diff --git a/src/transformers/commands/add_new_model_like.py b/src/transformers/commands/add_new_model_like.py
index 5dd5e7dcb82..30d273bec50 100644
--- a/src/transformers/commands/add_new_model_like.py
+++ b/src/transformers/commands/add_new_model_like.py
@@ -23,9 +23,8 @@ from itertools import chain
 from pathlib import Path
 from typing import Any, Callable, Dict, List, Optional, Pattern, Tuple, Union
 
-import transformers.models.auto as auto_module
-from transformers.models.auto.configuration_auto import model_type_to_module_name
-
+from ..models import auto as auto_module
+from ..models.auto.configuration_auto import model_type_to_module_name
 from ..utils import is_flax_available, is_tf_available, is_torch_available, logging
 from . import BaseTransformersCLICommand
 
diff --git a/src/transformers/image_transforms.py b/src/transformers/image_transforms.py
index d09f29b7904..0ae19c43c74 100644
--- a/src/transformers/image_transforms.py
+++ b/src/transformers/image_transforms.py
@@ -18,7 +18,7 @@ from typing import Iterable, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.image_utils import (
+from .image_utils import (
     ChannelDimension,
     ImageInput,
     get_channel_dimension_axis,
@@ -26,8 +26,8 @@ from transformers.image_utils import (
     infer_channel_dimension_format,
     to_numpy_array,
 )
-from transformers.utils import ExplicitEnum, TensorType, is_jax_tensor, is_tf_tensor, is_torch_tensor
-from transformers.utils.import_utils import (
+from .utils import ExplicitEnum, TensorType, is_jax_tensor, is_tf_tensor, is_torch_tensor
+from .utils.import_utils import (
     is_flax_available,
     is_tf_available,
     is_torch_available,
diff --git a/src/transformers/modeling_tf_utils.py b/src/transformers/modeling_tf_utils.py
index 5c73a387d2b..29c4d67510c 100644
--- a/src/transformers/modeling_tf_utils.py
+++ b/src/transformers/modeling_tf_utils.py
@@ -33,8 +33,6 @@ import tensorflow as tf
 from huggingface_hub import Repository, list_repo_files
 from packaging.version import parse
 
-from transformers.utils.hub import convert_file_size_to_int, get_checkpoint_shard_files
-
 from . import DataCollatorWithPadding, DefaultDataCollator
 from .activations_tf import get_tf_activation
 from .configuration_utils import PretrainedConfig
@@ -63,6 +61,7 @@ from .utils import (
     requires_backends,
     working_or_temp_dir,
 )
+from .utils.hub import convert_file_size_to_int, get_checkpoint_shard_files
 
 
 if parse(tf.__version__) >= parse("2.11.0"):
diff --git a/src/transformers/modeling_utils.py b/src/transformers/modeling_utils.py
index 73e6cf00ef8..e7b0aab459f 100644
--- a/src/transformers/modeling_utils.py
+++ b/src/transformers/modeling_utils.py
@@ -32,9 +32,6 @@ from packaging import version
 from torch import Tensor, nn
 from torch.nn import CrossEntropyLoss
 
-from transformers.utils.hub import convert_file_size_to_int, get_checkpoint_shard_files
-from transformers.utils.import_utils import ENV_VARS_TRUE_VALUES, is_sagemaker_mp_enabled
-
 from .activations import get_activation
 from .configuration_utils import PretrainedConfig
 from .deepspeed import deepspeed_config, is_deepspeed_zero3_enabled
@@ -73,7 +70,8 @@ from .utils import (
     logging,
     replace_return_docstrings,
 )
-from .utils.import_utils import importlib_metadata
+from .utils.hub import convert_file_size_to_int, get_checkpoint_shard_files
+from .utils.import_utils import ENV_VARS_TRUE_VALUES, importlib_metadata, is_sagemaker_mp_enabled
 from .utils.quantization_config import BitsAndBytesConfig
 from .utils.versions import require_version_core
 
diff --git a/src/transformers/models/albert/convert_albert_original_tf_checkpoint_to_pytorch.py b/src/transformers/models/albert/convert_albert_original_tf_checkpoint_to_pytorch.py
index 8823a86fc8c..eecada8b432 100644
--- a/src/transformers/models/albert/convert_albert_original_tf_checkpoint_to_pytorch.py
+++ b/src/transformers/models/albert/convert_albert_original_tf_checkpoint_to_pytorch.py
@@ -19,8 +19,8 @@ import argparse
 
 import torch
 
-from transformers import AlbertConfig, AlbertForPreTraining, load_tf_weights_in_albert
-from transformers.utils import logging
+from ...utils import logging
+from . import AlbertConfig, AlbertForPreTraining, load_tf_weights_in_albert
 
 
 logging.set_verbosity_info()
diff --git a/src/transformers/models/beit/image_processing_beit.py b/src/transformers/models/beit/image_processing_beit.py
index 6f73679a71b..4fafc5fda66 100644
--- a/src/transformers/models/beit/image_processing_beit.py
+++ b/src/transformers/models/beit/image_processing_beit.py
@@ -19,9 +19,6 @@ from typing import Any, Dict, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_torch_available, is_torch_tensor, is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -34,7 +31,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, is_torch_tensor, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/bit/image_processing_bit.py b/src/transformers/models/bit/image_processing_bit.py
index 0394ecb411e..66b308c1bcd 100644
--- a/src/transformers/models/bit/image_processing_bit.py
+++ b/src/transformers/models/bit/image_processing_bit.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,8 +38,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
-from ...utils.import_utils import is_vision_available
+from ...utils import TensorType, is_vision_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/blip/image_processing_blip.py b/src/transformers/models/blip/image_processing_blip.py
index 539d6d19860..59ea4ac7798 100644
--- a/src/transformers/models/blip/image_processing_blip.py
+++ b/src/transformers/models/blip/image_processing_blip.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import convert_to_rgb, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -33,7 +30,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/blip/modeling_blip_text.py b/src/transformers/models/blip/modeling_blip_text.py
index 3ec4a994d78..356d4e52719 100644
--- a/src/transformers/models/blip/modeling_blip_text.py
+++ b/src/transformers/models/blip/modeling_blip_text.py
@@ -22,20 +22,19 @@ import torch.utils.checkpoint
 from torch import Tensor, device, nn
 from torch.nn import CrossEntropyLoss
 
