Learning rate log callback

CHANGELOG.md

@@ -9,9 +9,9 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 ### Added
 
 - Added `auto_select_gpus` flag to trainer that enables automatic selection of available GPUs on exclusive mode systems.
-- Added learning rate finder ([#1347](https://github.com/PyTorchLightning/pytorch-lightning/pull/1347))
-
+- Added learining rate finder ([#1347](https://github.com/PyTorchLightning/pytorch-lightning/pull/1347))
 - Added `terminate_on_nan` flag to trainer that performs a NaN check with each training iteration when set to `True`. ([#1475](https://github.com/PyTorchLightning/pytorch-lightning/pull/1475))
+- Added callback for logging learning rates ([#1498](https://github.com/PyTorchLightning/pytorch-lightning/pull/1498))
 
 ### Changed
 

docs/source/callbacks.rst

@@ -78,3 +78,10 @@ We successfully extended functionality without polluting our super clean
         _save_model,
         _abc_impl,
         check_monitor_top_k,
+
+---------
+
+.. automodule:: pytorch_lightning.callbacks.lr_logger
+    :noindex:
+    :exclude-members:
+        _extract_lr

pytorch_lightning/callbacks/__init__.py

@@ -2,10 +2,12 @@
 from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.callbacks.gradient_accumulation_scheduler import GradientAccumulationScheduler
 from pytorch_lightning.callbacks.model_checkpoint import ModelCheckpoint
+from pytorch_lightning.callbacks.lr_logger import LearningRateLogger
 
 __all__ = [
     'Callback',
     'EarlyStopping',
     'ModelCheckpoint',
     'GradientAccumulationScheduler',
+    'LearningRateLogger'
 ]

pytorch_lightning/callbacks/lr_logger.py

@@ -0,0 +1,117 @@
+r"""
+
+Logging of learning rates
+=========================
+
+Log learning rate for lr schedulers during training
+
+"""
+
+from pytorch_lightning.callbacks.base import Callback
+from pytorch_lightning.utilities.exceptions import MisconfigurationException
+
+
+class LearningRateLogger(Callback):
+    r"""
+    Automatically logs learning rate for learning rate schedulers during training.
+
+    Example::
+
+        >>> from pytorch_lightning import Trainer
+        >>> from pytorch_lightning.callbacks import LearningRateLogger
+        >>> lr_logger = LearningRateLogger()
+        >>> trainer = Trainer(callbacks=[lr_logger])
+
+    Logging names are automatically determined based on optimizer class name.
+    In case of multiple optimizers of same type, they will be named `Adam`,
+    `Adam-1` etc. If a optimizer has multiple parameter groups they will
+    be named `Adam/pg1`, `Adam/pg2` etc. To control naming, pass in a
+    `name` keyword in the construction of the learning rate schdulers
+
+    Example::
+
+        def configure_optimizer(self):
+            optimizer = torch.optim.Adam(...)
+            lr_scheduler = {'scheduler': torch.optim.lr_schedulers.LambdaLR(optimizer, ...)
+                            'name': 'my_logging_name'}
+            return [optimizer], [lr_scheduler]
+    """
+    def __init__(self):
+        self.lrs = {}
+        self.names = []
+
+    def on_train_start(self, trainer, pl_module):
+        """ Called before training, determines unique names for all lr
+            schedulers in the case of multiple of the same type or in
+            the case of multiple parameter groups
+        """
+        if trainer.lr_schedulers == []:
+            raise MisconfigurationException(
+                'Cannot use LearningRateLogger callback with models that have no'
+                ' learning rate schedulers. Please see documentation for'
+                ' `configure_optimizers` method.')
+
+        if not trainer.logger:
+            raise MisconfigurationException(
+                'Cannot use LearningRateLogger callback with Trainer that have no logger.')
+
+        # Create uniqe names in the case we have multiple of the same learning
+        # rate schduler + multiple parameter groups
+        names = []
+        for scheduler in trainer.lr_schedulers:
+            sch = scheduler['scheduler']
+            if 'name' in scheduler:
+                name = scheduler['name']
+            else:
+                opt_name = 'lr-' + sch.optimizer.__class__.__name__
+                name = opt_name
+                counter = 0
+                # Multiple schduler of the same type
+                while True:
+                    counter += 1
+                    if name not in names:
+                        break
+                    name = opt_name + '-' + str(counter)
+
+            # Multiple param groups for the same schduler
+            param_groups = sch.optimizer.param_groups
+            if len(param_groups) != 1:
+                for i, pg in enumerate(param_groups):
+                    temp = name + '/pg' + str(i + 1)
+                    names.append(temp)
+            else:
+                names.append(name)
+
+            self.names.append(name)
+
+        # Initialize for storing values
+        for name in names:
+            self.lrs[name] = []
+
+    def on_batch_start(self, trainer, pl_module):
+        latest_stat = self._extract_lr(trainer, 'step')
+        if trainer.logger and latest_stat:
+            trainer.logger.log_metrics(latest_stat, step=trainer.global_step)
+
+    def on_epoch_start(self, trainer, pl_module):
+        latest_stat = self._extract_lr(trainer, 'epoch')
+        if trainer.logger and latest_stat:
+            trainer.logger.log_metrics(latest_stat, step=trainer.global_step)
+
+    def _extract_lr(self, trainer, interval):
+        """ Extracts learning rates for lr schedulers and saves information
+            into dict structure. """
+        latest_stat = {}
+        for name, scheduler in zip(self.names, trainer.lr_schedulers):
+            if scheduler['interval'] == interval:
+                param_groups = scheduler['scheduler'].optimizer.param_groups
+                if len(param_groups) != 1:
+                    for i, pg in enumerate(param_groups):
+                        lr = pg['lr']
+                        key = f'{name}/{i + 1}'
+                        self.lrs[key].append(lr)
+                        latest_stat[key] = lr
+                else:
+                    self.lrs[name].append(param_groups[0]['lr'])
+                    latest_stat[name] = param_groups[0]['lr']
+        return latest_stat

