docformatter: config with black

.actions/assistant.py

@@ -86,6 +86,7 @@ def adjust(self, unfreeze: str) -> str:
         'arrow>=1.2.0'
         >>> _RequirementWithComment("arrow").adjust("major")
         'arrow'
+
         """
         out = str(self)
         if self.strict:
@@ -115,6 +116,7 @@ def _parse_requirements(strs: Union[str, Iterable[str]]) -> Iterator[_Requiremen
     >>> txt = '\\n'.join(txt)
     >>> [r.adjust('none') for r in _parse_requirements(txt)]
     ['this', 'example', 'foo  # strict', 'thing']
+
     """
     lines = yield_lines(strs)
     pip_argument = None
@@ -149,6 +151,7 @@ def load_requirements(path_dir: str, file_name: str = "base.txt", unfreeze: str
     >>> path_req = os.path.join(_PROJECT_ROOT, "requirements")
     >>> load_requirements(path_req, "docs.txt", unfreeze="major")  # doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
     ['sphinx<...]
+
     """
     assert unfreeze in {"none", "major", "all"}
     path = Path(path_dir) / file_name
@@ -165,6 +168,7 @@ def load_readme_description(path_dir: str, homepage: str, version: str) -> str:
 
     >>> load_readme_description(_PROJECT_ROOT, "", "")  # doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
     '...PyTorch Lightning is just organized PyTorch...'
+
     """
     path_readme = os.path.join(path_dir, "README.md")
     with open(path_readme, encoding="utf-8") as fo:
@@ -244,6 +248,7 @@ def _load_aggregate_requirements(req_dir: str = "requirements", freeze_requireme
     """Load all base requirements from all particular packages and prune duplicates.
 
     >>> _load_aggregate_requirements(os.path.join(_PROJECT_ROOT, "requirements"))
+
     """
     requires = [
         load_requirements(d, unfreeze="none" if freeze_requirements else "major")
@@ -300,6 +305,7 @@ def _replace_imports(lines: List[str], mapping: List[Tuple[str, str]], lightning
      'http://pytorch_lightning.ai', \
      'from lightning_fabric import __version__', \
      '@lightning.ai']
+
     """
     out = lines[:]
     for source_import, target_import in mapping:

.pre-commit-config.yaml

@@ -60,7 +60,8 @@ repos:
     rev: v1.7.3
     hooks:
       - id: docformatter
-        args: [--in-place, --wrap-summaries=115, --wrap-descriptions=120]
+        additional_dependencies: [tomli]
+        args: ["--in-place"]
 
   - repo: https://github.com/asottile/yesqa
     rev: v1.5.0

docs/source-app/examples/file_server/app.py

@@ -23,6 +23,7 @@ def __init__(
             drive: The drive can share data inside your application.
             base_dir: The local directory where the data will be stored.
             chunk_size: The quantity of bytes to download/upload at once.
+
         """
         super().__init__(
             cloud_build_config=L.BuildConfig(["flask, flask-cors"]),
@@ -238,4 +239,5 @@ def test_file_server_in_cloud():
 
         # 2. By calling logs = get_logs_fn(),
         # you get all the logs currently on the admin page.
+
         """

docs/source-app/examples/github_repo_runner/app.py

@@ -36,6 +36,7 @@ def __init__(
             script_args: The arguments to be provided to the script.
             requirements: The python requirements tp run the script.
             cloud_compute: The object to select the cloud instance.
+
         """
         super().__init__(
             script_path=script_path,

docs/source-app/examples/model_server_app/locust_component.py

@@ -10,6 +10,7 @@ def __init__(self, num_users: int = 100):
 
         Arguments:
             num_users: Number of users emulated by Locust
+
         """
         # Note: Using the default port 8089 of Locust.
         super().__init__(

docs/source-app/examples/model_server_app/model_server.py

@@ -18,6 +18,7 @@ class MLServer(LightningWork):
             Example: "mlserver_sklearn.SKLearnModel".
             Learn more here: $ML_SERVER_URL/tree/master/runtimes
         workers: Number of server worker.
+
     """
 
     def __init__(
@@ -51,6 +52,7 @@ def run(self, model_path: Path):
 