-from transformers.activations import ACT2FN
-from transformers.modeling_outputs import (
+from ...activations import ACT2FN
+from ...modeling_outputs import (
     BaseModelOutputWithPastAndCrossAttentions,
     BaseModelOutputWithPoolingAndCrossAttentions,
     CausalLMOutputWithCrossAttentions,
 )
-from transformers.modeling_utils import (
+from ...modeling_utils import (
     PreTrainedModel,
     apply_chunking_to_forward,
     find_pruneable_heads_and_indices,
     prune_linear_layer,
 )
-from transformers.utils import logging
-
+from ...utils import logging
 from .configuration_blip import BlipTextConfig
 
 
diff --git a/src/transformers/models/blip_2/configuration_blip_2.py b/src/transformers/models/blip_2/configuration_blip_2.py
index 9db098adf1a..8a80510db9a 100644
--- a/src/transformers/models/blip_2/configuration_blip_2.py
+++ b/src/transformers/models/blip_2/configuration_blip_2.py
@@ -18,9 +18,8 @@ import copy
 import os
 from typing import Union
 
-from transformers.models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
-
 from ...configuration_utils import PretrainedConfig
+from ...models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
 from ...utils import logging
 from ..auto import CONFIG_MAPPING
 
diff --git a/src/transformers/models/bloom/configuration_bloom.py b/src/transformers/models/bloom/configuration_bloom.py
index f2ea93c1168..17395625e01 100644
--- a/src/transformers/models/bloom/configuration_bloom.py
+++ b/src/transformers/models/bloom/configuration_bloom.py
@@ -18,15 +18,13 @@ from typing import TYPE_CHECKING, Any, List, Mapping, Optional
 
 from packaging import version
 
-from transformers import is_torch_available
-
 
 if TYPE_CHECKING:
-    from transformers import PreTrainedTokenizer, TensorType
+    from ... import PreTrainedTokenizer, TensorType
 
 from ...configuration_utils import PretrainedConfig
 from ...onnx import OnnxConfigWithPast, PatchingSpec
-from ...utils import logging
+from ...utils import is_torch_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/bridgetower/image_processing_bridgetower.py b/src/transformers/models/bridgetower/image_processing_bridgetower.py
index 6cf988114e2..76e0876c76c 100644
--- a/src/transformers/models/bridgetower/image_processing_bridgetower.py
+++ b/src/transformers/models/bridgetower/image_processing_bridgetower.py
@@ -19,9 +19,6 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import PaddingMode, center_crop, normalize, pad, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -36,7 +33,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/chinese_clip/image_processing_chinese_clip.py b/src/transformers/models/chinese_clip/image_processing_chinese_clip.py
index a21372b7533..5aa4eafbbba 100644
--- a/src/transformers/models/chinese_clip/image_processing_chinese_clip.py
+++ b/src/transformers/models/chinese_clip/image_processing_chinese_clip.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,8 +38,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
-from ...utils.import_utils import is_vision_available
+from ...utils import TensorType, is_vision_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/clip/image_processing_clip.py b/src/transformers/models/clip/image_processing_clip.py
index b06e121758f..f9b5f3edde7 100644
--- a/src/transformers/models/clip/image_processing_clip.py
+++ b/src/transformers/models/clip/image_processing_clip.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,8 +38,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
-from ...utils.import_utils import is_vision_available
+from ...utils import TensorType, is_vision_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/conditional_detr/image_processing_conditional_detr.py b/src/transformers/models/conditional_detr/image_processing_conditional_detr.py
index 2a70e45edf0..cc4a24cb05d 100644
--- a/src/transformers/models/conditional_detr/image_processing_conditional_detr.py
+++ b/src/transformers/models/conditional_detr/image_processing_conditional_detr.py
@@ -22,9 +22,9 @@ from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Un
 
 import numpy as np
 
-from transformers.feature_extraction_utils import BatchFeature
-from transformers.image_processing_utils import BaseImageProcessor, get_size_dict
-from transformers.image_transforms import (
+from ...feature_extraction_utils import BatchFeature
+from ...image_processing_utils import BaseImageProcessor, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     center_to_corners_format,
     corners_to_center_format,
@@ -36,7 +36,7 @@ from transformers.image_transforms import (
     rgb_to_id,
     to_channel_dimension_format,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     ChannelDimension,
@@ -50,7 +50,9 @@ from transformers.image_utils import (
     valid_coco_panoptic_annotations,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
+    ExplicitEnum,
+    TensorType,
     is_flax_available,
     is_jax_tensor,
     is_scipy_available,
@@ -60,7 +62,6 @@ from transformers.utils import (
     is_torch_tensor,
     is_vision_available,
 )
-from transformers.utils.generic import ExplicitEnum, TensorType
 
 
 if is_torch_available():
diff --git a/src/transformers/models/conditional_detr/modeling_conditional_detr.py b/src/transformers/models/conditional_detr/modeling_conditional_detr.py
index 36e6e942158..eb67b1e25bd 100644
--- a/src/transformers/models/conditional_detr/modeling_conditional_detr.py
+++ b/src/transformers/models/conditional_detr/modeling_conditional_detr.py
@@ -49,7 +49,7 @@ if is_timm_available():
     from timm import create_model
 
 if is_vision_available():
-    from transformers.image_transforms import center_to_corners_format
+    from ...image_transforms import center_to_corners_format
 
 logger = logging.get_logger(__name__)
 
diff --git a/src/transformers/models/convnext/image_processing_convnext.py b/src/transformers/models/convnext/image_processing_convnext.py
index a46bdcfef75..6d1892ef87f 100644
--- a/src/transformers/models/convnext/image_processing_convnext.py
+++ b/src/transformers/models/convnext/image_processing_convnext.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/convnext/modeling_tf_convnext.py b/src/transformers/models/convnext/modeling_tf_convnext.py
index 671de6ca9b4..00db1f0b788 100644
--- a/src/transformers/models/convnext/modeling_tf_convnext.py
+++ b/src/transformers/models/convnext/modeling_tf_convnext.py
@@ -20,8 +20,6 @@ from typing import Dict, Optional, Tuple, Union
 import numpy as np
 import tensorflow as tf
 
-from transformers import shape_list
-
 from ...activations_tf import get_tf_activation
 from ...modeling_tf_outputs import TFBaseModelOutput, TFBaseModelOutputWithPooling, TFSequenceClassifierOutput
 from ...modeling_tf_utils import (
@@ -32,6 +30,7 @@ from ...modeling_tf_utils import (
     keras_serializable,
     unpack_inputs,
 )
+from ...tf_utils import shape_list
 from ...utils import add_start_docstrings, add_start_docstrings_to_model_forward, logging, replace_return_docstrings
 from .configuration_convnext import ConvNextConfig
 
diff --git a/src/transformers/models/data2vec/modeling_tf_data2vec_vision.py b/src/transformers/models/data2vec/modeling_tf_data2vec_vision.py
index 8bf8a885504..eca13d433f2 100644
--- a/src/transformers/models/data2vec/modeling_tf_data2vec_vision.py
+++ b/src/transformers/models/data2vec/modeling_tf_data2vec_vision.py
@@ -22,8 +22,6 @@ from typing import Dict, List, Optional, Tuple, Union
 import numpy as np
 import tensorflow as tf
 
