Big file_utils cleanup (#16396)

* Big file_utils cleanup

* This one still needs to be treated separately

src/transformers/models/deberta/modeling_deberta.py

@@ -871,7 +871,7 @@ DEBERTA_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/deberta/modeling_tf_deberta.py

@@ -1063,7 +1063,7 @@ DEBERTA_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~transformers.file_utils.ModelOutput``] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput``] instead of a plain tuple.
 """
 
 

src/transformers/models/deberta_v2/modeling_deberta_v2.py

@@ -965,7 +965,7 @@ DEBERTA_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/deberta_v2/modeling_tf_deberta_v2.py

@@ -1164,7 +1164,7 @@ DEBERTA_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~transformers.file_utils.ModelOutput``] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput``] instead of a plain tuple.
 """
 
 

src/transformers/models/decision_transformer/__init__.py

@@ -18,7 +18,7 @@
 from typing import TYPE_CHECKING
 
 # rely on isort to merge the imports
-from ...file_utils import _LazyModule, is_torch_available
+from ...utils import _LazyModule, is_torch_available
 
 
 _import_structure = {

src/transformers/models/decision_transformer/modeling_decision_transformer.py

@@ -25,14 +25,14 @@ from packaging import version
 from torch import nn
 
 from ...activations import ACT2FN
-from ...file_utils import (
+from ...modeling_utils import Conv1D, PreTrainedModel, find_pruneable_heads_and_indices, prune_conv1d_layer
+from ...utils import (
     ModelOutput,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
+    logging,
     replace_return_docstrings,
 )
-from ...modeling_utils import Conv1D, PreTrainedModel, find_pruneable_heads_and_indices, prune_conv1d_layer
-from ...utils import logging
 
 
 if version.parse(torch.__version__) >= version.parse("1.6"):

src/transformers/models/deit/feature_extraction_deit.py

@@ -107,7 +107,7 @@ class DeiTFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin):
                 tensor. In case of a NumPy array/PyTorch tensor, each image should be of shape (C, H, W), where C is a
                 number of channels, H and W are image height and width.
 
-            return_tensors (`str` or [`~file_utils.TensorType`], *optional*, defaults to `'np'`):
+            return_tensors (`str` or [`~utils.TensorType`], *optional*, defaults to `'np'`):
                 If set, will return tensors of a particular framework. Acceptable values are:
 
                 - `'tf'`: Return TensorFlow `tf.constant` objects.

src/transformers/models/deit/modeling_deit.py

@@ -460,7 +460,7 @@ DEIT_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/detr/feature_extraction_detr.py

@@ -455,7 +455,7 @@ class DetrFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin):
                 - 1 for pixels that are real (i.e. **not masked**),
                 - 0 for pixels that are padding (i.e. **masked**).
 
-            return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
+            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors instead of NumPy arrays. If set to `'pt'`, return PyTorch `torch.Tensor`
                 objects.
 
@@ -638,7 +638,7 @@ class DetrFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin):
         Args:
             pixel_values_list (`List[torch.Tensor]`):
                 List of images (pixel values) to be padded. Each image should be a tensor of shape (C, H, W).
-            return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
+            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors instead of NumPy arrays. If set to `'pt'`, return PyTorch `torch.Tensor`
                 objects.
 

src/transformers/models/detr/modeling_detr.py

@@ -868,7 +868,7 @@ DETR_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 
@@ -932,7 +932,7 @@ class DetrEncoder(DetrPreTrainedModel):
                 Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors
                 for more detail.
             return_dict (`bool`, *optional*):
-                Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+                Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
         """
         output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
         output_hidden_states = (
@@ -1054,7 +1054,7 @@ class DetrDecoder(DetrPreTrainedModel):
                 Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors
                 for more detail.
             return_dict (`bool`, *optional*):
-                Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+                Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
         """
         output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
         output_hidden_states = (

src/transformers/models/distilbert/modeling_distilbert.py

@@ -446,7 +446,7 @@ DISTILBERT_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/distilbert/modeling_flax_distilbert.py

@@ -89,7 +89,7 @@ DISTILBERT_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/distilbert/modeling_tf_distilbert.py

@@ -508,8 +508,8 @@ DISTILBERT_INPUTS_DOCSTRING = r"""
             more detail. This argument can be used only in eager mode, in graph mode the value in the config will be
             used instead.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple. This argument can be used
-            in eager mode, in graph mode the value will always be set to True.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in
+            eager mode, in graph mode the value will always be set to True.
         training (`bool`, *optional*, defaults to `False`):
             Whether or not to use the model in training mode (some modules like dropout modules have different
             behaviors between training and evaluation).

