Sort imports with isort.

This is the result of:

    $ isort --recursive examples templates transformers utils hubconf.py setup.py

transformers/modeling_gpt2.py

@@ -30,9 +30,10 @@ import torch.nn as nn
 from torch.nn import CrossEntropyLoss
 from torch.nn.parameter import Parameter
 
-from .modeling_utils import PreTrainedModel, Conv1D, prune_conv1d_layer, SequenceSummary
 from .configuration_gpt2 import GPT2Config
 from .file_utils import add_start_docstrings
+from .modeling_utils import Conv1D, PreTrainedModel, SequenceSummary, prune_conv1d_layer
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_mmbt.py

@@ -25,6 +25,7 @@ from torch.nn import CrossEntropyLoss, MSELoss
 
 from .file_utils import add_start_docstrings
 
+
 logger = logging.getLogger(__name__)
 
 

transformers/modeling_openai.py

@@ -30,9 +30,10 @@ import torch.nn as nn
 from torch.nn import CrossEntropyLoss
 from torch.nn.parameter import Parameter
 
-from .modeling_utils import PreTrainedModel, Conv1D, prune_conv1d_layer, SequenceSummary
 from .configuration_openai import OpenAIGPTConfig
 from .file_utils import add_start_docstrings
+from .modeling_utils import Conv1D, PreTrainedModel, SequenceSummary, prune_conv1d_layer
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_roberta.py

@@ -23,9 +23,10 @@ import torch
 import torch.nn as nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
-from .modeling_bert import BertEmbeddings, BertLayerNorm, BertModel, BertPreTrainedModel, gelu
 from .configuration_roberta import RobertaConfig
 from .file_utils import add_start_docstrings
+from .modeling_bert import BertEmbeddings, BertLayerNorm, BertModel, BertPreTrainedModel, gelu
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_t5.py

@@ -16,23 +16,24 @@
 
 from __future__ import absolute_import, division, print_function, unicode_literals
 
+import copy
+import itertools
 import json
 import logging
 import math
 import os
 import sys
-import copy
-import itertools
 from io import open
 
 import torch
-from torch import nn
 import torch.nn.functional as F
+from torch import nn
 from torch.nn import CrossEntropyLoss, MSELoss
 
-from .modeling_utils import PreTrainedModel, prune_linear_layer
 from .configuration_t5 import T5Config
-from .file_utils import add_start_docstrings, DUMMY_INPUTS, DUMMY_MASK
+from .file_utils import DUMMY_INPUTS, DUMMY_MASK, add_start_docstrings
+from .modeling_utils import PreTrainedModel, prune_linear_layer
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_albert.py

@@ -22,11 +22,10 @@ import sys
 import tensorflow as tf
 
 from .configuration_albert import AlbertConfig
-from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
-from .modeling_tf_bert import ACT2FN, TFBertSelfAttention
 from .file_utils import add_start_docstrings
+from .modeling_tf_bert import ACT2FN, TFBertSelfAttention
+from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
 
