rename logging -> loggers (#767)

* move logging >> loggers

* add warning

* fix tests

* logging alias

* formatting

* formatting

pl_examples/basic_examples/lightning_module_template.py

@@ -1,7 +1,7 @@
 """
 Example template for defining a system
 """
-import logging
+import logging as log
 import os
 from argparse import ArgumentParser
 from collections import OrderedDict
@@ -215,17 +215,17 @@ def __dataloader(self, train):
 
     @pl.data_loader
     def train_dataloader(self):
-        logging.info('training data loader called')
+        log.info('Training data loader called.')
         return self.__dataloader(train=True)
 
     @pl.data_loader
     def val_dataloader(self):
-        logging.info('val data loader called')
+        log.info('Validation data loader called.')
         return self.__dataloader(train=False)
 
     @pl.data_loader
     def test_dataloader(self):
-        logging.info('test data loader called')
+        log.info('Test data loader called.')
         return self.__dataloader(train=False)
 
     @staticmethod

pytorch_lightning/__init__.py

@@ -25,14 +25,15 @@
     # We are not importing the rest of the scikit during the build
     # process, as it may not be compiled yet
 else:
+    import logging as log
+    log.basicConfig(level=log.INFO)
+
     from .trainer.trainer import Trainer
     from .core.lightning import LightningModule
     from .core.decorators import data_loader
-    import logging
 
     __all__ = [
         'Trainer',
         'LightningModule',
         'data_loader',
     ]
-    logging.basicConfig(level=logging.INFO)

pytorch_lightning/callbacks/pt_callbacks.py

@@ -7,8 +7,9 @@
 
 import os
 import shutil
-import logging
+import logging as log
 import warnings
+
 import numpy as np
 
 from pytorch_lightning.overrides.data_parallel import LightningDistributedDataParallel
@@ -113,7 +114,7 @@ def __init__(self, monitor='val_loss',
 
         if mode not in ['auto', 'min', 'max']:
             if self.verbose > 0:
-                logging.info(f'EarlyStopping mode {mode} is unknown, fallback to auto mode.')
+                log.info(f'EarlyStopping mode {mode} is unknown, fallback to auto mode.')
             mode = 'auto'
 
         if mode == 'min':
@@ -175,7 +176,7 @@ def on_epoch_end(self, epoch, logs=None):
 
     def on_train_end(self, logs=None):
         if self.stopped_epoch > 0 and self.verbose > 0:
-            logging.info(f'Epoch {self.stopped_epoch + 1:05d}: early stopping')
+            log.info(f'Epoch {self.stopped_epoch + 1:05d}: early stopping')
 
 
 class ModelCheckpoint(Callback):
@@ -351,21 +352,21 @@ def on_epoch_end(self, epoch, logs=None):
                         else:
                             self.best = max(self.best_k_models.values())
                         if self.verbose > 0:
-                            logging.info(
+                            log.info(
                                 f'\nEpoch {epoch:05d}: {self.monitor} reached'
                                 f' {current:0.5f} (best {self.best:0.5f}), saving model to'
                                 f' {filepath} as top {self.save_top_k}')
                         self._save_model(filepath)
 
                     else:
                         if self.verbose > 0:
-                            logging.info(
+                            log.info(
                                 f'\nEpoch {epoch:05d}: {self.monitor}'
                                 f' was not in top {self.save_top_k}')
 
             else:
                 if self.verbose > 0:
-                    logging.info(f'\nEpoch {epoch:05d}: saving model to {filepath}')
+                    log.info(f'\nEpoch {epoch:05d}: saving model to {filepath}')
                 self._save_model(filepath)
 
 
@@ -417,6 +418,6 @@ def on_epoch_begin(self, epoch, trainer):
 #     losses = [10, 9, 8, 8, 6, 4.3, 5, 4.4, 2.8, 2.5]
 #     for i, loss in enumerate(losses):
 #         should_stop = c.on_epoch_end(i, logs={'val_loss': loss})
-#         logging.info(loss)
+#         log.info(loss)
 #         if should_stop:
 #             break

pytorch_lightning/core/lightning.py

@@ -1,10 +1,10 @@
 import collections
-import logging
+import logging as log
+import csv
 import os
 import warnings
 from abc import ABC, abstractmethod
 from argparse import Namespace
-import csv
 
 import torch
 import torch.distributed as dist
@@ -1130,7 +1130,7 @@ def __init__(self, hparams):
 
     def summarize(self, mode):
         model_summary = ModelSummary(self, mode=mode)
-        logging.info('\n' + model_summary.__str__())
+        log.info('\n' + model_summary.__str__())
 
     def freeze(self):
         r"""
@@ -1213,7 +1213,7 @@ def on_save_checkpoint(self, checkpoint):
 