src/transformers/models/dpr/modeling_dpr.py

@@ -398,7 +398,7 @@ DPR_ENCODERS_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 DPR_READER_INPUTS_DOCSTRING = r"""
@@ -434,7 +434,7 @@ DPR_READER_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/dpr/modeling_tf_dpr.py

@@ -487,8 +487,8 @@ TF_DPR_ENCODERS_INPUTS_DOCSTRING = r"""
             more detail. This argument can be used only in eager mode, in graph mode the value in the config will be
             used instead.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple. This argument can be used
-            in eager mode, in graph mode the value will always be set to True.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in
+            eager mode, in graph mode the value will always be set to True.
         training (`bool`, *optional*, defaults to `False`):
             Whether or not to use the model in training mode (some modules like dropout modules have different
             behaviors between training and evaluation).
@@ -523,8 +523,8 @@ TF_DPR_READER_INPUTS_DOCSTRING = r"""
             more detail. This argument can be used only in eager mode, in graph mode the value in the config will be
             used instead.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple. This argument can be used
-            in eager mode, in graph mode the value will always be set to True.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in
+            eager mode, in graph mode the value will always be set to True.
         training (`bool`, *optional*, defaults to `False`):
             Whether or not to use the model in training mode (some modules like dropout modules have different
             behaviors between training and evaluation).

src/transformers/models/dpr/tokenization_dpr.py

@@ -144,7 +144,7 @@ CUSTOM_DPR_READER_DOCSTRING = r"""
             The passages titles to be encoded. This can be a string or a list of strings if there are several passages.
         texts (`str` or `List[str]`):
             The passages texts to be encoded. This can be a string or a list of strings if there are several passages.
-        padding (`bool`, `str` or [`~file_utils.PaddingStrategy`], *optional*, defaults to `False`):
+        padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`):
             Activates and controls padding. Accepts the following values:
 
             - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single sequence
@@ -174,7 +174,7 @@ CUSTOM_DPR_READER_DOCSTRING = r"""
                 If left unset or set to `None`, this will use the predefined model maximum length if a maximum length
                 is required by one of the truncation/padding parameters. If the model has no specific maximum input
                 length (like XLNet) truncation/padding to a maximum length will be deactivated.
-        return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
+        return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors instead of list of python integers. Acceptable values are:
 
                 - `'tf'`: Return TensorFlow `tf.constant` objects.

src/transformers/models/dpr/tokenization_dpr_fast.py

@@ -145,7 +145,7 @@ CUSTOM_DPR_READER_DOCSTRING = r"""
             The passages titles to be encoded. This can be a string or a list of strings if there are several passages.
         texts (`str` or `List[str]`):
             The passages texts to be encoded. This can be a string or a list of strings if there are several passages.
-        padding (`bool`, `str` or [`~file_utils.PaddingStrategy`], *optional*, defaults to `False`):
+        padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`):
             Activates and controls padding. Accepts the following values:
 
             - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single sequence
@@ -175,7 +175,7 @@ CUSTOM_DPR_READER_DOCSTRING = r"""
                 If left unset or set to `None`, this will use the predefined model maximum length if a maximum length
                 is required by one of the truncation/padding parameters. If the model has no specific maximum input
                 length (like XLNet) truncation/padding to a maximum length will be deactivated.
-        return_tensors (`str` or [`~file_utils.TensorType`], *optional*):
+        return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors instead of list of python integers. Acceptable values are:
 
                 - `'tf'`: Return TensorFlow `tf.constant` objects.

src/transformers/models/electra/modeling_electra.py

@@ -794,7 +794,7 @@ ELECTRA_INPUTS_DOCSTRING = r"""
             Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
             more detail.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/electra/modeling_flax_electra.py

@@ -134,7 +134,7 @@ ELECTRA_INPUTS_DOCSTRING = r"""
             - 0 indicates the head is **masked**.
 
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
 
 """
 

src/transformers/models/electra/modeling_tf_electra.py

@@ -907,8 +907,8 @@ ELECTRA_INPUTS_DOCSTRING = r"""
             more detail. This argument can be used only in eager mode, in graph mode the value in the config will be
             used instead.
         return_dict (`bool`, *optional*):
-            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple. This argument can be used
-            in eager mode, in graph mode the value will always be set to True.
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in
+            eager mode, in graph mode the value will always be set to True.
         training (`bool`, *optional*, defaults to `False`):
             Whether or not to use the model in training mode (some modules like dropout modules have different
             behaviors between training and evaluation).