-from transformers.tf_utils import shape_list, stable_softmax
-
 from ...activations_tf import get_tf_activation
 from ...modeling_tf_outputs import (
     TFBaseModelOutput,
@@ -39,6 +37,7 @@ from ...modeling_tf_utils import (
     keras_serializable,
     unpack_inputs,
 )
+from ...tf_utils import shape_list, stable_softmax
 from ...utils import (
     add_code_sample_docstrings,
     add_start_docstrings,
diff --git a/src/transformers/models/deformable_detr/image_processing_deformable_detr.py b/src/transformers/models/deformable_detr/image_processing_deformable_detr.py
index 4db2e27647a..87e6ef508c6 100644
--- a/src/transformers/models/deformable_detr/image_processing_deformable_detr.py
+++ b/src/transformers/models/deformable_detr/image_processing_deformable_detr.py
@@ -22,9 +22,9 @@ from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Un
 
 import numpy as np
 
-from transformers.feature_extraction_utils import BatchFeature
-from transformers.image_processing_utils import BaseImageProcessor, get_size_dict
-from transformers.image_transforms import (
+from ...feature_extraction_utils import BatchFeature
+from ...image_processing_utils import BaseImageProcessor, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     center_to_corners_format,
     corners_to_center_format,
@@ -36,7 +36,7 @@ from transformers.image_transforms import (
     rgb_to_id,
     to_channel_dimension_format,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     ChannelDimension,
@@ -50,7 +50,9 @@ from transformers.image_utils import (
     valid_coco_panoptic_annotations,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
+    ExplicitEnum,
+    TensorType,
     is_flax_available,
     is_jax_tensor,
     is_scipy_available,
@@ -60,7 +62,6 @@ from transformers.utils import (
     is_torch_tensor,
     is_vision_available,
 )
-from transformers.utils.generic import ExplicitEnum, TensorType
 
 
 if is_torch_available():
diff --git a/src/transformers/models/deit/image_processing_deit.py b/src/transformers/models/deit/image_processing_deit.py
index 77d7d2bb2ca..3f9d5ee2d0a 100644
--- a/src/transformers/models/deit/image_processing_deit.py
+++ b/src/transformers/models/deit/image_processing_deit.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -33,7 +30,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/detr/image_processing_detr.py b/src/transformers/models/detr/image_processing_detr.py
index 75132b9a2f3..71e80e8c748 100644
--- a/src/transformers/models/detr/image_processing_detr.py
+++ b/src/transformers/models/detr/image_processing_detr.py
@@ -22,8 +22,8 @@ from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Un
 
 import numpy as np
 
-from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
-from transformers.image_transforms import (
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     center_to_corners_format,
     corners_to_center_format,
@@ -35,7 +35,7 @@ from transformers.image_transforms import (
     rgb_to_id,
     to_channel_dimension_format,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     ChannelDimension,
@@ -49,7 +49,9 @@ from transformers.image_utils import (
     valid_coco_panoptic_annotations,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
+    ExplicitEnum,
+    TensorType,
     is_flax_available,
     is_jax_tensor,
     is_scipy_available,
@@ -59,7 +61,6 @@ from transformers.utils import (
     is_torch_tensor,
     is_vision_available,
 )
-from transformers.utils.generic import ExplicitEnum, TensorType
 
 
 if is_torch_available():
diff --git a/src/transformers/models/distilbert/modeling_distilbert.py b/src/transformers/models/distilbert/modeling_distilbert.py
index 7dbf95b0692..84db89e0f4b 100755
--- a/src/transformers/models/distilbert/modeling_distilbert.py
+++ b/src/transformers/models/distilbert/modeling_distilbert.py
@@ -27,9 +27,8 @@ import torch
 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
-from transformers.configuration_utils import PretrainedConfig
-
 from ...activations import get_activation
+from ...configuration_utils import PretrainedConfig
 from ...deepspeed import is_deepspeed_zero3_enabled
 from ...modeling_outputs import (
     BaseModelOutput,
diff --git a/src/transformers/models/dpt/image_processing_dpt.py b/src/transformers/models/dpt/image_processing_dpt.py
index d6bcfe9c5e3..4f5dbc44507 100644
--- a/src/transformers/models/dpt/image_processing_dpt.py
+++ b/src/transformers/models/dpt/image_processing_dpt.py
@@ -19,9 +19,6 @@ from typing import Dict, Iterable, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -37,7 +34,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_torch_available():
diff --git a/src/transformers/models/efficientformer/image_processing_efficientformer.py b/src/transformers/models/efficientformer/image_processing_efficientformer.py
index 5694fb166e3..81e3d798d11 100644
--- a/src/transformers/models/efficientformer/image_processing_efficientformer.py
+++ b/src/transformers/models/efficientformer/image_processing_efficientformer.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -39,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/efficientnet/image_processing_efficientnet.py b/src/transformers/models/efficientnet/image_processing_efficientnet.py
index 0769fb820e2..f4d2a88ee4d 100644
--- a/src/transformers/models/efficientnet/image_processing_efficientnet.py
+++ b/src/transformers/models/efficientnet/image_processing_efficientnet.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -33,7 +30,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/flava/image_processing_flava.py b/src/transformers/models/flava/image_processing_flava.py
index 433f09f6ad8..3b5755d9b4d 100644
--- a/src/transformers/models/flava/image_processing_flava.py
+++ b/src/transformers/models/flava/image_processing_flava.py
@@ -21,9 +21,6 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -36,7 +33,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/flava/modeling_flava.py b/src/transformers/models/flava/modeling_flava.py
index 720e0659725..6681de3b322 100644
--- a/src/transformers/models/flava/modeling_flava.py
+++ b/src/transformers/models/flava/modeling_flava.py
@@ -24,13 +24,12 @@ import torch
 import torch.utils.checkpoint
 from torch import nn
 
-from transformers.utils.doc import add_code_sample_docstrings
-
 from ...activations import ACT2FN
 from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithPooling
 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,
diff --git a/src/transformers/models/glpn/image_processing_glpn.py b/src/transformers/models/glpn/image_processing_glpn.py
index ba279c3c4ff..7b0f316d2b5 100644
--- a/src/transformers/models/glpn/image_processing_glpn.py
+++ b/src/transformers/models/glpn/image_processing_glpn.py
@@ -19,13 +19,17 @@ from typing import List, Optional, Union
 import numpy as np
 import PIL.Image
 
