Big file_utils cleanup (#16396)

* Big file_utils cleanup

* This one still needs to be treated separately

src/transformers/models/sew_d/modeling_sew_d.py

@@ -1340,7 +1340,7 @@ SEWD_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/speech_encoder_decoder/modeling_flax_speech_encoder_decoder.py

@@ -120,7 +120,7 @@ SPEECH_ENCODER_DECODER_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*):
-            If set to `True`, the model will return a [`~file_utils.FlaxSeq2SeqLMOutput`] instead of a plain tuple.
+            If set to `True`, the model will return a [`~utils.FlaxSeq2SeqLMOutput`] instead of a plain tuple.
 """
 
 SPEECH_ENCODER_DECODER_ENCODE_INPUTS_DOCSTRING = r"""
@@ -145,7 +145,7 @@ SPEECH_ENCODER_DECODER_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*):
-            If set to `True`, the model will return a [`~file_utils.FlaxBaseModelOutput`] instead of a plain tuple.
+            If set to `True`, the model will return a [`~utils.FlaxBaseModelOutput`] instead of a plain tuple.
 """
 
 SPEECH_ENCODER_DECODER_DECODE_INPUTS_DOCSTRING = r"""
@@ -191,8 +191,8 @@ SPEECH_ENCODER_DECODER_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*):
-            If set to `True`, the model will return a [`~file_utils.FlaxCausalLMOutputWithCrossAttentions`] instead of
-            a plain tuple.
+            If set to `True`, the model will return a [`~utils.FlaxCausalLMOutputWithCrossAttentions`] instead of a
+            plain tuple.
 """
 
 

src/transformers/models/speech_encoder_decoder/modeling_speech_encoder_decoder.py

@@ -142,7 +142,7 @@ SPEECH_ENCODER_DECODER_INPUTS_DOCSTRING = r"""
             [`Speech2TextTokenizer`] should be used for extracting the fbank features, padding and conversion into a
             tensor of type `torch.FloatTensor`. See [`~Speech2TextTokenizer.__call__`]
         return_dict (`bool`, *optional*):
-            If set to `True`, the model will return a [`~file_utils.Seq2SeqLMOutput`] instead of a plain tuple.
+            If set to `True`, the model will return a [`~utils.Seq2SeqLMOutput`] instead of a plain tuple.
         kwargs: (*optional*) Remaining dictionary of keyword arguments. Keyword arguments come in two flavors:
 
             - Without a prefix which will be input as `**encoder_kwargs` for the encoder forward function.

src/transformers/models/speech_to_text/feature_extraction_speech_to_text.py

@@ -142,7 +142,7 @@ class Speech2TextFeatureExtractor(SequenceFeatureExtractor):
             raw_speech (`np.ndarray`, `List[float]`, `List[np.ndarray]`, `List[List[float]]`):
                 The sequence or batch of sequences to be padded. Each sequence can be a numpy array, a list of float
                 values, a list of numpy arrays or a list of list of float values.
-            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:
 
@@ -174,7 +174,7 @@ class Speech2TextFeatureExtractor(SequenceFeatureExtractor):
 
                 </Tip>
 
-            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/speech_to_text/modeling_speech_to_text.py

@@ -679,7 +679,7 @@ SPEECH_TO_TEXT_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.
 """
 
 
@@ -757,7 +757,7 @@ class Speech2TextEncoder(Speech2TextPreTrainedModel):
                 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 = (
@@ -974,7 +974,7 @@ class Speech2TextDecoder(Speech2TextPreTrainedModel):
                 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/speech_to_text/modeling_tf_speech_to_text.py

@@ -710,8 +710,8 @@ SPEECH_TO_TEXT_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).
@@ -821,7 +821,7 @@ class TFSpeech2TextEncoder(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.
         """
         if input_features is None:
             raise ValueError("You have to specify input_features")
@@ -1017,7 +1017,7 @@ class TFSpeech2TextDecoder(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.
         """
 
         if input_ids is not None and inputs_embeds is not None:

src/transformers/models/speech_to_text_2/modeling_speech_to_text_2.py

@@ -581,7 +581,7 @@ class Speech2Text2Decoder(Speech2Text2PreTrainedModel):
                 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 = (
@@ -850,7 +850,7 @@ class Speech2Text2ForCausalLM(Speech2Text2PreTrainedModel):
                 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/splinter/modeling_splinter.py

@@ -600,7 +600,7 @@ SPLINTER_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/squeezebert/modeling_squeezebert.py

@@ -538,7 +538,7 @@ SQUEEZEBERT_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/swin/modeling_swin.py

@@ -873,7 +873,7 @@ SWIN_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/t5/modeling_flax_t5.py

@@ -800,7 +800,7 @@ T5_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.
 """
 
 T5_DECODE_INPUTS_DOCSTRING = r"""
@@ -841,7 +841,7 @@ T5_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.
 """
 
 
@@ -899,7 +899,7 @@ T5_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/t5/modeling_t5.py

@@ -1211,7 +1211,7 @@ T5_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.
 """
 
 T5_ENCODER_INPUTS_DOCSTRING = r"""
@@ -1248,7 +1248,7 @@ T5_ENCODER_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.
 """
 
 # Warning message for FutureWarning: head_mask was separated into two input args - head_mask, decoder_head_mask

src/transformers/models/t5/modeling_tf_t5.py

@@ -1047,8 +1047,8 @@ T5_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).
@@ -1088,7 +1088,7 @@ T5_ENCODER_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/tapas/modeling_tapas.py

@@ -845,7 +845,7 @@ TAPAS_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/tapas/modeling_tf_tapas.py

@@ -947,8 +947,8 @@ TAPAS_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/tapas/tokenization_tapas.py

@@ -146,7 +146,7 @@ def whitespace_tokenize(text):
 TAPAS_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
@@ -176,7 +176,7 @@ TAPAS_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/transfo_xl/modeling_tf_transfo_xl.py

@@ -863,8 +863,8 @@ TRANSFO_XL_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/transfo_xl/modeling_transfo_xl.py

@@ -760,7 +760,7 @@ TRANSFO_XL_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/trocr/modeling_trocr.py

@@ -609,7 +609,7 @@ class TrOCRDecoder(TrOCRPreTrainedModel):
                 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 = (
@@ -884,7 +884,7 @@ class TrOCRForCausalLM(TrOCRPreTrainedModel):
                 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/unispeech/modeling_unispeech.py

@@ -1076,7 +1076,7 @@ UNISPEECH_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.
 """