         Arguments:
             model_path: The path to the trained model.
+
         """
         # 1: Use the host and port at runtime so it works in the cloud.
         # $ML_SERVER_URL/blob/master/mlserver/settings.py#L50

examples/app/hpo/utils.py

@@ -15,6 +15,7 @@ def download_data(url: str, path: str = "data/", verbose: bool = False) -> None:
 
     Usage:
         download_file('http://web4host.net/5MB.zip')
+
     """
     if url == "NEED_TO_BE_CREATED":
         raise NotImplementedError

examples/app/layout/app.py

@@ -5,6 +5,7 @@
 lightning run app examples/layout/demo.py
 
 This starts one server for each flow that returns a UI. Access the UI at the link printed in the terminal.
+
 """
 
 import os

examples/fabric/build_your_own_trainer/trainer.py

@@ -137,6 +137,7 @@ def fit(
                 If not specified, no validation will run.
             ckpt_path: Path to previous checkpoints to resume training from.
                 If specified, will always look for the latest checkpoint within the given directory.
+
         """
         self.fabric.launch()
 
@@ -207,6 +208,7 @@ def train_loop(
                 If greater then the number of batches in the ``train_loader``, this has no effect.
             scheduler_cfg: The learning rate scheduler configuration.
                 Have a look at :meth:`lightning.pytorch.LightninModule.configure_optimizers` for supported values.
+
         """
         self.fabric.call("on_train_epoch_start")
         iterable = self.progbar_wrapper(
@@ -268,6 +270,7 @@ def val_loop(
             val_loader: The dataloader yielding the validation batches.
             limit_batches: Limits the batches during this validation epoch.
                 If greater then the number of batches in the ``val_loader``, this has no effect.
+
         """
         # no validation if val_loader wasn't passed
         if val_loader is None:
@@ -311,13 +314,14 @@ def val_loop(
         torch.set_grad_enabled(True)
 
     def training_step(self, model: L.LightningModule, batch: Any, batch_idx: int) -> torch.Tensor:
-        """A single training step, running forward and backward. The optimizer step is called separately, as this
-        is given as a closure to the optimizer step.
+        """A single training step, running forward and backward. The optimizer step is called separately, as this is
+        given as a closure to the optimizer step.
 
         Args:
             model: the lightning module to train
             batch: the batch to run the forward on
             batch_idx: index of the current batch w.r.t the current epoch
+
         """
         outputs: Union[torch.Tensor, Mapping[str, Any]] = model.training_step(batch, batch_idx=batch_idx)
 
@@ -347,6 +351,7 @@ def step_scheduler(
                 Have a look at :meth:`lightning.pytorch.LightningModule.configure_optimizers` for supported values.
             level: whether we are trying to step on epoch- or step-level
             current_value: Holds the current_epoch if ``level==epoch``, else holds the ``global_step``
+
         """
 
         # no scheduler
@@ -395,6 +400,7 @@ def progbar_wrapper(self, iterable: Iterable, total: int, **kwargs: Any):
         Args:
             iterable: the iterable to wrap with tqdm
             total: the total length of the iterable, necessary in case the number of batches was limited.
+
         """
         if self.fabric.is_global_zero:
             return tqdm(iterable, total=total, **kwargs)
@@ -406,6 +412,7 @@ def load(self, state: Optional[Mapping], path: str) -> None:
         Args:
             state: a mapping contaning model, optimizer and lr scheduler
             path: the path to load the checkpoint from
+
         """
         if state is None:
             state = {}
@@ -458,6 +465,7 @@ def _parse_optimizers_schedulers(
         Args:
             configure_optim_output: The output of ``configure_optimizers``.
                 For supported values, please refer to :meth:`lightning.pytorch.LightningModule.configure_optimizers`.
+
         """
         _lr_sched_defaults = {"interval": "epoch", "frequency": 1, "monitor": "val_loss"}
 