-from transformers.image_utils import PILImageResampling
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature
 from ...image_transforms import rescale, resize, to_channel_dimension_format
-from ...image_utils import ChannelDimension, get_image_size, make_list_of_images, to_numpy_array, valid_images
-from ...utils import logging
+from ...image_utils import (
+    ChannelDimension,
+    PILImageResampling,
+    get_image_size,
+    make_list_of_images,
+    to_numpy_array,
+    valid_images,
+)
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/gpt2/configuration_gpt2.py b/src/transformers/models/gpt2/configuration_gpt2.py
index 4ad37c5fdf0..f1e2d0ae150 100644
--- a/src/transformers/models/gpt2/configuration_gpt2.py
+++ b/src/transformers/models/gpt2/configuration_gpt2.py
@@ -17,8 +17,7 @@
 from collections import OrderedDict
 from typing import Any, List, Mapping, Optional
 
-from transformers import PreTrainedTokenizer, TensorType, is_torch_available
-
+from ... import PreTrainedTokenizer, TensorType, is_torch_available
 from ...configuration_utils import PretrainedConfig
 from ...onnx import OnnxConfigWithPast, PatchingSpec
 from ...utils import logging
diff --git a/src/transformers/models/hubert/modeling_hubert.py b/src/transformers/models/hubert/modeling_hubert.py
index dc88f9359d4..cb51d0479f8 100755
--- a/src/transformers/models/hubert/modeling_hubert.py
+++ b/src/transformers/models/hubert/modeling_hubert.py
@@ -23,9 +23,8 @@ import torch.utils.checkpoint
 from torch import nn
 from torch.nn import CrossEntropyLoss
 
-from transformers.deepspeed import is_deepspeed_zero3_enabled
-
 from ...activations import ACT2FN
+from ...deepspeed import is_deepspeed_zero3_enabled
 from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput
 from ...modeling_utils import PreTrainedModel
 from ...pytorch_utils import torch_int_div
diff --git a/src/transformers/models/ibert/configuration_ibert.py b/src/transformers/models/ibert/configuration_ibert.py
index fe46e3fca61..249061ceae3 100644
--- a/src/transformers/models/ibert/configuration_ibert.py
+++ b/src/transformers/models/ibert/configuration_ibert.py
@@ -18,9 +18,8 @@
 from collections import OrderedDict
 from typing import Mapping
 
-from transformers.onnx import OnnxConfig
-
 from ...configuration_utils import PretrainedConfig
+from ...onnx import OnnxConfig
 from ...utils import logging
 
 
diff --git a/src/transformers/models/imagegpt/image_processing_imagegpt.py b/src/transformers/models/imagegpt/image_processing_imagegpt.py
index 451b4a265fe..9cfa40078bd 100644
--- a/src/transformers/models/imagegpt/image_processing_imagegpt.py
+++ b/src/transformers/models/imagegpt/image_processing_imagegpt.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -31,7 +28,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/jukebox/tokenization_jukebox.py b/src/transformers/models/jukebox/tokenization_jukebox.py
index bd4d6721daf..63399adf162 100644
--- a/src/transformers/models/jukebox/tokenization_jukebox.py
+++ b/src/transformers/models/jukebox/tokenization_jukebox.py
@@ -25,11 +25,10 @@ from typing import Any, Dict, List, Optional, Tuple, Union
 import numpy as np
 import regex
 
-from transformers.utils.generic import _is_jax, _is_numpy
-
 from ...tokenization_utils import AddedToken, PreTrainedTokenizer
 from ...tokenization_utils_base import BatchEncoding
 from ...utils import TensorType, is_flax_available, is_tf_available, is_torch_available, logging
+from ...utils.generic import _is_jax, _is_numpy
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/layoutlm/configuration_layoutlm.py b/src/transformers/models/layoutlm/configuration_layoutlm.py
index f297adea142..c6fc01b7650 100644
--- a/src/transformers/models/layoutlm/configuration_layoutlm.py
+++ b/src/transformers/models/layoutlm/configuration_layoutlm.py
@@ -16,11 +16,9 @@
 from collections import OrderedDict
 from typing import Any, List, Mapping, Optional
 
-from transformers import PretrainedConfig, PreTrainedTokenizer, TensorType
-
-from ... import is_torch_available
+from ... import PretrainedConfig, PreTrainedTokenizer
 from ...onnx import OnnxConfig, PatchingSpec
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/layoutlmv2/image_processing_layoutlmv2.py b/src/transformers/models/layoutlmv2/image_processing_layoutlmv2.py
index ca01b3670d0..6ad0968b612 100644
--- a/src/transformers/models/layoutlmv2/image_processing_layoutlmv2.py
+++ b/src/transformers/models/layoutlmv2/image_processing_layoutlmv2.py
@@ -18,9 +18,6 @@ from typing import Dict, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import resize, to_channel_dimension_format, to_pil_image
 from ...image_utils import (
@@ -32,7 +29,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import is_pytesseract_available, logging, requires_backends
+from ...utils import TensorType, is_pytesseract_available, is_vision_available, logging, requires_backends
 
 
 if is_vision_available():
diff --git a/src/transformers/models/layoutlmv3/image_processing_layoutlmv3.py b/src/transformers/models/layoutlmv3/image_processing_layoutlmv3.py
index aec81de1a99..7152abf06c4 100644
--- a/src/transformers/models/layoutlmv3/image_processing_layoutlmv3.py
+++ b/src/transformers/models/layoutlmv3/image_processing_layoutlmv3.py
@@ -18,9 +18,6 @@ from typing import Dict, Iterable, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import normalize, rescale, resize, to_channel_dimension_format, to_pil_image
 from ...image_utils import (
@@ -34,7 +31,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import is_pytesseract_available, logging, requires_backends
+from ...utils import TensorType, is_pytesseract_available, is_vision_available, logging, requires_backends
 
 
 if is_vision_available():
diff --git a/src/transformers/models/layoutlmv3/modeling_layoutlmv3.py b/src/transformers/models/layoutlmv3/modeling_layoutlmv3.py
index 56e4fa1e68a..db6618caaea 100644
--- a/src/transformers/models/layoutlmv3/modeling_layoutlmv3.py
+++ b/src/transformers/models/layoutlmv3/modeling_layoutlmv3.py
@@ -24,18 +24,16 @@ import torch.nn.functional as F
 import torch.utils.checkpoint
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
-from transformers import apply_chunking_to_forward
-from transformers.modeling_outputs import (
+from ...activations import ACT2FN
+from ...modeling_outputs import (
     BaseModelOutput,
     QuestionAnsweringModelOutput,
     SequenceClassifierOutput,
     TokenClassifierOutput,
 )
-from transformers.modeling_utils import PreTrainedModel
-from transformers.utils import logging
-
-from ...activations import ACT2FN
-from ...utils import add_start_docstrings, add_start_docstrings_to_model_forward, replace_return_docstrings
+from ...modeling_utils import PreTrainedModel
+from ...pytorch_utils import apply_chunking_to_forward
+from ...utils import add_start_docstrings, add_start_docstrings_to_model_forward, logging, replace_return_docstrings
 from .configuration_layoutlmv3 import LayoutLMv3Config
 