-import logging
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_auto.py

@@ -29,62 +29,61 @@ from .configuration_auto import (
     XLMConfig,
     XLNetConfig,
 )
-
+from .file_utils import add_start_docstrings
+from .modeling_tf_albert import (
+    TF_ALBERT_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFAlbertForMaskedLM,
+    TFAlbertForSequenceClassification,
+    TFAlbertModel,
+)
 from .modeling_tf_bert import (
-    TFBertModel,
+    TF_BERT_PRETRAINED_MODEL_ARCHIVE_MAP,
     TFBertForMaskedLM,
-    TFBertForSequenceClassification,
     TFBertForQuestionAnswering,
+    TFBertForSequenceClassification,
     TFBertForTokenClassification,
-    TF_BERT_PRETRAINED_MODEL_ARCHIVE_MAP,
-)
-from .modeling_tf_openai import TFOpenAIGPTModel, TFOpenAIGPTLMHeadModel, TF_OPENAI_GPT_PRETRAINED_MODEL_ARCHIVE_MAP
-from .modeling_tf_gpt2 import TFGPT2Model, TFGPT2LMHeadModel, TF_GPT2_PRETRAINED_MODEL_ARCHIVE_MAP
-from .modeling_tf_transfo_xl import (
-    TFTransfoXLModel,
-    TFTransfoXLLMHeadModel,
-    TF_TRANSFO_XL_PRETRAINED_MODEL_ARCHIVE_MAP,
-)
-from .modeling_tf_xlnet import (
-    TFXLNetModel,
-    TFXLNetLMHeadModel,
-    TFXLNetForSequenceClassification,
-    TFXLNetForQuestionAnsweringSimple,
-    TFXLNetForTokenClassification,
-    TF_XLNET_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFBertModel,
 )
-from .modeling_tf_xlm import (
-    TFXLMModel,
-    TFXLMWithLMHeadModel,
-    TFXLMForSequenceClassification,
-    TFXLMForQuestionAnsweringSimple,
-    TF_XLM_PRETRAINED_MODEL_ARCHIVE_MAP,
+from .modeling_tf_ctrl import TF_CTRL_PRETRAINED_MODEL_ARCHIVE_MAP, TFCTRLLMHeadModel, TFCTRLModel
+from .modeling_tf_distilbert import (
+    TF_DISTILBERT_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFDistilBertForMaskedLM,
+    TFDistilBertForQuestionAnswering,
+    TFDistilBertForSequenceClassification,
+    TFDistilBertForTokenClassification,
+    TFDistilBertModel,
 )
+from .modeling_tf_gpt2 import TF_GPT2_PRETRAINED_MODEL_ARCHIVE_MAP, TFGPT2LMHeadModel, TFGPT2Model
+from .modeling_tf_openai import TF_OPENAI_GPT_PRETRAINED_MODEL_ARCHIVE_MAP, TFOpenAIGPTLMHeadModel, TFOpenAIGPTModel
 from .modeling_tf_roberta import (
-    TFRobertaModel,
+    TF_ROBERTA_PRETRAINED_MODEL_ARCHIVE_MAP,
     TFRobertaForMaskedLM,
     TFRobertaForSequenceClassification,
     TFRobertaForTokenClassification,
-    TF_ROBERTA_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFRobertaModel,
 )
-from .modeling_tf_distilbert import (
-    TFDistilBertModel,
-    TFDistilBertForQuestionAnswering,
-    TFDistilBertForMaskedLM,
-    TFDistilBertForSequenceClassification,
-    TFDistilBertForTokenClassification,
-    TF_DISTILBERT_PRETRAINED_MODEL_ARCHIVE_MAP,
+from .modeling_tf_t5 import TF_T5_PRETRAINED_MODEL_ARCHIVE_MAP, TFT5Model, TFT5WithLMHeadModel
+from .modeling_tf_transfo_xl import (
+    TF_TRANSFO_XL_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFTransfoXLLMHeadModel,
+    TFTransfoXLModel,
 )
-from .modeling_tf_ctrl import TFCTRLModel, TFCTRLLMHeadModel, TF_CTRL_PRETRAINED_MODEL_ARCHIVE_MAP
-from .modeling_tf_albert import (
-    TFAlbertModel,
-    TFAlbertForMaskedLM,
-    TFAlbertForSequenceClassification,
-    TF_ALBERT_PRETRAINED_MODEL_ARCHIVE_MAP,
+from .modeling_tf_xlm import (
+    TF_XLM_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFXLMForQuestionAnsweringSimple,
+    TFXLMForSequenceClassification,
+    TFXLMModel,
+    TFXLMWithLMHeadModel,
+)
+from .modeling_tf_xlnet import (
+    TF_XLNET_PRETRAINED_MODEL_ARCHIVE_MAP,
+    TFXLNetForQuestionAnsweringSimple,
+    TFXLNetForSequenceClassification,
+    TFXLNetForTokenClassification,
+    TFXLNetLMHeadModel,
+    TFXLNetModel,
 )
-from .modeling_tf_t5 import TFT5Model, TFT5WithLMHeadModel, TF_T5_PRETRAINED_MODEL_ARCHIVE_MAP
 
-from .file_utils import add_start_docstrings
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_bert.py

@@ -28,8 +28,9 @@ import numpy as np
 import tensorflow as tf
 
 from .configuration_bert import BertConfig
-from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
 from .file_utils import add_start_docstrings
+from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_ctrl.py

@@ -21,12 +21,14 @@ import logging
 import os
 import sys
 from io import open
+
 import numpy as np
 import tensorflow as tf
 
 from .configuration_ctrl import CTRLConfig
-from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list, TFSharedEmbeddings
 from .file_utils import add_start_docstrings
+from .modeling_tf_utils import TFPreTrainedModel, TFSharedEmbeddings, get_initializer, shape_list
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_distilbert.py

@@ -16,21 +16,21 @@
 """
 from __future__ import absolute_import, division, print_function, unicode_literals
 
+import copy
+import itertools
 import json
 import logging
 import math
-import copy
 import sys
 from io import open
 
-import itertools
-
 import numpy as np
 import tensorflow as tf
 
 from .configuration_distilbert import DistilBertConfig
-from .modeling_tf_utils import TFPreTrainedModel, TFSharedEmbeddings, shape_list, get_initializer
 from .file_utils import add_start_docstrings
+from .modeling_tf_utils import TFPreTrainedModel, TFSharedEmbeddings, get_initializer, shape_list
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_gpt2.py

@@ -28,16 +28,17 @@ from io import open
 import numpy as np
 import tensorflow as tf
 
+from .configuration_gpt2 import GPT2Config
+from .file_utils import add_start_docstrings
 from .modeling_tf_utils import (
-    TFPreTrainedModel,
     TFConv1D,
-    TFSharedEmbeddings,
+    TFPreTrainedModel,
     TFSequenceSummary,
-    shape_list,
+    TFSharedEmbeddings,
     get_initializer,
+    shape_list,
 )
-from .configuration_gpt2 import GPT2Config
-from .file_utils import add_start_docstrings
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_openai.py

@@ -28,16 +28,17 @@ from io import open
 import numpy as np
 import tensorflow as tf
 
