Reorganize file utils (#16264)

* Split file_utils in several submodules

* Fixes

* Add back more objects

* More fixes

* Who exactly decided to import that from there?

* Second suggestion to code with code review

* Revert wront move

* Fix imports

* Adapt all imports

* Adapt all imports everywhere

* Revert this import, will fix in a separate commit

src/transformers/models/ctrl/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_tf_available, is_torch_available
+from ...utils import _LazyModule, is_tf_available, is_torch_available
 
 
 _import_structure = {

src/transformers/models/ctrl/modeling_ctrl.py

@@ -22,10 +22,9 @@ import torch
 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast, SequenceClassifierOutput
 from ...modeling_utils import Conv1D, PreTrainedModel, find_pruneable_heads_and_indices, prune_linear_layer
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_ctrl import CTRLConfig
 
 

src/transformers/models/ctrl/modeling_tf_ctrl.py

@@ -21,7 +21,6 @@ from typing import Tuple
 import numpy as np
 import tensorflow as tf
 
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_tf_outputs import TFBaseModelOutputWithPast, TFCausalLMOutputWithPast, TFSequenceClassifierOutput
 from ...modeling_tf_utils import (
     TFCausalLanguageModelingLoss,
@@ -33,7 +32,7 @@ from ...modeling_tf_utils import (
     unpack_inputs,
 )
 from ...tf_utils import shape_list
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_ctrl import CTRLConfig
 
 

src/transformers/models/data2vec/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_torch_available
+from ...utils import _LazyModule, is_torch_available
 
 
 _import_structure = {

src/transformers/models/data2vec/modeling_data2vec_audio.py

@@ -27,16 +27,16 @@ from torch.nn import CrossEntropyLoss
 
 from ...activations import ACT2FN
 from ...deepspeed import is_deepspeed_zero3_enabled
-from ...file_utils import (
+from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput, TokenClassifierOutput
+from ...modeling_utils import PreTrainedModel
+from ...pytorch_utils import torch_int_div
+from ...utils import (
     ModelOutput,
     add_code_sample_docstrings,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
+    logging,
 )
-from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput, TokenClassifierOutput
-from ...modeling_utils import PreTrainedModel
-from ...pytorch_utils import torch_int_div
-from ...utils import logging
 from .configuration_data2vec_audio import Data2VecAudioConfig
 
 

src/transformers/models/data2vec/modeling_data2vec_text.py

@@ -24,12 +24,6 @@ from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
 from ...activations import ACT2FN, gelu
-from ...file_utils import (
-    add_code_sample_docstrings,
-    add_start_docstrings,
-    add_start_docstrings_to_model_forward,
-    replace_return_docstrings,
-)
 from ...modeling_outputs import (
     BaseModelOutputWithPastAndCrossAttentions,
     BaseModelOutputWithPoolingAndCrossAttentions,
@@ -46,7 +40,13 @@ from ...modeling_utils import (
     find_pruneable_heads_and_indices,
     prune_linear_layer,
 )
-from ...utils import logging
+from ...utils import (
+    add_code_sample_docstrings,
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    logging,
+    replace_return_docstrings,
+)
 from .configuration_data2vec_text import Data2VecTextConfig
 
 

src/transformers/models/deberta/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_tf_available, is_tokenizers_available, is_torch_available
+from ...utils import _LazyModule, is_tf_available, is_tokenizers_available, is_torch_available
 
 
 _import_structure = {

src/transformers/models/deberta/modeling_deberta.py

@@ -23,7 +23,6 @@ from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
 from ...activations import ACT2FN
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_outputs import (
     BaseModelOutput,
     MaskedLMOutput,
@@ -33,7 +32,7 @@ from ...modeling_outputs import (
 )
 from ...modeling_utils import PreTrainedModel
 from ...pytorch_utils import softmax_backward_data
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_deberta import DebertaConfig
 
 

src/transformers/models/deberta/modeling_tf_deberta.py

@@ -22,7 +22,6 @@ import numpy as np
 import tensorflow as tf
 
 from ...activations_tf import get_tf_activation
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_tf_outputs import (
     TFBaseModelOutput,
     TFMaskedLMOutput,
@@ -41,7 +40,7 @@ from ...modeling_tf_utils import (
     unpack_inputs,
 )
 from ...tf_utils import shape_list
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_deberta import DebertaConfig
 
 

src/transformers/models/deberta_v2/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_tf_available, is_torch_available
+from ...utils import _LazyModule, is_tf_available, is_torch_available
 
