Centralize logging (#6434)

* Logging

* Style

* hf_logging > utils.logging

* Address @thomwolf's comments

* Update test

* Update src/transformers/benchmark/benchmark_utils.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Revert bad change
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

src/transformers/modeling_tf_transfo_xl.py

@@ -17,7 +17,6 @@
 """
 
 
-import logging
 from dataclasses import dataclass
 from typing import List, Optional, Tuple
 
@@ -28,9 +27,10 @@ from .file_utils import ModelOutput, add_code_sample_docstrings, add_start_docst
 from .modeling_tf_transfo_xl_utilities import TFAdaptiveSoftmaxMask
 from .modeling_tf_utils import TFPreTrainedModel, get_initializer, keras_serializable, shape_list
 from .tokenization_utils import BatchEncoding
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "TransfoXLConfig"
 _TOKENIZER_FOR_DOC = "TransfoXLTokenizer"

src/transformers/modeling_tf_utils.py

@@ -15,7 +15,6 @@
 # limitations under the License.
 """TF general model utils."""
 import functools
-import logging
 import os
 import warnings
 from typing import Dict, List, Optional, Union
@@ -29,9 +28,10 @@ from .configuration_utils import PretrainedConfig
 from .file_utils import DUMMY_INPUTS, TF2_WEIGHTS_NAME, WEIGHTS_NAME, cached_path, hf_bucket_url, is_remote_url
 from .generation_tf_utils import TFGenerationMixin
 from .modeling_tf_pytorch_utils import load_pytorch_checkpoint_in_tf2_model
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 
 class TFModelUtilsMixin:

src/transformers/modeling_tf_xlm.py

@@ -17,7 +17,6 @@
 
 
 import itertools
-import logging
 import math
 import warnings
 from dataclasses import dataclass
@@ -54,9 +53,10 @@ from .modeling_tf_utils import (
     shape_list,
 )
 from .tokenization_utils import BatchEncoding
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "XLMConfig"
 _TOKENIZER_FOR_DOC = "XLMTokenizer"

src/transformers/modeling_tf_xlm_roberta.py

@@ -15,9 +15,6 @@
 # limitations under the License.
 """ TF 2.0  XLM-RoBERTa model. """
 
-
-import logging
-
 from .configuration_xlm_roberta import XLMRobertaConfig
 from .file_utils import add_start_docstrings
 from .modeling_tf_roberta import (
@@ -28,9 +25,10 @@ from .modeling_tf_roberta import (
     TFRobertaForTokenClassification,
     TFRobertaModel,
 )
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 TF_XLM_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST = [
     # See all XLM-RoBERTa models at https://huggingface.co/models?filter=xlm-roberta

src/transformers/modeling_tf_xlnet.py

@@ -17,7 +17,6 @@
 """
 
 
-import logging
 from dataclasses import dataclass
 from typing import List, Optional, Tuple
 
@@ -47,9 +46,10 @@ from .modeling_tf_utils import (
     shape_list,
 )
 from .tokenization_utils import BatchEncoding
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "XLNetConfig"
 _TOKENIZER_FOR_DOC = "XLNetTokenizer"

src/transformers/modeling_transfo_xl.py

@@ -19,7 +19,6 @@
 """
 
 
-import logging
 from dataclasses import dataclass
 from typing import List, Optional, Tuple
 
@@ -31,9 +30,10 @@ from .configuration_transfo_xl import TransfoXLConfig
 from .file_utils import ModelOutput, add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_callable
 from .modeling_transfo_xl_utilities import ProjectedAdaptiveLogSoftmax
 from .modeling_utils import PreTrainedModel
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "TransfoXLConfig"
 _TOKENIZER_FOR_DOC = "TransfoXLTokenizer"

src/transformers/modeling_utils.py

@@ -15,7 +15,6 @@
 # limitations under the License.
 
 import inspect
-import logging
 import os
 import re
 from dataclasses import dataclass
@@ -41,9 +40,10 @@ from .file_utils import (
     replace_return_docstrings,
 )
 from .generation_utils import GenerationMixin
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 
 try:

src/transformers/modeling_xlm.py

@@ -17,7 +17,6 @@
 
 
 import itertools
-import logging
 import math
 import warnings
 from dataclasses import dataclass
@@ -54,9 +53,10 @@ from .modeling_utils import (
     find_pruneable_heads_and_indices,
     prune_linear_layer,
 )
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "XLMConfig"
 _TOKENIZER_FOR_DOC = "XLMTokenizer"

src/transformers/modeling_xlm_roberta.py

@@ -15,9 +15,6 @@
 # limitations under the License.
 """PyTorch XLM-RoBERTa model. """
 
-
-import logging
-
 from .configuration_xlm_roberta import XLMRobertaConfig
 from .file_utils import add_start_docstrings
 from .modeling_roberta import (
@@ -28,9 +25,10 @@ from .modeling_roberta import (
     RobertaForTokenClassification,
     RobertaModel,
 )
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 XLM_ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST = [
     "xlm-roberta-base",

src/transformers/modeling_xlnet.py

@@ -17,7 +17,6 @@
 """
 
 
-import logging
 from dataclasses import dataclass
 from typing import List, Optional, Tuple
 
@@ -43,9 +42,10 @@ from .modeling_utils import (
     SequenceSummary,
     apply_chunking_to_forward,
 )
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 _CONFIG_FOR_DOC = "XLNetConfig"
 _TOKENIZER_FOR_DOC = "XLNetTokenizer"

src/transformers/optimization.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """PyTorch optimization for BERT model."""
 
