change default logger to dedicated one

Fix test


Fix format

Author: cmpute

pl_examples/basic_examples/lightning_module_template.py

@@ -1,7 +1,6 @@
 """
 Example template for defining a system
 """
-import logging as log
 import os
 from argparse import ArgumentParser
 from collections import OrderedDict
@@ -14,6 +13,7 @@
 from torch.utils.data import DataLoader
 from torchvision.datasets import MNIST
 
+from pytorch_lightning import _logger as log
 from pytorch_lightning.core import LightningModule
 
 

pytorch_lightning/__init__.py

@@ -24,6 +24,9 @@
     # We are not importing the rest of the scikit during the build
     # process, as it may not be compiled yet
 else:
+    import logging
+    _logger = logging.getLogger("lightning")
+
     from .core import LightningModule
     from .trainer import Trainer
     from .callbacks import Callback

pytorch_lightning/callbacks/early_stopping.py

@@ -5,12 +5,12 @@
 
 """
 
-import logging as log
 import warnings
 
 import numpy as np
 
 from .base import Callback
+from pytorch_lightning import _logger as log
 
 
 class EarlyStopping(Callback):

pytorch_lightning/callbacks/model_checkpoint.py

@@ -5,7 +5,6 @@
 Automatically save model checkpoints during training.
 """
 
-import logging as log
 import os
 import shutil
 import warnings
@@ -14,6 +13,7 @@
 import numpy as np
 
 from pytorch_lightning.callbacks.base import Callback
+from pytorch_lightning import _logger as log
 
 
 class ModelCheckpoint(Callback):

pytorch_lightning/core/lightning.py

@@ -1,6 +1,5 @@
 import collections
 import inspect
-import logging as log
 import os
 import warnings
 from abc import ABC, abstractmethod
@@ -15,6 +14,7 @@
 from torch.optim.optimizer import Optimizer
 from torch.utils.data import DataLoader
 
+from pytorch_lightning import _logger as log
 from pytorch_lightning.core.grads import GradInformation
 from pytorch_lightning.core.hooks import ModelHooks
 from pytorch_lightning.core.memory import ModelSummary

pytorch_lightning/core/memory.py

@@ -3,7 +3,6 @@
 """
 
 import gc
-import logging as log
 import os
 import subprocess
 from subprocess import PIPE
@@ -15,6 +14,8 @@
 
 import pytorch_lightning as pl
 
+from pytorch_lightning import _logger as log
+
 
 class ModelSummary(object):
 

pytorch_lightning/core/saving.py

@@ -1,9 +1,10 @@
 import csv
-import logging as log
 import os
 from argparse import Namespace
 from typing import Union, Dict, Any
 
+from pytorch_lightning import _logger as log
+
 
 class ModelIO(object):
 

pytorch_lightning/loggers/comet.py

@@ -6,7 +6,6 @@
 -------------
 """
 
-import logging as log
 from argparse import Namespace
 from typing import Optional, Dict, Union, Any
 
@@ -27,8 +26,9 @@
 import torch
 from torch import is_tensor
 
+from pytorch_lightning import _logger as log
+from pytorch_lightning.loggers.base import LightningLoggerBase, rank_zero_only
 from pytorch_lightning.utilities.debugging import MisconfigurationException
-from .base import LightningLoggerBase, rank_zero_only
 
 
 class CometLogger(LightningLoggerBase):

pytorch_lightning/loggers/mlflow.py

@@ -23,7 +23,6 @@ def any_lightning_module_function_or_hook(...):
         self.logger.experiment.whatever_ml_flow_supports(...)
 
 """
-import logging as log
 from argparse import Namespace
 from time import time
 from typing import Optional, Dict, Any, Union
@@ -34,7 +33,8 @@ def any_lightning_module_function_or_hook(...):
     raise ImportError('You want to use `mlflow` logger which is not installed yet,'
                       ' install it with `pip install mlflow`.')
 
-from .base import LightningLoggerBase, rank_zero_only
+from pytorch_lightning import _logger as log
+from pytorch_lightning.loggers.base import LightningLoggerBase, rank_zero_only
 
 
 class MLFlowLogger(LightningLoggerBase):

pytorch_lightning/loggers/neptune.py

@@ -6,7 +6,6 @@
 NeptuneLogger
 --------------
 """
-import logging as log
 from argparse import Namespace
 from typing import Optional, List, Dict, Any, Union, Iterable
 
@@ -20,6 +19,7 @@
 import torch
 from torch import is_tensor
 
+from pytorch_lightning import _logger as log
 from pytorch_lightning.loggers.base import LightningLoggerBase, rank_zero_only