 
diff --git a/src/transformers/models/levit/image_processing_levit.py b/src/transformers/models/levit/image_processing_levit.py
index 6aef221a16a..6e142729418 100644
--- a/src/transformers/models/levit/image_processing_levit.py
+++ b/src/transformers/models/levit/image_processing_levit.py
@@ -18,8 +18,6 @@ from typing import Dict, Iterable, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -39,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/lxmert/modeling_tf_lxmert.py b/src/transformers/models/lxmert/modeling_tf_lxmert.py
index 5510c7ff845..9408188100f 100644
--- a/src/transformers/models/lxmert/modeling_tf_lxmert.py
+++ b/src/transformers/models/lxmert/modeling_tf_lxmert.py
@@ -23,8 +23,6 @@ from typing import Dict, Optional, Tuple, Union
 import numpy as np
 import tensorflow as tf
 
-from transformers.tf_utils import stable_softmax
-
 from ...activations_tf import get_tf_activation
 from ...modeling_tf_utils import (
     TFModelInputType,
@@ -34,6 +32,7 @@ from ...modeling_tf_utils import (
     shape_list,
     unpack_inputs,
 )
+from ...tf_utils import stable_softmax
 from ...utils import (
     ModelOutput,
     add_code_sample_docstrings,
diff --git a/src/transformers/models/markuplm/configuration_markuplm.py b/src/transformers/models/markuplm/configuration_markuplm.py
index 935625b1fa1..1455150598a 100644
--- a/src/transformers/models/markuplm/configuration_markuplm.py
+++ b/src/transformers/models/markuplm/configuration_markuplm.py
@@ -14,9 +14,8 @@
 # limitations under the License.
 """ MarkupLM model configuration"""
 
-from transformers.utils import logging
-
 from ...configuration_utils import PretrainedConfig
+from ...utils import logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/markuplm/modeling_markuplm.py b/src/transformers/models/markuplm/modeling_markuplm.py
index 205e6aefc12..18ad4d6978e 100755
--- a/src/transformers/models/markuplm/modeling_markuplm.py
+++ b/src/transformers/models/markuplm/modeling_markuplm.py
@@ -23,13 +23,13 @@ import torch.utils.checkpoint
 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
-from transformers.activations import ACT2FN
-from transformers.file_utils import (
+from ...activations import ACT2FN
+from ...file_utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
     replace_return_docstrings,
 )
-from transformers.modeling_outputs import (
+from ...modeling_outputs import (
     BaseModelOutputWithPastAndCrossAttentions,
     BaseModelOutputWithPoolingAndCrossAttentions,
     MaskedLMOutput,
@@ -37,14 +37,13 @@ from transformers.modeling_outputs import (
     SequenceClassifierOutput,
     TokenClassifierOutput,
 )
-from transformers.modeling_utils import (
+from ...modeling_utils import (
     PreTrainedModel,
     apply_chunking_to_forward,
     find_pruneable_heads_and_indices,
     prune_linear_layer,
 )
-from transformers.utils import logging
-
+from ...utils import logging
 from .configuration_markuplm import MarkupLMConfig
 
 
diff --git a/src/transformers/models/mask2former/image_processing_mask2former.py b/src/transformers/models/mask2former/image_processing_mask2former.py
index b27ef5207b3..07642faf24b 100644
--- a/src/transformers/models/mask2former/image_processing_mask2former.py
+++ b/src/transformers/models/mask2former/image_processing_mask2former.py
@@ -20,8 +20,8 @@ from typing import Any, Dict, Iterable, List, Optional, Set, Tuple, Union
 
 import numpy as np
 
-from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
-from transformers.image_transforms import (
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     get_resize_output_image_size,
     normalize,
@@ -31,7 +31,7 @@ from transformers.image_transforms import (
     to_channel_dimension_format,
     to_numpy_array,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     ChannelDimension,
     ImageInput,
     PILImageResampling,
@@ -40,7 +40,7 @@ from transformers.image_utils import (
     is_batched,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     TensorType,
diff --git a/src/transformers/models/mask2former/modeling_mask2former.py b/src/transformers/models/mask2former/modeling_mask2former.py
index ca9712639df..f8a9c3c1ae4 100644
--- a/src/transformers/models/mask2former/modeling_mask2former.py
+++ b/src/transformers/models/mask2former/modeling_mask2former.py
@@ -24,9 +24,7 @@ import numpy as np
 import torch
 from torch import Tensor, nn
 
-from transformers import AutoBackbone, SwinConfig
-from transformers.utils import logging
-
+from ... import AutoBackbone, SwinConfig
 from ...activations import ACT2FN
 from ...file_utils import (
     ModelOutput,
@@ -38,6 +36,7 @@ from ...file_utils import (
 )
 from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithCrossAttentions
 from ...modeling_utils import PreTrainedModel
+from ...utils import logging
 from .configuration_mask2former import Mask2FormerConfig
 
 
diff --git a/src/transformers/models/maskformer/feature_extraction_maskformer.py b/src/transformers/models/maskformer/feature_extraction_maskformer.py
index 26aff086afd..848c8e12829 100644
--- a/src/transformers/models/maskformer/feature_extraction_maskformer.py
+++ b/src/transformers/models/maskformer/feature_extraction_maskformer.py
@@ -16,8 +16,7 @@
 
 import warnings
 
-from transformers.utils import logging
-
+from ...utils import logging
 from .image_processing_maskformer import MaskFormerImageProcessor
 
 
diff --git a/src/transformers/models/maskformer/image_processing_maskformer.py b/src/transformers/models/maskformer/image_processing_maskformer.py
index 373bee1ab21..ef4314869b4 100644
--- a/src/transformers/models/maskformer/image_processing_maskformer.py
+++ b/src/transformers/models/maskformer/image_processing_maskformer.py
@@ -20,8 +20,8 @@ from typing import TYPE_CHECKING, Any, Dict, Iterable, List, Optional, Set, Tupl
 
 import numpy as np
 
-from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
-from transformers.image_transforms import (
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     get_resize_output_image_size,
     normalize,
@@ -31,7 +31,7 @@ from transformers.image_transforms import (
     to_channel_dimension_format,
     to_numpy_array,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     ChannelDimension,
     ImageInput,
     PILImageResampling,
@@ -40,7 +40,7 @@ from transformers.image_utils import (
     make_list_of_images,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     TensorType,
diff --git a/src/transformers/models/maskformer/modeling_maskformer.py b/src/transformers/models/maskformer/modeling_maskformer.py
index 1482b76f209..878b046c180 100644
--- a/src/transformers/models/maskformer/modeling_maskformer.py
+++ b/src/transformers/models/maskformer/modeling_maskformer.py
@@ -24,9 +24,7 @@ import numpy as np
 import torch
 from torch import Tensor, nn
 
