Add dummy logger for internally disabling logging for some features

pytorch_lightning/loggers/base.py

@@ -292,6 +292,41 @@ def version(self) -> str:
         return '_'.join([str(logger.version) for logger in self._logger_iterable])
 
 
+class DummyExperiment(object):
+    """ Dummy experiment """
+    def nop(*args, **kw):
+        pass
+
+    def __getattr__(self, _):
+        return self.nop
+
+
+class DummyLogger(LightningLoggerBase):
+    """ Dummy logger for internal use. Is usefull if we want to disable users
+        logger for a feature, but still secure that users code can run """
+    def __init__(self):
+        super().__init__()
+        self._experiment = DummyExperiment()
+
+    @property
+    def experiment(self):
+        return self._experiment
+
+    def log_metrics(self, metrics, step):
+        pass
+
+    def log_hyperparams(self, params):
+        pass
+
+    @property
+    def name(self):
+        pass
+
+    @property
+    def version(self):
+        pass
+
+
 def merge_dicts(
         dicts: Sequence[Mapping],
         agg_key_funcs: Optional[Mapping[str, Callable[[Sequence[float]], float]]] = None,

pytorch_lightning/trainer/lr_finder.py

@@ -13,6 +13,7 @@
 
 from pytorch_lightning.core.lightning import LightningModule
 from pytorch_lightning.callbacks import Callback
+from pytorch_lightning.loggers.base import DummyLogger
 from pytorch_lightning import _logger as log
 from pytorch_lightning.utilities.exceptions import MisconfigurationException
 from pytorch_lightning.utilities import rank_zero_warn
@@ -133,7 +134,7 @@ def lr_find(self,
                                       progress_bar_refresh_rate=1)]
 
         # No logging
-        self.logger = None
+        self.logger = DummyLogger()
 
         # Max step set to number of iterations
         self.max_steps = num_training

pytorch_lightning/trainer/training_tricks.py

@@ -12,6 +12,7 @@
 from pytorch_lightning import _logger as log
 from pytorch_lightning.core.lightning import LightningModule
 from pytorch_lightning.callbacks import GradientAccumulationScheduler
+from pytorch_lightning.loggers.base import DummyLogger
 from pytorch_lightning.utilities.exceptions import MisconfigurationException
 from pytorch_lightning.utilities.memory import is_oom_error, garbage_collection_cuda
 
@@ -195,7 +196,7 @@ def __scale_batch_reset_params(self, model, steps_per_trial):
         self.auto_scale_batch_size = None  # prevent recursion
         self.max_steps = steps_per_trial  # take few steps
         self.weights_summary = None  # not needed before full run
-        self.logger = None  # not needed before full run
+        self.logger = DummyLogger()
         self.callbacks = []  # not needed before full run
         self.checkpoint_callback = False  # required for saving
         self.early_stop_callback = None