 def load_hparams_from_tags_csv(tags_csv):
     if not os.path.isfile(tags_csv):
-        logging.warning(f'Missing Tags: {tags_csv}.')
+        log.warning(f'Missing Tags: {tags_csv}.')
         return Namespace()
 
     tags = {}

pytorch_lightning/core/memory.py

@@ -3,7 +3,7 @@
 '''
 
 import gc
-import logging
+import logging as log
 import os
 import subprocess
 from subprocess import PIPE
@@ -214,7 +214,7 @@ def print_mem_stack():  # pragma: no cover
     for obj in gc.get_objects():
         try:
             if torch.is_tensor(obj) or (hasattr(obj, 'data') and torch.is_tensor(obj.data)):
-                logging.info(type(obj), obj.size())
+                log.info(type(obj), obj.size())
         except Exception:
             pass
 

pytorch_lightning/loggers/__init__.py

@@ -0,0 +1,113 @@
+"""
+Lightning supports most popular logging frameworks (Tensorboard, comet, weights and biases, etc...).
+To use a logger, simply pass it into the trainer.
+
+.. code-block:: python
+
+    from pytorch_lightning import loggers
+
+    # lightning uses tensorboard by default
+    tb_logger = loggers.TensorBoardLogger()
+    trainer = Trainer(logger=tb_logger)
+
+    # or choose from any of the others such as MLFlow, Comet, Neptune, Wandb
+    comet_logger = loggers.CometLogger()
+    trainer = Trainer(logger=comet_logger)
+
+.. note:: All loggers log by default to `os.getcwd()`. To change the path without creating a logger set
+    Trainer(default_save_path='/your/path/to/save/checkpoints')
+
+Custom logger
+-------------
+
+You can implement your own logger by writing a class that inherits from
+`LightningLoggerBase`. Use the `rank_zero_only` decorator to make sure that
+only the first process in DDP training logs data.
+
+.. code-block:: python
+
+    from pytorch_lightning.loggers import LightningLoggerBase, rank_zero_only
+
+    class MyLogger(LightningLoggerBase):
+
+        @rank_zero_only
+        def log_hyperparams(self, params):
+            # params is an argparse.Namespace
+            # your code to record hyperparameters goes here
+            pass
+
+        @rank_zero_only
+        def log_metrics(self, metrics, step):
+            # metrics is a dictionary of metric names and values
+            # your code to record metrics goes here
+            pass
+
+        def save(self):
+            # Optional. Any code necessary to save logger data goes here
+            pass
+
+        @rank_zero_only
+        def finalize(self, status):
+            # Optional. Any code that needs to be run after training
+            # finishes goes here
+
+
+If you write a logger than may be useful to others, please send
+a pull request to add it to Lighting!
+
+Using loggers
+-------------
+
+Call the logger anywhere from your LightningModule by doing:
+
+.. code-block:: python
+
+    def train_step(...):
+        # example
+        self.logger.experiment.whatever_method_summary_writer_supports(...)
+
+    def any_lightning_module_function_or_hook(...):
+        self.logger.experiment.add_histogram(...)
+
+Supported Loggers
+-----------------
+"""
+from os import environ
+
+from .base import LightningLoggerBase, rank_zero_only
+from .tensorboard import TensorBoardLogger
+
+__all__ = ['TensorBoardLogger']
+
+try:
+    # needed to prevent ImportError and duplicated logs.
+    environ["COMET_DISABLE_AUTO_LOGGING"] = "1"
+
+    from .comet import CometLogger
+    __all__.append('CometLogger')
+except ImportError:
+    del environ["COMET_DISABLE_AUTO_LOGGING"]
+
+try:
+    from .mlflow import MLFlowLogger
+    __all__.append('MLFlowLogger')
+except ImportError:
+    pass
+
+try:
+    from .neptune import NeptuneLogger
+    __all__.append('NeptuneLogger')
+except ImportError:
+    pass
+
+try:
+    from .test_tube import TestTubeLogger
+    __all__.append('TestTubeLogger')
+except ImportError:
+    pass
+
+try:
+    from .wandb import WandbLogger
+    __all__.append('WandbLogger')
+except ImportError:
+    pass

pytorch_lightning/logging/comet.py → pytorch_lightning/loggers/comet.py

@@ -31,7 +31,7 @@ def __init__(self, api_key=None, save_dir=None, workspace=None,
         .. code-block:: python
 
             # ONLINE MODE
-            from pytorch_lightning.logging import CometLogger
+            from pytorch_lightning.loggers import CometLogger
 
             # arguments made to CometLogger are passed on to the comet_ml.Experiment class
             comet_logger = CometLogger(
@@ -47,7 +47,7 @@ def __init__(self, api_key=None, save_dir=None, workspace=None,
         .. code-block:: python
 
