Big file_utils cleanup (#16396)

* Big file_utils cleanup

* This one still needs to be treated separately

src/transformers/models/ibert/modeling_ibert.py

@@ -722,7 +722,7 @@ IBERT_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/imagegpt/feature_extraction_imagegpt.py

@@ -117,7 +117,7 @@ class ImageGPTFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMix
                 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/imagegpt/modeling_imagegpt.py

@@ -608,7 +608,7 @@ IMAGEGPT_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/layoutlm/modeling_layoutlm.py

@@ -695,7 +695,7 @@ LAYOUTLM_INPUTS_DOCSTRING = r"""
             If set to `True`, the hidden states of all layers are returned. See `hidden_states` under returned tensors
             for more detail.
         return_dict (`bool`, *optional*):
-            If set to `True`, the model will return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+            If set to `True`, the model will return a [`~utils.ModelOutput`] instead of a plain tuple.
 """
 
 

src/transformers/models/layoutlm/modeling_tf_layoutlm.py

@@ -891,7 +891,7 @@ LAYOUTLM_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.
         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/layoutlmv2/feature_extraction_layoutlmv2.py

@@ -131,7 +131,7 @@ class LayoutLMv2FeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionM
                 The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch
                 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/layoutlmv2/modeling_layoutlmv2.py

@@ -682,7 +682,7 @@ LAYOUTLMV2_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/layoutlmv2/tokenization_layoutlmv2.py

@@ -60,7 +60,7 @@ PRETRAINED_INIT_CONFIGURATION = {
 LAYOUTLMV2_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING = r"""
             add_special_tokens (`bool`, *optional*, defaults to `True`):
                 Whether or not to encode the sequences with the special tokens relative to their model.
-            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
@@ -97,7 +97,7 @@ LAYOUTLMV2_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING = r"""
             pad_to_multiple_of (`int`, *optional*):
                 If set will pad the sequence to a multiple of the provided value. This is especially useful to enable
                 the use of Tensor Cores on NVIDIA hardware with compute capability >= 7.5 (Volta).
-            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/led/modeling_led.py

@@ -1590,7 +1590,7 @@ LED_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.
 """
 
 
@@ -1748,7 +1748,7 @@ class LEDEncoder(LEDPreTrainedModel):
                 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 = (
@@ -2003,7 +2003,7 @@ class LEDDecoder(LEDPreTrainedModel):
                 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/led/modeling_tf_led.py

@@ -1600,8 +1600,8 @@ LED_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).
@@ -1701,7 +1701,7 @@ class TFLEDEncoder(tf.keras.layers.Layer):
                 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.
         """
         inputs = input_processing(
             func=self.call,
@@ -1983,7 +1983,7 @@ class TFLEDDecoder(tf.keras.layers.Layer):
                 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.
         """
         inputs = input_processing(
             func=self.call,

src/transformers/models/longformer/modeling_longformer.py

@@ -1486,7 +1486,7 @@ LONGFORMER_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/longformer/modeling_tf_longformer.py

@@ -1974,8 +1974,8 @@ LONGFORMER_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/luke/modeling_luke.py

@@ -868,7 +868,7 @@ LUKE_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/luke/tokenization_luke.py

@@ -1108,7 +1108,7 @@ class LukeTokenizer(RobertaTokenizer):
                 List[int]]]*) so you can use this method during preprocessing as well as in a PyTorch Dataloader
                 collate function. Instead of `List[int]` you can have tensors (numpy arrays, PyTorch tensors or
                 TensorFlow tensors), see the note above for the return type.
-            padding (`bool`, `str` or [`~file_utils.PaddingStrategy`], *optional*, defaults to `True`):
+            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True`):
                  Select a strategy to pad the returned sequences (according to the model's padding side and padding
                  index) among:
 
@@ -1129,7 +1129,7 @@ class LukeTokenizer(RobertaTokenizer):
                 Whether to return the attention mask. If left to the default, will return the attention mask according
                 to the specific tokenizer's default, defined by the `return_outputs` attribute. [What are attention
                 masks?](../glossary#attention-mask)
-            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/lxmert/modeling_lxmert.py

@@ -875,7 +875,7 @@ LXMERT_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/lxmert/modeling_tf_lxmert.py

@@ -929,8 +929,8 @@ LXMERT_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/m2m_100/modeling_m2m_100.py

@@ -667,7 +667,7 @@ M2M_100_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.
 """
 
 
@@ -753,7 +753,7 @@ class M2M100Encoder(M2M100PreTrainedModel):
                 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 = (
@@ -952,7 +952,7 @@ class M2M100Decoder(M2M100PreTrainedModel):
                 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/marian/modeling_flax_marian.py

@@ -136,7 +136,7 @@ MARIAN_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.
 """
 
 
@@ -167,7 +167,7 @@ MARIAN_ENCODE_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.
 """
 
 MARIAN_DECODE_INPUTS_DOCSTRING = r"""
@@ -213,7 +213,7 @@ MARIAN_DECODE_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/marian/modeling_marian.py

@@ -634,7 +634,7 @@ MARIAN_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.
 """
 
 
@@ -723,7 +723,7 @@ class MarianEncoder(MarianPreTrainedModel):
                 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 = (
@@ -942,7 +942,7 @@ class MarianDecoder(MarianPreTrainedModel):
                 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 = (
@@ -1643,7 +1643,7 @@ class MarianForCausalLM(MarianPreTrainedModel):
                 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.
 
         Returns:
 

src/transformers/models/marian/modeling_tf_marian.py

@@ -654,8 +654,8 @@ MARIAN_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).
@@ -745,8 +745,8 @@ class TFMarianEncoder(tf.keras.layers.Layer):
                 for 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).
@@ -928,8 +928,8 @@ class TFMarianDecoder(tf.keras.layers.Layer):
                 for 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).