rename valid_dataloader [discussion]

README.md

@@ -141,7 +141,7 @@ To use lightning do 2 things:
             return DataLoader(MNIST(os.getcwd(), train=True, download=True, transform=transforms.ToTensor()), batch_size=32)
 
         @pl.data_loader
-        def val_dataloader(self):
+        def valid_dataloader(self):
             # OPTIONAL
             return DataLoader(MNIST(os.getcwd(), train=True, download=True, transform=transforms.ToTensor()), batch_size=32)
 

pl_examples/basic_examples/lightning_module_template.py

@@ -219,7 +219,7 @@ def train_dataloader(self):
         return self.__dataloader(train=True)
 
     @pl.data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         logging.info('val data loader called')
         return self.__dataloader(train=False)
 

pl_examples/full_examples/imagenet/imagenet_example.py

@@ -32,7 +32,7 @@
 class ImageNetLightningModel(pl.LightningModule):
 
     def __init__(self, hparams):
-        super(ImageNetLightningModel, self).__init__()
+        super(ImageNetLightningModel, self).__idatasetdatasetnit__()
         self.hparams = hparams
         self.model = models.__dict__[self.hparams.arch](pretrained=self.hparams.pretrained)
 
@@ -159,7 +159,7 @@ def train_dataloader(self):
         return train_loader
 
     @pl.data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         normalize = transforms.Normalize(
             mean=[0.485, 0.456, 0.406],
             std=[0.229, 0.224, 0.225],

pytorch_lightning/core/__init__.py

@@ -72,7 +72,7 @@ def train_dataloader(self):
                               transform=transforms.ToTensor()), batch_size=32)
 
         @pl.data_loader
-        def val_dataloader(self):
+        def valid_dataloader(self):
             # OPTIONAL
             # can also return a list of val dataloaders
             return DataLoader(MNIST(os.getcwd(), train=True, download=True,

pytorch_lightning/core/decorators.py

@@ -19,7 +19,7 @@ def _get_data_loader(self):
                 if (
                         value is not None and
                         not isinstance(value, list) and
-                        fn.__name__ in ['test_dataloader', 'val_dataloader']
+                        fn.__name__ in ['test_dataloader', 'valid_dataloader']
                 ):
                     value = [value]
             except AttributeError as e:

pytorch_lightning/core/lightning.py

@@ -310,7 +310,7 @@ def validation_step(self, batch, batch_idx):
             def validation_step(self, batch, batch_idx, dataset_idx):
                 # dataset_idx tells you which dataset this is.
 
-        The `dataset_idx` corresponds to the order of datasets returned in `val_dataloader`.
+        The `dataset_idx` corresponds to the order of datasets returned in `valid_dataloader`.
         """
         pass
 
@@ -850,13 +850,10 @@ def tng_dataloader(self):
 
         .. warning:: Deprecated in v0.5.0. use train_dataloader instead.
         """
-        try:
-            output = self.tng_dataloader()
-            warnings.warn("`tng_dataloader` has been renamed to `train_dataloader` since v0.5.0"
-                          " and will be removed in v0.8.0", DeprecationWarning)
-            return output
-        except NotImplementedError:
-            raise NotImplementedError
+        output = self.train_dataloader()
+        warnings.warn("`tng_dataloader` has been renamed to `train_dataloader` since v0.5.0"
+                      " and will be removed in v0.8.0", DeprecationWarning)
+        return output
 
     @data_loader
     def test_dataloader(self):
@@ -891,7 +888,7 @@ def test_dataloader(self):
         return None
 
     @data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         """Implement a PyTorch DataLoader.
 
         :return: PyTorch DataLoader or list of PyTorch Dataloaders.
@@ -908,7 +905,7 @@ def val_dataloader(self):
         .. code-block:: python
 
             @pl.data_loader
-            def val_dataloader(self):
+            def valid_dataloader(self):
                 transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (1.0,))])
                 dataset = MNIST(root='/path/to/mnist/', train=False, transform=transform, download=True)
                 loader = torch.utils.data.DataLoader(
@@ -921,14 +918,25 @@ def val_dataloader(self):
 
             # can also return multiple dataloaders
             @pl.data_loader
-            def val_dataloader(self):
+            def valid_dataloader(self):
                 return [loader_a, loader_b, ..., loader_n]
 
-        In the case where you return multiple `val_dataloaders`, the `validation_step`
+        In the case where you return multiple `valid_dataloaders`, the `validation_step`
          will have an arguement `dataset_idx` which matches the order here.
         """
         return None
 