 
 _import_structure = {

src/transformers/models/deberta_v2/modeling_deberta_v2.py

@@ -24,7 +24,6 @@ from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, LayerNorm, MSELoss
 
 from ...activations import ACT2FN
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_outputs import (
     BaseModelOutput,
     MaskedLMOutput,
@@ -34,7 +33,7 @@ from ...modeling_outputs import (
 )
 from ...modeling_utils import PreTrainedModel
 from ...pytorch_utils import softmax_backward_data
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_deberta_v2 import DebertaV2Config
 
 

src/transformers/models/deberta_v2/modeling_tf_deberta_v2.py

@@ -21,7 +21,6 @@ import numpy as np
 import tensorflow as tf
 
 from ...activations_tf import get_tf_activation
-from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
 from ...modeling_tf_outputs import (
     TFBaseModelOutput,
     TFMaskedLMOutput,
@@ -40,7 +39,7 @@ from ...modeling_tf_utils import (
     unpack_inputs,
 )
 from ...tf_utils import shape_list
-from ...utils import logging
+from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
 from .configuration_deberta_v2 import DebertaV2Config
 
 

src/transformers/models/deit/__init__.py

@@ -17,7 +17,7 @@
 # limitations under the License.
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_torch_available, is_vision_available
+from ...utils import _LazyModule, is_torch_available, is_vision_available
 
 
 _import_structure = {

src/transformers/models/deit/feature_extraction_deit.py

@@ -20,7 +20,6 @@ import numpy as np
 from PIL import Image
 
 from ...feature_extraction_utils import BatchFeature, FeatureExtractionMixin
-from ...file_utils import TensorType
 from ...image_utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
@@ -28,7 +27,7 @@ from ...image_utils import (
     ImageInput,
     is_torch_tensor,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)

src/transformers/models/deit/modeling_deit.py

@@ -26,16 +26,16 @@ from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
 from ...activations import ACT2FN
-from ...file_utils import (
+from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithPooling, MaskedLMOutput, SequenceClassifierOutput
+from ...modeling_utils import PreTrainedModel, find_pruneable_heads_and_indices, prune_linear_layer
+from ...utils import (
     ModelOutput,
     add_code_sample_docstrings,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
+    logging,
     replace_return_docstrings,
 )
-from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithPooling, MaskedLMOutput, SequenceClassifierOutput
-from ...modeling_utils import PreTrainedModel, find_pruneable_heads_and_indices, prune_linear_layer
-from ...utils import logging
 from .configuration_deit import DeiTConfig
 
 

src/transformers/models/detr/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_timm_available, is_vision_available
+from ...utils import _LazyModule, is_timm_available, is_vision_available
 
 
 _import_structure = {

src/transformers/models/detr/feature_extraction_detr.py

@@ -23,9 +23,8 @@ import numpy as np
 from PIL import Image
 
 from ...feature_extraction_utils import BatchFeature, FeatureExtractionMixin
-from ...file_utils import TensorType, is_torch_available
 from ...image_utils import ImageFeatureExtractionMixin, is_torch_tensor
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, logging
 
 
 if is_torch_available():

src/transformers/models/detr/modeling_detr.py

@@ -24,20 +24,20 @@ import torch
 from torch import Tensor, nn
 
 from ...activations import ACT2FN
-from ...file_utils import (
+from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithCrossAttentions, Seq2SeqModelOutput
+from ...modeling_utils import PreTrainedModel
+from ...pytorch_utils import torch_int_div
+from ...utils import (
     ModelOutput,
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
     is_scipy_available,
     is_timm_available,
     is_vision_available,
+    logging,
     replace_return_docstrings,
     requires_backends,
 )
-from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithCrossAttentions, Seq2SeqModelOutput
-from ...modeling_utils import PreTrainedModel
-from ...pytorch_utils import torch_int_div
-from ...utils import logging
 from .configuration_detr import DetrConfig
 
 

src/transformers/models/dialogpt/convert_dialogpt_original_pytorch_checkpoint_to_pytorch.py

@@ -17,7 +17,7 @@ import os
 
 import torch
 
-from transformers.file_utils import WEIGHTS_NAME
+from transformers.utils import WEIGHTS_NAME
 
 
 DIALOGPT_MODELS = ["small", "medium", "large"]

src/transformers/models/distilbert/__init__.py

@@ -18,7 +18,7 @@
 
 from typing import TYPE_CHECKING
 
-from ...file_utils import _LazyModule, is_flax_available, is_tf_available, is_tokenizers_available, is_torch_available
+from ...utils import _LazyModule, is_flax_available, is_tf_available, is_tokenizers_available, is_torch_available
 
 
 _import_structure = {