-import logging
 import math
 from typing import Callable, Iterable, Tuple
 
@@ -22,8 +21,10 @@ import torch
 from torch.optim import Optimizer
 from torch.optim.lr_scheduler import LambdaLR
 
+from .utils import logging
 
-logger = logging.getLogger(__name__)
+
+logger = logging.get_logger(__name__)
 
 
 def get_constant_schedule(optimizer: Optimizer, last_epoch: int = -1):

src/transformers/pipelines.py

@@ -16,7 +16,6 @@
 
 import csv
 import json
-import logging
 import os
 import pickle
 import sys
@@ -39,6 +38,7 @@ from .tokenization_auto import AutoTokenizer
 from .tokenization_bert import BasicTokenizer
 from .tokenization_utils import PreTrainedTokenizer
 from .tokenization_utils_base import BatchEncoding, PaddingStrategy
+from .utils import logging
 
 
 if is_tf_available():
@@ -80,7 +80,7 @@ if TYPE_CHECKING:
     from .modeling_utils import PreTrainedModel
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 
 def get_framework(model=None):

src/transformers/tokenization_albert.py

@@ -15,16 +15,16 @@
 """ Tokenization classes for ALBERT model."""
 
 
-import logging
 import os
 import unicodedata
 from shutil import copyfile
 from typing import List, Optional
 
 from .tokenization_utils import PreTrainedTokenizer
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 VOCAB_FILES_NAMES = {"vocab_file": "spiece.model"}
 
 PRETRAINED_VOCAB_FILES_MAP = {

src/transformers/tokenization_auto.py

@@ -15,7 +15,6 @@
 """ Auto Tokenizer class. """
 
 
-import logging
 from collections import OrderedDict
 
 from .configuration_auto import (
@@ -69,9 +68,10 @@ from .tokenization_transfo_xl import TransfoXLTokenizer, TransfoXLTokenizerFast
 from .tokenization_xlm import XLMTokenizer
 from .tokenization_xlm_roberta import XLMRobertaTokenizer
 from .tokenization_xlnet import XLNetTokenizer
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 
 TOKENIZER_MAPPING = OrderedDict(

src/transformers/tokenization_bart.py

@@ -13,14 +13,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import logging
 from typing import List, Optional
 
 from .tokenization_roberta import RobertaTokenizer, RobertaTokenizerFast
 from .tokenization_utils_base import BatchEncoding
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 
 # vocab and merges same as roberta

src/transformers/tokenization_bert.py

@@ -16,7 +16,6 @@
 
 
 import collections
-import logging
 import os
 import unicodedata
 from typing import List, Optional
@@ -25,9 +24,10 @@ from tokenizers import BertWordPieceTokenizer
 
 from .tokenization_utils import PreTrainedTokenizer, _is_control, _is_punctuation, _is_whitespace
 from .tokenization_utils_fast import PreTrainedTokenizerFast
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"}
 

src/transformers/tokenization_bert_japanese.py

@@ -16,15 +16,15 @@
 
 
 import collections
-import logging
 import os
 import unicodedata
 from typing import Optional
 
 from .tokenization_bert import BasicTokenizer, BertTokenizer, WordpieceTokenizer, load_vocab
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"}
 

src/transformers/tokenization_camembert.py

@@ -15,7 +15,6 @@
 """ Tokenization classes for Camembert model."""
 
 
-import logging
 import os
 from shutil import copyfile
 from typing import List, Optional
@@ -24,9 +23,10 @@ import sentencepiece as spm
 
 from .tokenization_utils import PreTrainedTokenizer
 from .tokenization_xlnet import SPIECE_UNDERLINE
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 VOCAB_FILES_NAMES = {"vocab_file": "sentencepiece.bpe.model"}
 

src/transformers/tokenization_ctrl.py

@@ -16,15 +16,15 @@
 
 
 import json
-import logging
 import os
 
 import regex as re
 
 from .tokenization_utils import PreTrainedTokenizer
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 VOCAB_FILES_NAMES = {
     "vocab_file": "vocab.json",

src/transformers/tokenization_distilbert.py

@@ -14,13 +14,11 @@
 # limitations under the License.
 """Tokenization classes for DistilBERT."""
 
-
-import logging
-
 from .tokenization_bert import BertTokenizer, BertTokenizerFast
+from .utils import logging
 
 
-logger = logging.getLogger(__name__)
+logger = logging.get_logger(__name__)
 
 VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"}