Merge branch 'master' into bugfix/5487_auto_lr_ordering

docs/source/conf.py

@@ -26,9 +26,9 @@
 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
 
@@ -361,15 +361,18 @@ 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 LightningModule, Trainer
 from pytorch_lightning.utilities import (
     NATIVE_AMP_AVAILABLE,
     APEX_AVAILABLE,
     XLA_AVAILABLE,
     TPU_AVAILABLE,
+    _module_available,
 )
-TORCHVISION_AVAILABLE = importlib.util.find_spec("torchvision") is not None
-
+TORCHVISION_AVAILABLE = _module_available("torchvision")
 
 """
 coverage_skip_undoc_in_source = True

docs/source/style_guide.rst

@@ -35,16 +35,16 @@ A LightningModule can define both a system and a model.
 
 Here's a LightningModule that defines a model:
 
-.. code-block:: python
+.. testcode::
 
-    class LitModel(pl.LightningModule):
-        def __init__(self, num_layers: int = 3)
+    class LitModel(LightningModule):
+        def __init__(self, num_layers: int = 3):
             super().__init__()
-            self.layer_1 = nn.Linear(...)
-            self.layer_2 = nn.Linear(...)
-            self.layer_3 = nn.Linear(...)
+            self.layer_1 = nn.Linear()
+            self.layer_2 = nn.Linear()
+            self.layer_3 = nn.Linear()
 
-Here's a lightningModule that defines a system:
+Here's a LightningModule that defines a system:
 
 .. code-block:: python
 
@@ -74,9 +74,9 @@ sensible defaults in the init so that the user doesn't have to guess.
 
 Here's an example where a user will have to go hunt through files to figure out how to init this LightningModule.
 
-.. code-block:: python
+.. testcode::
 
-    class LitModel(pl.LightningModule):
+    class LitModel(LightningModule):
         def __init__(self, params):
             self.lr = params.lr
             self.coef_x = params.coef_x
@@ -85,10 +85,11 @@ Models defined as such leave you with many questions; what is coef_x? is it a st
 
 Instead, be explicit in your init
 
-.. code-block:: python
+.. testcode::
 
     class LitModel(pl.LightningModule):
-        def __init__(self, encoder: nn.Module, coeff_x: float = 0.2, lr: float = 1e-3)
+        def __init__(self, encoder: nn.Module, coef_x: float = 0.2, lr: float = 1e-3):
+            pass
 
 Now the user doesn't have to guess. Instead they know the value type and the model has a sensible default where the
 user can see the value immediately.

pytorch_lightning/core/lightning.py

@@ -1501,7 +1501,7 @@ def save_hyperparameters(self, *args, frame=None) -> None:
 
         Args:
             args: single object of `dict`, `NameSpace` or `OmegaConf`
-             or string names or argumenst from class `__init__`
+             or string names or arguments from class `__init__`
 
         >>> from collections import OrderedDict
         >>> class ManuallyArgsModel(LightningModule):