+from .configuration_openai import OpenAIGPTConfig
+from .file_utils import add_start_docstrings
 from .modeling_tf_utils import (
-    TFPreTrainedModel,
     TFConv1D,
-    TFSharedEmbeddings,
+    TFPreTrainedModel,
     TFSequenceSummary,
-    shape_list,
+    TFSharedEmbeddings,
     get_initializer,
+    shape_list,
 )
-from .configuration_openai import OpenAIGPTConfig
-from .file_utils import add_start_docstrings
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_pytorch_utils.py

@@ -20,8 +20,10 @@ from __future__ import absolute_import, division, print_function, unicode_litera
 import logging
 import os
 import re
+
 import numpy
 
+
 logger = logging.getLogger(__name__)
 
 

transformers/modeling_tf_roberta.py

@@ -22,10 +22,10 @@ import logging
 import tensorflow as tf
 
 from .configuration_roberta import RobertaConfig
-from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
 from .file_utils import add_start_docstrings
-
 from .modeling_tf_bert import TFBertEmbeddings, TFBertMainLayer, gelu, gelu_new
+from .modeling_tf_utils import TFPreTrainedModel, get_initializer, shape_list
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_t5.py

@@ -17,16 +17,17 @@
 
 from __future__ import absolute_import, division, print_function, unicode_literals
 
-import logging
-import math
 import copy
 import itertools
+import logging
+import math
 
 import tensorflow as tf
 
 from .configuration_t5 import T5Config
+from .file_utils import DUMMY_INPUTS, DUMMY_MASK, add_start_docstrings
 from .modeling_tf_utils import TFPreTrainedModel, TFSharedEmbeddings, shape_list
-from .file_utils import add_start_docstrings, DUMMY_INPUTS, DUMMY_MASK
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_transfo_xl.py

@@ -18,11 +18,11 @@
 
 from __future__ import absolute_import, division, print_function, unicode_literals
 
-import os
+import collections
 import json
-import math
 import logging
-import collections
+import math
+import os
 import sys
 from io import open
 
@@ -30,9 +30,10 @@ import numpy as np
 import tensorflow as tf
 
 from .configuration_transfo_xl import TransfoXLConfig
-from .modeling_tf_utils import TFPreTrainedModel, TFConv1D, TFSequenceSummary, shape_list, get_initializer
-from .modeling_tf_transfo_xl_utilities import TFAdaptiveSoftmaxMask
 from .file_utils import add_start_docstrings
+from .modeling_tf_transfo_xl_utilities import TFAdaptiveSoftmaxMask
+from .modeling_tf_utils import TFConv1D, TFPreTrainedModel, TFSequenceSummary, get_initializer, shape_list
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_transfo_xl_utilities.py

@@ -19,7 +19,6 @@
 from collections import defaultdict
 
 import numpy as np
-
 import tensorflow as tf
 
 from .modeling_tf_utils import shape_list

transformers/modeling_tf_utils.py

@@ -20,22 +20,23 @@ from __future__ import absolute_import, division, print_function, unicode_litera
 import logging
 import os
 
+import h5py
 import tensorflow as tf
 from tensorflow.python.keras.saving import hdf5_format
-import h5py
 
 from .configuration_utils import PretrainedConfig
 from .file_utils import (
+    DUMMY_INPUTS,
     TF2_WEIGHTS_NAME,
     TF_WEIGHTS_NAME,
     WEIGHTS_NAME,
-    DUMMY_INPUTS,
     cached_path,
     hf_bucket_url,
     is_remote_url,
 )
 from .modeling_tf_pytorch_utils import load_pytorch_checkpoint_in_tf2_model
 
+
 logger = logging.getLogger(__name__)
 
 

transformers/modeling_tf_xlm.py

@@ -16,24 +16,25 @@
 """
 from __future__ import absolute_import, division, print_function, unicode_literals
 
+import itertools
 import logging
 import math
 import os
 
-import itertools
 import numpy as np
 import tensorflow as tf
 
 from .configuration_xlm import XLMConfig
+from .file_utils import add_start_docstrings
 from .modeling_tf_utils import (
+    DUMMY_INPUTS,
     TFPreTrainedModel,
-    TFSharedEmbeddings,
     TFSequenceSummary,
-    shape_list,
+    TFSharedEmbeddings,
     get_initializer,
-    DUMMY_INPUTS,
+    shape_list,
 )
-from .file_utils import add_start_docstrings
+
 
 logger = logging.getLogger(__name__)
 

transformers/modeling_tf_xlnet.py

@@ -28,8 +28,8 @@ import numpy as np
 import tensorflow as tf
 
 from .configuration_xlnet import XLNetConfig
-from .modeling_tf_utils import TFPreTrainedModel, TFSharedEmbeddings, TFSequenceSummary, shape_list, get_initializer
 from .file_utils import add_start_docstrings
+from .modeling_tf_utils import TFPreTrainedModel, TFSequenceSummary, TFSharedEmbeddings, get_initializer, shape_list
 
 
 logger = logging.getLogger(__name__)