Add DefaultDataCollator to data_collator.rst

* Add DefaultDataCollator to data_collator.rst

docs/source/main_classes/data_collator.rst

@@ -29,6 +29,13 @@ Default data collator
 .. autofunction:: transformers.data.data_collator.default_data_collator
 
 
+DefaultDataCollator
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.data.data_collator.DefaultDataCollator
+    :members:
+
+
 DataCollatorWithPadding
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

src/transformers/__init__.py

@@ -92,6 +92,7 @@ _import_structure = {
         "DataCollatorForTokenClassification",
         "DataCollatorForWholeWordMask",
         "DataCollatorWithPadding",
+        "DefaultDataCollator",
         "default_data_collator",
     ],
     "feature_extraction_sequence_utils": ["SequenceFeatureExtractor"],
@@ -2087,6 +2088,7 @@ if TYPE_CHECKING:
         DataCollatorForTokenClassification,
         DataCollatorForWholeWordMask,
         DataCollatorWithPadding,
+        DefaultDataCollator,
         default_data_collator,
     )
     from .feature_extraction_sequence_utils import SequenceFeatureExtractor

src/transformers/data/__init__.py

@@ -24,6 +24,7 @@ from .data_collator import (
     DataCollatorForTokenClassification,
     DataCollatorForWholeWordMask,
     DataCollatorWithPadding,
+    DefaultDataCollator,
     default_data_collator,
 )
 from .metrics import glue_compute_metrics, xnli_compute_metrics

src/transformers/data/data_collator.py

@@ -72,6 +72,24 @@ def default_data_collator(features: List[InputDataClass], return_tensors="pt") -
 
 @dataclass
 class DefaultDataCollator(DataCollatorMixin):
+    """
+    Very simple data collator that simply collates batches of dict-like objects and performs special handling for
+    potential keys named:
+
+        - ``label``: handles a single value (int or float) per object
+        - ``label_ids``: handles a list of values per object
+
+    Does not do any additional preprocessing: property names of the input object will be used as corresponding inputs
+    to the model. See glue and ner for example of how it's useful.
+
+    This is an object (like other data collators) rather than a pure function like default_data_collator. This can be
+    helpful if you need to set a return_tensors value at initialization.
+
+    Args:
+        return_tensors (:obj:`str`):
+            The type of Tensor to return. Allowable values are "np", "pt" and "tf".
+    """
+
     return_tensors: str = "pt"
 
     def __call__(self, features: List[Dict[str, Any]], return_tensors=None) -> Dict[str, Any]:
@@ -214,6 +232,8 @@ class DataCollatorWithPadding:
 
             This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >=
             7.5 (Volta).
+        return_tensors (:obj:`str`):
+            The type of Tensor to return. Allowable values are "np", "pt" and "tf".
     """
 
     tokenizer: PreTrainedTokenizerBase
@@ -266,6 +286,8 @@ class DataCollatorForTokenClassification(DataCollatorMixin):
             7.5 (Volta).
         label_pad_token_id (:obj:`int`, `optional`, defaults to -100):
             The id to use when padding the labels (-100 will be automatically ignore by PyTorch loss functions).
+        return_tensors (:obj:`str`):
+            The type of Tensor to return. Allowable values are "np", "pt" and "tf".
     """
 
     tokenizer: PreTrainedTokenizerBase
@@ -519,6 +541,8 @@ class DataCollatorForSeq2Seq:
             7.5 (Volta).
         label_pad_token_id (:obj:`int`, `optional`, defaults to -100):
             The id to use when padding the labels (-100 will be automatically ignored by PyTorch loss functions).
+        return_tensors (:obj:`str`):
+            The type of Tensor to return. Allowable values are "np", "pt" and "tf".
     """
 
     tokenizer: PreTrainedTokenizerBase
@@ -591,6 +615,8 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
             The probability with which to (randomly) mask tokens in the input, when :obj:`mlm` is set to :obj:`True`.
         pad_to_multiple_of (:obj:`int`, `optional`):
             If set will pad the sequence to a multiple of the provided value.
+        return_tensors (:obj:`str`):
+            The type of Tensor to return. Allowable values are "np", "pt" and "tf".
 
     .. note::