docs cleaning - testcode (#5595)

* testcode - python

* revert

* simple

* testcode @rst

* pl

* fix

* pip

* update

* conf

* conf

* nn.

* typo

.circleci/config.yml

@@ -18,7 +18,10 @@ references:
         pyenv global 3.7.3
         python --version
         pip install -r requirements/docs.txt
-        cd docs; make clean; make html --debug --jobs 2 SPHINXOPTS="-W"
+        pip list
+        cd docs
+        make clean
+        make html --jobs 2 SPHINXOPTS="-W"
 
   checkout_ml_testing: &checkout_ml_testing
    run:

docs/source/conf.py

@@ -25,9 +25,9 @@
 PATH_ROOT = os.path.join(PATH_HERE, '..', '..')
 sys.path.insert(0, os.path.abspath(PATH_ROOT))
 
-builtins.__LIGHTNING_SETUP__ = True
-
 SPHINX_MOCK_REQUIREMENTS = int(os.environ.get('SPHINX_MOCK_REQUIREMENTS', True))
+if SPHINX_MOCK_REQUIREMENTS:
+    builtins.__LIGHTNING_SETUP__ = True
 
 import pytorch_lightning  # noqa: E402
 
@@ -360,7 +360,10 @@ def package_list_from_file(file):
 import importlib
 import os
 import torch
+from torch import nn
 
+import pytorch_lightning as pl
+from pytorch_lightning import LightningDataModule, LightningModule, Trainer
 from pytorch_lightning.utilities import (
     _NATIVE_AMP_AVAILABLE,
     _APEX_AVAILABLE,
@@ -369,6 +372,5 @@ def package_list_from_file(file):
 )
 _TORCHVISION_AVAILABLE = importlib.util.find_spec("torchvision") is not None
 
-
 """
 coverage_skip_undoc_in_source = True

docs/source/converting.rst

@@ -24,8 +24,8 @@ Move the model architecture and forward pass to your :ref:`lightning_module`.
 
         def __init__(self):
             super().__init__()
-            self.layer_1 = torch.nn.Linear(28 * 28, 128)
-            self.layer_2 = torch.nn.Linear(128, 10)
+            self.layer_1 = nn.Linear(28 * 28, 128)
+            self.layer_2 = nn.Linear(128, 10)
 
         def forward(self, x):
             x = x.view(x.size(0), -1)

docs/source/early_stopping.rst

@@ -49,7 +49,7 @@ To enable it:
 
 -   You can customize the callbacks behaviour by changing its parameters.
 
-.. code-block:: python
+.. testcode::
 
     early_stop_callback = EarlyStopping(
        monitor='val_accuracy',

docs/source/hyperparameters.rst

@@ -161,9 +161,9 @@ improve readability and reproducibility.
             def __init__(self, hparams, *args, **kwargs):
                 super().__init__()
                 self.hparams = hparams
-                self.layer_1 = torch.nn.Linear(28 * 28, self.hparams.layer_1_dim)
-                self.layer_2 = torch.nn.Linear(self.hparams.layer_1_dim, self.hparams.layer_2_dim)
-                self.layer_3 = torch.nn.Linear(self.hparams.layer_2_dim, 10)
+                self.layer_1 = nn.Linear(28 * 28, self.hparams.layer_1_dim)
+                self.layer_2 = nn.Linear(self.hparams.layer_1_dim, self.hparams.layer_2_dim)
+                self.layer_3 = nn.Linear(self.hparams.layer_2_dim, 10)
             def train_dataloader(self):
                 return DataLoader(mnist_train, batch_size=self.hparams.batch_size)
 
@@ -182,9 +182,9 @@ improve readability and reproducibility.
                 super().__init__()
                 self.save_hyperparameters(conf)
 
-                self.layer_1 = torch.nn.Linear(28 * 28, self.hparams.layer_1_dim)
-                self.layer_2 = torch.nn.Linear(self.hparams.layer_1_dim, self.hparams.layer_2_dim)
-                self.layer_3 = torch.nn.Linear(self.hparams.layer_2_dim, 10)
+                self.layer_1 = nn.Linear(28 * 28, self.hparams.layer_1_dim)
+                self.layer_2 = nn.Linear(self.hparams.layer_1_dim, self.hparams.layer_2_dim)
+                self.layer_3 = nn.Linear(self.hparams.layer_2_dim, 10)
 
         conf = OmegaConf.create(...)
         model = LitMNIST(conf)
@@ -225,7 +225,7 @@ polluting the ``main.py`` file, the ``LightningModule`` lets you define argument
 
         def __init__(self, layer_1_dim, **kwargs):
             super().__init__()
-            self.layer_1 = torch.nn.Linear(28 * 28, layer_1_dim)
+            self.layer_1 = nn.Linear(28 * 28, layer_1_dim)
 