-from transformers import AutoBackbone
-from transformers.utils import logging
-
+from ... import AutoBackbone
 from ...activations import ACT2FN
 from ...modeling_outputs import BaseModelOutputWithCrossAttentions
 from ...modeling_utils import PreTrainedModel
@@ -35,6 +33,7 @@ from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
     is_scipy_available,
+    logging,
     replace_return_docstrings,
     requires_backends,
 )
diff --git a/src/transformers/models/mobilenet_v1/image_processing_mobilenet_v1.py b/src/transformers/models/mobilenet_v1/image_processing_mobilenet_v1.py
index c332b96c86a..9ba6bb1685c 100644
--- a/src/transformers/models/mobilenet_v1/image_processing_mobilenet_v1.py
+++ b/src/transformers/models/mobilenet_v1/image_processing_mobilenet_v1.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -39,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/mobilenet_v2/image_processing_mobilenet_v2.py b/src/transformers/models/mobilenet_v2/image_processing_mobilenet_v2.py
index 7b24547749a..be8020a7ea4 100644
--- a/src/transformers/models/mobilenet_v2/image_processing_mobilenet_v2.py
+++ b/src/transformers/models/mobilenet_v2/image_processing_mobilenet_v2.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_torch_available, is_torch_tensor
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, is_torch_tensor, logging
 
 
 if is_torch_available():
diff --git a/src/transformers/models/mobilevit/image_processing_mobilevit.py b/src/transformers/models/mobilevit/image_processing_mobilevit.py
index b600009c2ea..e121c2ae7ba 100644
--- a/src/transformers/models/mobilevit/image_processing_mobilevit.py
+++ b/src/transformers/models/mobilevit/image_processing_mobilevit.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_torch_available, is_torch_tensor, is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, get_resize_output_image_size, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -32,7 +29,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, is_torch_tensor, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/nezha/configuration_nezha.py b/src/transformers/models/nezha/configuration_nezha.py
index 8d191b7d96e..f41a9b2bf89 100644
--- a/src/transformers/models/nezha/configuration_nezha.py
+++ b/src/transformers/models/nezha/configuration_nezha.py
@@ -1,4 +1,4 @@
-from transformers import PretrainedConfig
+from ... import PretrainedConfig
 
 
 NEZHA_PRETRAINED_CONFIG_ARCHIVE_MAP = {
diff --git a/src/transformers/models/oneformer/image_processing_oneformer.py b/src/transformers/models/oneformer/image_processing_oneformer.py
index 237a3dbad46..67eadd4e841 100644
--- a/src/transformers/models/oneformer/image_processing_oneformer.py
+++ b/src/transformers/models/oneformer/image_processing_oneformer.py
@@ -21,8 +21,8 @@ from typing import Any, Dict, Iterable, List, Optional, Set, Tuple, Union
 import numpy as np
 from huggingface_hub import hf_hub_download
 
-from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
-from transformers.image_transforms import (
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     get_resize_output_image_size,
     normalize,
@@ -32,7 +32,7 @@ from transformers.image_transforms import (
     to_channel_dimension_format,
     to_numpy_array,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     ChannelDimension,
     ImageInput,
     PILImageResampling,
@@ -41,7 +41,7 @@ from transformers.image_utils import (
     make_list_of_images,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     TensorType,
diff --git a/src/transformers/models/oneformer/modeling_oneformer.py b/src/transformers/models/oneformer/modeling_oneformer.py
index 8e41ff86921..6536a54ba5f 100644
--- a/src/transformers/models/oneformer/modeling_oneformer.py
+++ b/src/transformers/models/oneformer/modeling_oneformer.py
@@ -24,9 +24,7 @@ import torch
 from torch import Tensor, nn
 from torch.cuda.amp import autocast
 
-from transformers import AutoBackbone
-from transformers.utils import logging
-
+from ... import AutoBackbone
 from ...activations import ACT2FN
 from ...modeling_outputs import BaseModelOutput
 from ...modeling_utils import PreTrainedModel
@@ -35,6 +33,7 @@ from ...utils import (
     add_start_docstrings,
     add_start_docstrings_to_model_forward,
     is_scipy_available,
+    logging,
     replace_return_docstrings,
     requires_backends,
 )
diff --git a/src/transformers/models/oneformer/processing_oneformer.py b/src/transformers/models/oneformer/processing_oneformer.py
index bc392a77c14..c4479110ae7 100644
--- a/src/transformers/models/oneformer/processing_oneformer.py
+++ b/src/transformers/models/oneformer/processing_oneformer.py
@@ -18,9 +18,8 @@ Image/Text processor class for OneFormer
 
 from typing import List
 
-from transformers.utils import is_torch_available
-
 from ...processing_utils import ProcessorMixin
+from ...utils import is_torch_available
 
 
 if is_torch_available():
diff --git a/src/transformers/models/owlvit/image_processing_owlvit.py b/src/transformers/models/owlvit/image_processing_owlvit.py
index cb5ffc4ddfd..2d86599fffe 100644
--- a/src/transformers/models/owlvit/image_processing_owlvit.py
+++ b/src/transformers/models/owlvit/image_processing_owlvit.py
@@ -19,8 +19,8 @@ from typing import Dict, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
-from transformers.image_transforms import (
+from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
+from ...image_transforms import (
     center_crop,
     center_to_corners_format,
     normalize,
@@ -29,7 +29,7 @@ from transformers.image_transforms import (
     to_channel_dimension_format,
     to_numpy_array,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     OPENAI_CLIP_MEAN,
     OPENAI_CLIP_STD,
     ChannelDimension,
@@ -38,7 +38,7 @@ from transformers.image_utils import (
     make_list_of_images,
     valid_images,
 )
-from transformers.utils import TensorType, is_torch_available, logging
+from ...utils import TensorType, is_torch_available, logging
 
 
 if is_torch_available():
diff --git a/src/transformers/models/owlvit/processing_owlvit.py b/src/transformers/models/owlvit/processing_owlvit.py
index 04b8c191acd..e41dc16354b 100644
--- a/src/transformers/models/owlvit/processing_owlvit.py
+++ b/src/transformers/models/owlvit/processing_owlvit.py
@@ -21,10 +21,9 @@ from typing import List
 
