Examples: using new API

README.md

@@ -145,7 +145,7 @@ To use lightning do 2 things:
             y_hat = self.forward(x)
             return {'val_loss': F.cross_entropy(y_hat, y)}
 
-        def validation_end(self, outputs):
+        def validation_epoch_end(self, outputs):
             # OPTIONAL
             avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()
             tensorboard_logs = {'val_loss': avg_loss}
@@ -157,7 +157,7 @@ To use lightning do 2 things:
             y_hat = self.forward(x)
             return {'test_loss': F.cross_entropy(y_hat, y)}
 
-        def test_end(self, outputs):
+        def test_epoch_end(self, outputs):
             # OPTIONAL
             avg_loss = torch.stack([x['test_loss'] for x in outputs]).mean()
             tensorboard_logs = {'test_loss': avg_loss}
@@ -268,7 +268,7 @@ def validation_step(self, batch, batch_idx):
 **And you also decide how to collate the output of all validation steps**    
 
 ```python
-def validation_end(self, outputs):
+def validation_epoch_end(self, outputs):
     """
     Called at the end of validation to aggregate outputs
     :param outputs: list of individual outputs of each validation step

docs/source/early_stopping.rst

@@ -4,7 +4,7 @@ Early stopping
 Default behavior
 ----------------
 By default early stopping will be enabled if `'val_loss'`
-is found in `validation_end()` return dict. Otherwise
+is found in `validation_epoch_end()` return dict. Otherwise
 training will proceed with early stopping disabled.
 
 Enable Early Stopping
@@ -16,7 +16,7 @@ There are two ways to enable early stopping.
 .. code-block:: python
 
     # A) Set early_stop_callback to True. Will look for 'val_loss'
-    # in validation_end() return dict. If it is not found an error is raised.
+    # in validation_epoch_end() return dict. If it is not found an error is raised.
     trainer = Trainer(early_stop_callback=True)
 
     # B) Or configure your own callback

docs/source/experiment_reporting.rst

@@ -87,7 +87,7 @@ Here we show the validation loss in the progress bar
 
 .. code-block:: python
 
-   def validation_end(self, outputs):
+   def validation_epoch_end(self, outputs):
       loss = some_loss()
       ...
 

docs/source/introduction_guide.rst

@@ -603,7 +603,7 @@ sample split in the `train_dataloader` method.
         loss = F.nll_loss(logits, y)
         return {'val_loss': loss}
 
-      def validation_end(self, outputs):
+      def validation_epoch_end(self, outputs):
         avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()
         tensorboard_logs = {'val_loss': avg_loss}
         return {'avg_val_loss': avg_loss, 'log': tensorboard_logs}
@@ -657,7 +657,7 @@ Just like the validation loop, we define exactly the same steps for testing:
         loss = F.nll_loss(logits, y)
         return {'val_loss': loss}
 
-      def test_end(self, outputs):
+      def test_epoch_end(self, outputs):
         avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()
         tensorboard_logs = {'val_loss': avg_loss}
         return {'avg_val_loss': avg_loss, 'log': tensorboard_logs}

pl_examples/basic_examples/lightning_module_template.py

@@ -143,7 +143,7 @@ def validation_step(self, batch, batch_idx):
         # can also return just a scalar instead of a dict (return loss_val)
         return output
 
-    def validation_end(self, outputs):
+    def validation_epoch_end(self, outputs):
         """
         Called at the end of validation to aggregate outputs
         :param outputs: list of individual outputs of each validation step

pl_examples/full_examples/imagenet/imagenet_example.py

@@ -84,7 +84,7 @@ def validation_step(self, batch, batch_idx):
 
         return output
 
-    def validation_end(self, outputs):
+    def validation_epoch_end(self, outputs):
 
         tqdm_dict = {}
 

tests/models/debug.py

@@ -34,7 +34,7 @@ def validation_step(self, batch, batch_idx):
         y_hat = self.forward(x)
         return {'val_loss': self.my_loss(y_hat, y)}
 
-    def validation_end(self, outputs):
+    def validation_epoch_end(self, outputs):
         avg_loss = torch.stack([x for x in outputs['val_loss']]).mean()
         return avg_loss
 

tests/models/mixins.py

@@ -2,7 +2,6 @@
 
 import torch
 from torch import optim
-from pytorch_lightning.core.decorators import data_loader
 
 
 class LightValidationStepMixin:
@@ -64,7 +63,7 @@ class LightValidationMixin(LightValidationStepMixin):
     when val_dataloader returns a single dataloader
     """
 
-    def validation_end(self, outputs):
+    def validation_epoch_end(self, outputs):
         """
         Called at the end of validation to aggregate outputs
         :param outputs: list of individual outputs of each validation step
@@ -163,7 +162,7 @@ class LightValidationMultipleDataloadersMixin(LightValidationStepMultipleDataloa
     when val_dataloader returns multiple dataloaders
     """
 
-    def validation_end(self, outputs):
+    def validation_epoch_end(self, outputs):
         """
         Called at the end of validation to aggregate outputs
         :param outputs: list of individual outputs of each validation step
@@ -271,7 +270,7 @@ def test_step(self, batch, batch_idx, *args, **kwargs):
 class LightTestMixin(LightTestStepMixin):
     """Ritch test mixin."""
 
-    def test_end(self, outputs):
+    def test_epoch_end(self, outputs):
         """
         Called at the end of validation to aggregate outputs
         :param outputs: list of individual outputs of each validation step
@@ -561,7 +560,7 @@ def validation_step(self, batch, batch_idx, dataloader_idx, **kwargs):
 
 class LightTestMultipleDataloadersMixin(LightTestStepMultipleDataloadersMixin):
 
-    def test_end(self, outputs):
+    def test_epoch_end(self, outputs):
         """
         Called at the end of validation to aggregate outputs
         :param outputs: list of individual outputs of each validation step

tests/trainer/test_trainer.py

@@ -604,7 +604,7 @@ class LocalModelNoEnd(LightTrainDataloader, LightTestDataloader, LightEmptyTestS
         pass
 
     class LocalModelNoStep(LightTrainDataloader, TestModelBase):
-        def test_end(self, outputs):
+        def test_epoch_end(self, outputs):
             return {}
 
     # Misconfig when neither test_step or test_end is implemented