+    @data_loader
+    def val_dataloader(self):
+        """Implement a PyTorch DataLoader.
+
+        .. warning:: Deprecated in v0.5.0. use valid_dataloader instead.
+        """
+        output = self.valid_dataloader()
+        warnings.warn("`val_dataloader` has been renamed to `valid_dataloader` since v0.5.0"
+                      " and will be removed in v0.8.0", DeprecationWarning)
+        return output
+
     @classmethod
     def load_from_metrics(cls, weights_path, tags_csv, map_location=None):
         """Primary way of loading model from csv weights path.

pytorch_lightning/testing/model_mixins.py

@@ -7,12 +7,12 @@
 
 class LightningValidationStepMixin:
     """
-    Add val_dataloader and validation_step methods for the case
-    when val_dataloader returns a single dataloader
+    Add valid_dataloader and validation_step methods for the case
+    when valid_dataloader returns a single dataloader
     """
 
     @data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         return self._dataloader(train=False)
 
     def validation_step(self, batch, batch_idx):
@@ -61,8 +61,8 @@ def validation_step(self, batch, batch_idx):
 
 class LightningValidationMixin(LightningValidationStepMixin):
     """
-    Add val_dataloader, validation_step, and validation_end methods for the case
-    when val_dataloader returns a single dataloader
+    Add valid_dataloader, validation_step, and validation_end methods for the case
+    when valid_dataloader returns a single dataloader
     """
 
     def validation_end(self, outputs):
@@ -101,12 +101,12 @@ def validation_end(self, outputs):
 
 class LightningValidationStepMultipleDataloadersMixin:
     """
-    Add val_dataloader and validation_step methods for the case
-    when val_dataloader returns multiple dataloaders
+    Add valid_dataloader and validation_step methods for the case
+    when valid_dataloader returns multiple dataloaders
     """
 
     @data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         return [self._dataloader(train=False), self._dataloader(train=False)]
 
     def validation_step(self, batch, batch_idx, dataloader_idx):
@@ -161,8 +161,8 @@ def validation_step(self, batch, batch_idx, dataloader_idx):
 
 class LightningValidationMultipleDataloadersMixin(LightningValidationStepMultipleDataloadersMixin):
     """
-    Add val_dataloader, validation_step, and validation_end methods for the case
-    when val_dataloader returns multiple dataloaders
+    Add valid_dataloader, validation_step, and validation_end methods for the case
+    when valid_dataloader returns multiple dataloaders
     """
 
     def validation_end(self, outputs):

pytorch_lightning/trainer/data_loading.py

@@ -82,27 +82,27 @@ def init_train_dataloader(self, model):
                 self.shown_warnings.add(msg)
                 warnings.warn(msg)
 
-    def init_val_dataloader(self, model):
+    def init_valid_dataloader(self, model):
         """
         Dataloaders are provided by the model
         :param model:
         :return:
         """
-        self.get_val_dataloaders = model.val_dataloader
+        self.get_valid_dataloaders = model.valid_dataloader
 
         # determine number of validation batches
         # val datasets could be none, 1 or 2+
-        if self.get_val_dataloaders() is not None:
-            self.num_val_batches = sum(len(dataloader) for dataloader in self.get_val_dataloaders())
+        if self.get_valid_dataloaders() is not None:
+            self.num_val_batches = sum(len(dataloader) for dataloader in self.get_valid_dataloaders())
             self.num_val_batches = int(self.num_val_batches * self.val_percent_check)
             self.num_val_batches = max(1, self.num_val_batches)
 
         on_ddp = self.use_ddp or self.use_ddp2
-        if on_ddp and self.get_val_dataloaders() is not None:
-            for dataloader in self.get_val_dataloaders():
+        if on_ddp and self.get_valid_dataloaders() is not None:
+            for dataloader in self.get_valid_dataloaders():
                 if not isinstance(dataloader.sampler, DistributedSampler):
                     msg = """
-                    Your val_dataloader(s) don't use DistributedSampler.
+                    Your valid_dataloader(s) don't use DistributedSampler.
 