tests/trainer/test_callbacks.py

@@ -1,11 +1,12 @@
 import tests.base.utils as tutils
 from pytorch_lightning import Callback
 from pytorch_lightning import Trainer, LightningModule
-from pytorch_lightning.callbacks import EarlyStopping
+from pytorch_lightning.callbacks import EarlyStopping, LearningRateLogger
 from tests.base import (
     LightTrainDataloader,
     LightTestMixin,
     LightValidationMixin,
+    LightTestOptimizersWithMixedSchedulingMixin,
     TestModelBase
 )
 
@@ -181,3 +182,55 @@ def training_step(self, *args, **kwargs):
 
     assert result == 1, 'training failed to complete'
     assert trainer.current_epoch < trainer.max_epochs
+
+
+def test_lr_logger_single_lr(tmpdir):
+    """ Test that learning rates are extracted and logged for single lr scheduler"""
+    tutils.reset_seed()
+
+    class CurrentTestModel(LightTrainDataloader, TestModelBase):
+        pass
+
+    hparams = tutils.get_default_hparams()
+    model = CurrentTestModel(hparams)
+
+    lr_logger = LearningRateLogger()
+    trainer = Trainer(
+        default_root_dir=tmpdir,
+        max_epochs=5,
+        val_percent_check=0.1,
+        train_percent_check=0.5,
+        callbacks=[lr_logger]
+    )
+    results = trainer.fit(model)
+
+    assert lr_logger.lrs, 'No learning rates logged'
+    assert len(lr_logger.lrs) == len(trainer.lr_schedulers), \
+        'Number of learning rates logged does not match number of lr schedulers'
+
+
+def test_lr_logger_multi_lrs(tmpdir):
+    """ Test that learning rates are extracted and logged for multi lr schedulers """
+    tutils.reset_seed()
+
+    class CurrentTestModel(LightTestOptimizersWithMixedSchedulingMixin,
+                           LightTrainDataloader,
+                           TestModelBase):
+        pass
+
+    hparams = tutils.get_default_hparams()
+    model = CurrentTestModel(hparams)
+
+    lr_logger = LearningRateLogger()
+    trainer = Trainer(
+        default_root_dir=tmpdir,
+        max_epochs=1,
+        val_percent_check=0.1,
+        train_percent_check=0.5,
+        callbacks=[lr_logger]
+    )
+    results = trainer.fit(model)
+
+    assert lr_logger.lrs, 'No learning rates logged'
+    assert len(lr_logger.lrs) == len(trainer.lr_schedulers), \
+        'Number of learning rates logged does not match number of lr schedulers'