 import numpy as np
 
-from transformers import is_flax_available, is_tf_available, is_torch_available
-
 from ...processing_utils import ProcessorMixin
 from ...tokenization_utils_base import BatchEncoding
+from ...utils import is_flax_available, is_tf_available, is_torch_available
 
 
 class OwlViTProcessor(ProcessorMixin):
diff --git a/src/transformers/models/perceiver/image_processing_perceiver.py b/src/transformers/models/perceiver/image_processing_perceiver.py
index 59b7fd5332b..ecfd7026156 100644
--- a/src/transformers/models/perceiver/image_processing_perceiver.py
+++ b/src/transformers/models/perceiver/image_processing_perceiver.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import center_crop, normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -34,7 +31,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/poolformer/image_processing_poolformer.py b/src/transformers/models/poolformer/image_processing_poolformer.py
index 2548c03da46..d92ffaa3ddc 100644
--- a/src/transformers/models/poolformer/image_processing_poolformer.py
+++ b/src/transformers/models/poolformer/image_processing_poolformer.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/realm/retrieval_realm.py b/src/transformers/models/realm/retrieval_realm.py
index 4bdf19454f0..c84e7af08f5 100644
--- a/src/transformers/models/realm/retrieval_realm.py
+++ b/src/transformers/models/realm/retrieval_realm.py
@@ -20,8 +20,7 @@ from typing import Optional, Union
 import numpy as np
 from huggingface_hub import hf_hub_download
 
-from transformers import AutoTokenizer
-
+from ... import AutoTokenizer
 from ...utils import logging
 
 
diff --git a/src/transformers/models/segformer/image_processing_segformer.py b/src/transformers/models/segformer/image_processing_segformer.py
index 1920c6668b0..36d171f8e2f 100644
--- a/src/transformers/models/segformer/image_processing_segformer.py
+++ b/src/transformers/models/segformer/image_processing_segformer.py
@@ -19,9 +19,6 @@ from typing import Any, Dict, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_torch_available, is_torch_tensor, is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -34,7 +31,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_torch_available, is_torch_tensor, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/sew/modeling_sew.py b/src/transformers/models/sew/modeling_sew.py
index 6b6d3487bb3..7f7a90a8a1e 100644
--- a/src/transformers/models/sew/modeling_sew.py
+++ b/src/transformers/models/sew/modeling_sew.py
@@ -24,9 +24,8 @@ import torch.utils.checkpoint
 from torch import nn
 from torch.nn import CrossEntropyLoss
 
-from transformers.deepspeed import is_deepspeed_zero3_enabled
-
 from ...activations import ACT2FN
+from ...deepspeed import is_deepspeed_zero3_enabled
 from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput
 from ...modeling_utils import PreTrainedModel
 from ...pytorch_utils import torch_int_div
diff --git a/src/transformers/models/sew_d/modeling_sew_d.py b/src/transformers/models/sew_d/modeling_sew_d.py
index 5513fec19d5..02a8477d146 100644
--- a/src/transformers/models/sew_d/modeling_sew_d.py
+++ b/src/transformers/models/sew_d/modeling_sew_d.py
@@ -25,9 +25,8 @@ import torch.utils.checkpoint
 from torch import nn
 from torch.nn import CrossEntropyLoss, LayerNorm
 
-from transformers.deepspeed import is_deepspeed_zero3_enabled
-
 from ...activations import ACT2FN
+from ...deepspeed import is_deepspeed_zero3_enabled
 from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput
 from ...modeling_utils import PreTrainedModel
 from ...pytorch_utils import softmax_backward_data, torch_int_div
diff --git a/src/transformers/models/swin2sr/image_processing_swin2sr.py b/src/transformers/models/swin2sr/image_processing_swin2sr.py
index 24ee846e459..5af2ed41063 100644
--- a/src/transformers/models/swin2sr/image_processing_swin2sr.py
+++ b/src/transformers/models/swin2sr/image_processing_swin2sr.py
@@ -18,12 +18,10 @@ from typing import Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature
 from ...image_transforms import get_image_size, pad, rescale, to_channel_dimension_format
 from ...image_utils import ChannelDimension, ImageInput, make_list_of_images, to_numpy_array, valid_images
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/tvlt/feature_extraction_tvlt.py b/src/transformers/models/tvlt/feature_extraction_tvlt.py
index 7911fa93719..ac219502f1b 100644
--- a/src/transformers/models/tvlt/feature_extraction_tvlt.py
+++ b/src/transformers/models/tvlt/feature_extraction_tvlt.py
@@ -20,8 +20,8 @@ from typing import List, Optional, Union
 import numpy as np
 from numpy.fft import fft
 
-from transformers.feature_extraction_sequence_utils import BatchFeature, SequenceFeatureExtractor
-from transformers.utils import TensorType, logging
+from ...feature_extraction_sequence_utils import BatchFeature, SequenceFeatureExtractor
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/tvlt/image_processing_tvlt.py b/src/transformers/models/tvlt/image_processing_tvlt.py
index d01a49f3510..d07ca31e2ff 100644
--- a/src/transformers/models/tvlt/image_processing_tvlt.py
+++ b/src/transformers/models/tvlt/image_processing_tvlt.py
@@ -17,8 +17,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -38,7 +36,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/upernet/modeling_upernet.py b/src/transformers/models/upernet/modeling_upernet.py
index 1c2d37b0f2e..a00866f77dc 100644
--- a/src/transformers/models/upernet/modeling_upernet.py
+++ b/src/transformers/models/upernet/modeling_upernet.py
@@ -20,8 +20,7 @@ import torch
 from torch import nn
 from torch.nn import CrossEntropyLoss
 
-from transformers import AutoBackbone
-
+from ... import AutoBackbone
 from ...modeling_outputs import SemanticSegmenterOutput
 from ...modeling_utils import BackboneMixin, PreTrainedModel
 from ...utils import add_start_docstrings, add_start_docstrings_to_model_forward, replace_return_docstrings
diff --git a/src/transformers/models/videomae/image_processing_videomae.py b/src/transformers/models/videomae/image_processing_videomae.py
index e3edfdf00d2..3bc1ab5dd65 100644
--- a/src/transformers/models/videomae/image_processing_videomae.py
+++ b/src/transformers/models/videomae/image_processing_videomae.py
@@ -18,9 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,7 +37,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/vilt/image_processing_vilt.py b/src/transformers/models/vilt/image_processing_vilt.py
index 783197c7e95..87b6e682e93 100644
--- a/src/transformers/models/vilt/image_processing_vilt.py
+++ b/src/transformers/models/vilt/image_processing_vilt.py
@@ -19,9 +19,6 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
 
 import numpy as np
 
-from transformers.utils import is_vision_available
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import PaddingMode, normalize, pad, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -36,7 +33,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, is_vision_available, logging
 