         @staticmethod
         def add_model_specific_args(parent_parser):

docs/source/introduction_guide.rst

@@ -80,9 +80,9 @@ Let's first start with the model. In this case, we'll design a 3-layer neural ne
         super().__init__()
 
         # mnist images are (1, 28, 28) (channels, width, height)
-        self.layer_1 = torch.nn.Linear(28 * 28, 128)
-        self.layer_2 = torch.nn.Linear(128, 256)
-        self.layer_3 = torch.nn.Linear(256, 10)
+        self.layer_1 = nn.Linear(28 * 28, 128)
+        self.layer_2 = nn.Linear(128, 256)
+        self.layer_3 = nn.Linear(256, 10)
 
       def forward(self, x):
         batch_size, channels, width, height = x.size()
@@ -118,7 +118,7 @@ equivalent to a pure PyTorch Module except it has added functionality. However,
 
 Now we add the training_step which has all our training loop logic
 
-.. testcode:: python
+.. testcode::
 
     class LitMNIST(LightningModule):
 
@@ -225,7 +225,7 @@ In this case, it's better to group the full definition of a dataset into a `Data
 - Val dataloader(s)
 - Test dataloader(s)
 
-.. testcode:: python
+.. testcode::
 
     class MyDataModule(LightningDataModule):
 
@@ -420,9 +420,9 @@ For clarity, we'll recall that the full LightningModule now looks like this.
     class LitMNIST(LightningModule):
         def __init__(self):
             super().__init__()
-            self.layer_1 = torch.nn.Linear(28 * 28, 128)
-            self.layer_2 = torch.nn.Linear(128, 256)
-            self.layer_3 = torch.nn.Linear(256, 10)
+            self.layer_1 = nn.Linear(28 * 28, 128)
+            self.layer_2 = nn.Linear(128, 256)
+            self.layer_3 = nn.Linear(256, 10)
 
         def forward(self, x):
             batch_size, channels, width, height = x.size()

docs/source/lightning_module.rst

@@ -96,7 +96,7 @@ Here are the only required methods.
     ...
     ...     def __init__(self):
     ...         super().__init__()
-    ...         self.l1 = torch.nn.Linear(28 * 28, 10)
+    ...         self.l1 = nn.Linear(28 * 28, 10)
     ...
     ...     def forward(self, x):
     ...         return torch.relu(self.l1(x.view(x.size(0), -1)))

docs/source/metrics.rst

@@ -141,9 +141,11 @@ This metrics API is independent of PyTorch Lightning. Metrics can directly be us
 
     .. testcode::
 
+        from pytorch_lightning.metrics import Accuracy
+
         def __init__(self):
             ...
-            metric = pl.metrics.Accuracy()
+            metric = Accuracy()
             self.train_acc = metric.clone()
             self.val_acc = metric.clone()
             self.test_acc = metric.clone()
@@ -164,7 +166,6 @@ be moved to the same device as the input of the metric:
 
 .. code-block:: python
 
-    import torch
     from pytorch_lightning.metrics import Accuracy
 
     target = torch.tensor([1, 1, 0, 0], device=torch.device("cuda", 0))
@@ -186,13 +187,15 @@ as child modules. Instead of ``list`` use :class:`~torch.nn.ModuleList` and inst
 
 .. testcode::
 
+    from pytorch_lightning.metrics import Accuracy
+
     class MyModule(LightningModule):
         def __init__(self):
             ...
             # valid ways metrics will be identified as child modules
-            self.metric1 = pl.metrics.Accuracy()
-            self.metric2 = torch.nn.ModuleList(pl.metrics.Accuracy())
-            self.metric3 = torch.nn.ModuleDict({'accuracy': Accuracy()})
+            self.metric1 = Accuracy()
+            self.metric2 = nn.ModuleList(Accuracy())
+            self.metric3 = nn.ModuleDict({'accuracy': Accuracy()})
 
         def training_step(self, batch, batch_idx):
             # all metrics will be on the same device as the input batch
@@ -222,7 +225,7 @@ from the base ``Metric`` class.
 
 Example implementation:
 
-.. code-block:: python
+.. testcode::
 
     from pytorch_lightning.metrics import Metric
 
@@ -281,8 +284,8 @@ Example:
 .. testoutput::
     :options: +NORMALIZE_WHITESPACE
 
-    {'Accuracy': tensor(0.1250), 
-     'Precision': tensor(0.0667), 
+    {'Accuracy': tensor(0.1250),
+     'Precision': tensor(0.0667),
      'Recall': tensor(0.1111)}
 
 Similarly it can also reduce the amount of code required to log multiple metrics

docs/source/multi_gpu.rst

@@ -698,7 +698,7 @@ This should be kept within the ``sequential_module`` variable within your ``Ligh
     class MyModel(LightningModule):
         def __init__(self):
             ...
-            self.sequential_module = torch.nn.Sequential(my_layers)
+            self.sequential_module = nn.Sequential(my_layers)
 
     # Split my module across 4 gpus, one layer each
     model = MyModel()