             # OFFLINE MODE
-            from pytorch_lightning.logging import CometLogger
+            from pytorch_lightning.loggers import CometLogger
 
             # arguments made to CometLogger are passed on to the comet_ml.Experiment class
             comet_logger = CometLogger(

pytorch_lightning/logging/comet_logger.py → pytorch_lightning/loggers/comet_logger.py

@@ -7,4 +7,4 @@
 warnings.warn("`comet_logger` module has been renamed to `comet` since v0.6.0"
               " and will be removed in v0.8.0", DeprecationWarning)
 
-from pytorch_lightning.logging.comet import CometLogger  # noqa: E402
+from pytorch_lightning.loggers.comet import CometLogger  # noqa: E402

pytorch_lightning/logging/mlflow.py → pytorch_lightning/loggers/mlflow.py

@@ -3,7 +3,7 @@
 
 .. code-block:: python
 
-    from pytorch_lightning.logging import MLFlowLogger
+    from pytorch_lightning.loggers import MLFlowLogger
     mlf_logger = MLFlowLogger(
         experiment_name="default",
         tracking_uri="file:/."

pytorch_lightning/logging/mlflow_logger.py → pytorch_lightning/loggers/mlflow_logger.py

@@ -7,4 +7,4 @@
 warnings.warn("`mlflow_logger` module has been renamed to `mlflow` since v0.6.0"
               " and will be removed in v0.8.0", DeprecationWarning)
 
-from pytorch_lightning.logging.mlflow import MLFlowLogger  # noqa: E402
+from pytorch_lightning.loggers.mlflow import MLFlowLogger  # noqa: E402

pytorch_lightning/logging/neptune.py → pytorch_lightning/loggers/neptune.py

@@ -6,7 +6,7 @@
 
 .. code-block:: python
 
-    from pytorch_lightning.logging import NeptuneLogger
+    from pytorch_lightning.loggers import NeptuneLogger
     # arguments made to NeptuneLogger are passed on to the neptune.experiments.Experiment class
 
     neptune_logger = NeptuneLogger(
@@ -48,7 +48,7 @@ def any_lightning_module_function_or_hook(...):
 from torch import is_tensor
 
 # from .base import LightningLoggerBase, rank_zero_only
-from pytorch_lightning.logging.base import LightningLoggerBase, rank_zero_only
+from pytorch_lightning.loggers.base import LightningLoggerBase, rank_zero_only
 
 logger = getLogger(__name__)
 
@@ -66,7 +66,7 @@ def __init__(self, api_key=None, project_name=None, offline_mode=False,
         .. code-block:: python
 
             # ONLINE MODE
-            from pytorch_lightning.logging import NeptuneLogger
+            from pytorch_lightning.loggers import NeptuneLogger
             # arguments made to NeptuneLogger are passed on to the neptune.experiments.Experiment class
 
             neptune_logger = NeptuneLogger(
@@ -81,7 +81,7 @@ def __init__(self, api_key=None, project_name=None, offline_mode=False,
         .. code-block:: python
 
             # OFFLINE MODE
-            from pytorch_lightning.logging import NeptuneLogger
+            from pytorch_lightning.loggers import NeptuneLogger
             # arguments made to NeptuneLogger are passed on to the neptune.experiments.Experiment class
 
             neptune_logger = NeptuneLogger(

pytorch_lightning/logging/test_tube.py → pytorch_lightning/loggers/test_tube.py

@@ -6,7 +6,7 @@
 
 .. code-block:: python
 
-    from pytorch_lightning.logging import TestTubeLogger
+    from pytorch_lightning.loggers import TestTubeLogger
     tt_logger = TestTubeLogger(
         save_dir=".",
         name="default",

pytorch_lightning/logging/test_tube_logger.py → pytorch_lightning/loggers/test_tube_logger.py

@@ -7,4 +7,4 @@
 warnings.warn("`test_tube_logger` module has been renamed to `test_tube` since v0.6.0"
               " and will be removed in v0.8.0", DeprecationWarning)
 
-from pytorch_lightning.logging.test_tube import TestTubeLogger  # noqa: E402
+from pytorch_lightning.loggers.test_tube import TestTubeLogger  # noqa: E402

pytorch_lightning/logging/wandb.py → pytorch_lightning/loggers/wandb.py

@@ -25,7 +25,7 @@ class WandbLogger(LightningLoggerBase):
     --------
     .. code-block:: python
 
-        from pytorch_lightning.logging import WandbLogger
+        from pytorch_lightning.loggers import WandbLogger
         from pytorch_lightning import Trainer
 
         wandb_logger = WandbLogger()