@@ -511,6 +519,7 @@ def _format_iterable(
             prog_bar: a progressbar (on global rank zero) or an iterable (every other rank).
             candidates: the values to add as postfix strings to the progressbar.
             prefix: the prefix to add to each of these values.
+
         """
         if isinstance(prog_bar, tqdm) and candidates is not None:
             postfix_str = ""

examples/fabric/image_classifier/train_fabric.py

@@ -25,6 +25,7 @@
 Accelerate your training loop by setting the ``--accelerator``, ``--strategy``, ``--devices`` options directly from
 the command line. See ``lightning run model --help`` or learn more from the documentation:
 https://lightning.ai/docs/fabric.
+
 """
 
 import argparse

examples/pytorch/basics/autoencoder.py

@@ -14,6 +14,7 @@
 """MNIST autoencoder example.
 
 To run: python autoencoder.py --trainer.max_epochs=50
+
 """
 from os import path
 from typing import Optional, Tuple

examples/pytorch/basics/backbone_image_classifier.py

@@ -14,6 +14,7 @@
 """MNIST backbone image classifier example.
 
 To run: python backbone_image_classifier.py --trainer.max_epochs=50
+
 """
 from os import path
 from typing import Optional

examples/pytorch/basics/profiler_example.py

@@ -20,6 +20,7 @@
 * With PyTorch Tensorboard Profiler (Instructions are here: https://github.com/pytorch/kineto/tree/master/tb_plugin)
     1. pip install tensorboard torch-tb-profiler
     2. tensorboard --logdir={FOLDER}
+
 """
 
 from os import path

examples/pytorch/domain_templates/computer_vision_fine_tuning.py

@@ -12,9 +12,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Computer vision example on Transfer Learning. This computer vision example illustrates how one could fine-tune a
-pre-trained network (by default, a ResNet50 is used) using pytorch-lightning. For the sake of this example, the
-'cats and dogs dataset' (~60MB, see `DATA_URL` below) and the proposed network (denoted by `TransferLearningModel`,
-see below) is trained for 15 epochs.
+pre-trained network (by default, a ResNet50 is used) using pytorch-lightning. For the sake of this example, the 'cats
+and dogs dataset' (~60MB, see `DATA_URL` below) and the proposed network (denoted by `TransferLearningModel`, see
+below) is trained for 15 epochs.
 
 The training consists of three stages.
 
@@ -37,6 +37,7 @@
 
 To run:
     python computer_vision_fine_tuning.py fit
+
 """
 
 import logging
@@ -97,6 +98,7 @@ def __init__(self, dl_path: Union[str, Path] = "data", num_workers: int = 0, bat
             dl_path: root directory where to download the data
             num_workers: number of CPU workers
             batch_size: number of sample in a batch
+
         """
         super().__init__()
 
@@ -174,6 +176,7 @@ def __init__(
             milestones: List of two epochs milestones
             lr: Initial learning rate
             lr_scheduler_gamma: Factor by which the learning rate is reduced at each milestone
+
         """
         super().__init__()
         self.backbone = backbone
@@ -209,6 +212,7 @@ def forward(self, x):
         """Forward pass.
 
         Returns logits.
+
         """
         # 1. Feature extraction:
         x = self.feature_extractor(x)

examples/pytorch/domain_templates/generative_adversarial_net.py

@@ -16,6 +16,7 @@
 After a few epochs, launch TensorBoard to see the images being generated at every batch:
 
 tensorboard --logdir default
+
 """
 from argparse import ArgumentParser, Namespace
 

examples/pytorch/domain_templates/imagenet.py

@@ -28,6 +28,7 @@
 
     python imagenet.py --help
     python imagenet.py fit --help
+
 """
 import os
 from typing import Optional