 
 if is_vision_available():
diff --git a/src/transformers/models/vit/image_processing_vit.py b/src/transformers/models/vit/image_processing_vit.py
index 5ca0d932880..66d68e7f82d 100644
--- a/src/transformers/models/vit/image_processing_vit.py
+++ b/src/transformers/models/vit/image_processing_vit.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import normalize, rescale, resize, to_channel_dimension_format
 from ...image_utils import (
@@ -32,7 +30,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
+from ...utils import TensorType, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/vit_hybrid/image_processing_vit_hybrid.py b/src/transformers/models/vit_hybrid/image_processing_vit_hybrid.py
index 1babc3677a9..a45ece6e5a3 100644
--- a/src/transformers/models/vit_hybrid/image_processing_vit_hybrid.py
+++ b/src/transformers/models/vit_hybrid/image_processing_vit_hybrid.py
@@ -18,8 +18,6 @@ from typing import Dict, List, Optional, Union
 
 import numpy as np
 
-from transformers.utils.generic import TensorType
-
 from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
 from ...image_transforms import (
     center_crop,
@@ -40,8 +38,7 @@ from ...image_utils import (
     to_numpy_array,
     valid_images,
 )
-from ...utils import logging
-from ...utils.import_utils import is_vision_available
+from ...utils import TensorType, is_vision_available, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/models/yolos/image_processing_yolos.py b/src/transformers/models/yolos/image_processing_yolos.py
index a8fb00aee5c..6b6de49c68a 100644
--- a/src/transformers/models/yolos/image_processing_yolos.py
+++ b/src/transformers/models/yolos/image_processing_yolos.py
@@ -20,9 +20,9 @@ from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Un
 
 import numpy as np
 
-from transformers.feature_extraction_utils import BatchFeature
-from transformers.image_processing_utils import BaseImageProcessor, get_size_dict
-from transformers.image_transforms import (
+from ...feature_extraction_utils import BatchFeature
+from ...image_processing_utils import BaseImageProcessor, get_size_dict
+from ...image_transforms import (
     PaddingMode,
     center_to_corners_format,
     corners_to_center_format,
@@ -34,7 +34,7 @@ from transformers.image_transforms import (
     rgb_to_id,
     to_channel_dimension_format,
 )
-from transformers.image_utils import (
+from ...image_utils import (
     IMAGENET_DEFAULT_MEAN,
     IMAGENET_DEFAULT_STD,
     ChannelDimension,
@@ -48,7 +48,9 @@ from transformers.image_utils import (
     valid_coco_panoptic_annotations,
     valid_images,
 )
-from transformers.utils import (
+from ...utils import (
+    ExplicitEnum,
+    TensorType,
     is_flax_available,
     is_jax_tensor,
     is_scipy_available,
@@ -57,8 +59,8 @@ from transformers.utils import (
     is_torch_available,
     is_torch_tensor,
     is_vision_available,
+    logging,
 )
-from transformers.utils.generic import ExplicitEnum, TensorType
 
 
 if is_torch_available():
@@ -74,6 +76,7 @@ if is_scipy_available():
     import scipy.special
     import scipy.stats
 
+logger = logging.get_logger(__name__)
 
 AnnotationType = Dict[str, Union[int, str, List[Dict]]]
 
diff --git a/src/transformers/pipelines/pt_utils.py b/src/transformers/pipelines/pt_utils.py
index a2ce6fc7f21..4a95d050ec8 100644
--- a/src/transformers/pipelines/pt_utils.py
+++ b/src/transformers/pipelines/pt_utils.py
@@ -2,7 +2,7 @@ import numpy as np
 import torch
 from torch.utils.data import Dataset, IterableDataset
 
-from transformers.utils.generic import ModelOutput
+from ..utils.generic import ModelOutput
 
 
 class PipelineDataset(Dataset):
diff --git a/src/transformers/pipelines/text_generation.py b/src/transformers/pipelines/text_generation.py
index 2800515c821..f95acf7d307 100644
--- a/src/transformers/pipelines/text_generation.py
+++ b/src/transformers/pipelines/text_generation.py
@@ -1,8 +1,7 @@
 import enum
 import warnings
 
-from transformers import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING
-
+from .. import MODEL_FOR_CAUSAL_LM_MAPPING, TF_MODEL_FOR_CAUSAL_LM_MAPPING
 from ..utils import add_end_docstrings, is_tf_available
 from .base import PIPELINE_INIT_ARGS, Pipeline
 
diff --git a/src/transformers/sagemaker/training_args_sm.py b/src/transformers/sagemaker/training_args_sm.py
index e4a356a25b1..3daac7859b5 100644
--- a/src/transformers/sagemaker/training_args_sm.py
+++ b/src/transformers/sagemaker/training_args_sm.py
@@ -20,8 +20,8 @@ from dataclasses import dataclass, field
 
 import torch
 
-from transformers.training_args import TrainingArguments
-from transformers.utils import cached_property, is_sagemaker_dp_enabled, logging
+from ..training_args import TrainingArguments
+from ..utils import cached_property, is_sagemaker_dp_enabled, logging
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/transformers/utils/bitsandbytes.py b/src/transformers/utils/bitsandbytes.py
index 9c183bf5d54..efd9abf6ce5 100644
--- a/src/transformers/utils/bitsandbytes.py
+++ b/src/transformers/utils/bitsandbytes.py
@@ -1,6 +1,6 @@
 from copy import deepcopy
 
-from transformers.utils import is_accelerate_available, is_bitsandbytes_available
+from .import_utils import is_accelerate_available, is_bitsandbytes_available
 
 
 if is_bitsandbytes_available():
diff --git a/src/transformers/utils/hub.py b/src/transformers/utils/hub.py
index 2bee24324ca..3403867eafe 100644
--- a/src/transformers/utils/hub.py
+++ b/src/transformers/utils/hub.py
@@ -49,8 +49,6 @@ from huggingface_hub.utils import (
 )
 from requests.exceptions import HTTPError
 
-from transformers.utils.logging import tqdm
-
 from . import __version__, logging
 from .generic import working_or_temp_dir
 from .import_utils import (
@@ -61,6 +59,7 @@ from .import_utils import (
     is_torch_available,
     is_training_run_on_sagemaker,
 )
+from .logging import tqdm
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
diff --git a/src/transformers/utils/import_utils.py b/src/transformers/utils/import_utils.py
index 45ee725778a..6ef639831b8 100644
--- a/src/transformers/utils/import_utils.py
+++ b/src/transformers/utils/import_utils.py
@@ -29,9 +29,8 @@ from typing import Any
 
 from packaging import version
 
-from transformers.utils.versions import importlib_metadata
-
 from . import logging
+from .versions import importlib_metadata
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name