                     You're using multiple gpus and multiple nodes without using a
                     DistributedSampler to assign a subset of your data to each process.
@@ -177,7 +177,7 @@ def get_dataloaders(self, model):
 
         self.init_train_dataloader(model)
         self.init_test_dataloader(model)
-        self.init_val_dataloader(model)
+        self.init_valid_dataloader(model)
 
         if self.use_ddp or self.use_ddp2:
             # wait for all processes to catch up
@@ -186,7 +186,7 @@ def get_dataloaders(self, model):
             # load each dataloader
             self.get_train_dataloader()
             self.get_test_dataloaders()
-            self.get_val_dataloaders()
+            self.get_valid_dataloaders()
 
         # support IterableDataset for train data
         self.is_iterable_train_dataloader = (

pytorch_lightning/trainer/evaluation_loop.py

@@ -157,7 +157,7 @@ def __init__(self):
         self.current_epoch = None
         self.callback_metrics = None
         self.get_test_dataloaders = None
-        self.get_val_dataloaders = None
+        self.get_valid_dataloaders = None
 
     @abstractmethod
     def copy_trainer_model_properties(self, model):
@@ -290,7 +290,7 @@ def run_evaluation(self, test=False):
                 max_batches = self.num_test_batches
             else:
                 # val
-                dataloaders = self.get_val_dataloaders()
+                dataloaders = self.get_valid_dataloaders()
                 max_batches = self.num_val_batches
 
             # cap max batches to 1 when using fast_dev_run
@@ -349,7 +349,7 @@ def evaluation_forward(self, model, batch, batch_idx, dataloader_idx, test=False
         if test and len(self.get_test_dataloaders()) > 1:
             args.append(dataloader_idx)
 
-        elif not test and len(self.get_val_dataloaders()) > 1:
+        elif not test and len(self.get_valid_dataloaders()) > 1:
             args.append(dataloader_idx)
 
         # handle DP, DDP forward

pytorch_lightning/trainer/trainer.py

@@ -207,7 +207,7 @@ def __init__(
         self.num_test_batches = 0
         self.get_train_dataloader = None
         self.get_test_dataloaders = None
-        self.get_val_dataloaders = None
+        self.get_valid_dataloaders = None
         self.is_iterable_train_dataloader = False
 
         # training state
@@ -490,17 +490,17 @@ def run_pretrain_routine(self, model):
         # to make sure program won't crash during val
         ref_model.on_sanity_check_start()
         ref_model.on_train_start()
-        if self.get_val_dataloaders() is not None and self.num_sanity_val_steps > 0:
+        if self.get_valid_dataloaders() is not None and self.num_sanity_val_steps > 0:
             # init progress bars for validation sanity check
             pbar = tqdm.tqdm(desc='Validation sanity check',
-                             total=self.num_sanity_val_steps * len(self.get_val_dataloaders()),
+                             total=self.num_sanity_val_steps * len(self.get_valid_dataloaders()),
                              leave=False, position=2 * self.process_position,
                              disable=not self.show_progress_bar, dynamic_ncols=True, unit='batch')
             self.main_progress_bar = pbar
             # dummy validation progress bar
             self.val_progress_bar = tqdm.tqdm(disable=True)
 
-            self.evaluate(model, self.get_val_dataloaders(), self.num_sanity_val_steps, self.testing)
+            self.evaluate(model, self.get_valid_dataloaders(), self.num_sanity_val_steps, self.testing)
 
             # close progress bars
             self.main_progress_bar.close()

tests/debug.py

@@ -46,7 +46,7 @@ def train_dataloader(self):
         return DataLoader(MNIST('path/to/save', train=True), batch_size=32)
 
     @pl.data_loader
-    def val_dataloader(self):
+    def valid_dataloader(self):
         return DataLoader(MNIST('path/to/save', train=False), batch_size=32)
 
     @pl.data_loader

tests/test_restore_models.py

@@ -308,7 +308,7 @@ def assert_good_acc():
         # if model and state loaded correctly, predictions will be good even though we
         # haven't trained with the new loaded model
         trainer.model.eval()
-        for dataloader in trainer.get_val_dataloaders():
+        for dataloader in trainer.get_valid_dataloaders():
             tutils.run_prediction(dataloader, trainer.model)
 
     model.on_train_start = assert_good_acc

tests/test_trainer.py

@@ -338,8 +338,8 @@ def test_model_freeze_unfreeze():
     model.unfreeze()
 
 
-def test_multiple_val_dataloader(tmpdir):
-    """Verify multiple val_dataloader."""
+def test_multiple_valid_dataloader(tmpdir):
+    """Verify multiple valid_dataloader."""
     tutils.reset_seed()
 
     class CurrentTestModel(
@@ -367,11 +367,11 @@ class CurrentTestModel(
     assert result == 1
 
     # verify there are 2 val loaders
-    assert len(trainer.get_val_dataloaders()) == 2, \
-        'Multiple val_dataloaders not initiated properly'
+    assert len(trainer.get_valid_dataloaders()) == 2, \
+        'Multiple valid_dataloaders not initiated properly'
 
     # make sure predictions are good for each val set
-    for dataloader in trainer.get_val_dataloaders():
+    for dataloader in trainer.get_valid_dataloaders():
         tutils.run_prediction(dataloader, trainer.model)