lightning_template.utils.cli

lightning_template.utils.cli#

Submodules#

Classes#

LightningCLI

Implementation of a configurable command line tool for pytorch-lightning.

Trainer

Functions#

recursive_instantate_class(config)

Package Contents#

class lightning_template.utils.cli.LightningCLI(save_config_callback: Type[lightning.pytorch.cli.SaveConfigCallback] | None = SaveAndLogConfigCallback, trainer_class: Type[lightning_template.utils.cli.trainer._Trainer] | Callable[Ellipsis, lightning_template.utils.cli.trainer._Trainer] = Trainer, *args, **kwargs)#

Bases: lightning.pytorch.cli.LightningCLI

Implementation of a configurable command line tool for pytorch-lightning.

_setup_parser_kwargs(*args, **kwargs) → Tuple[Dict[str, Any], Dict[str, Any]]#
add_default_arguments_to_parser(parser: lightning.pytorch.cli.LightningArgumentParser) → None#

Adds default arguments to the parser.

static randomly_select_seed() → int#
_set_seed() → None#

Sets the seed.

before_instantiate_classes() → None#

Implement to run some code before instantiating the classes.

_add_configure_optimizers_method_to_model(*args, **kwargs) → None#

Overrides the model’s configure_optimizers() method if a single optimizer and optionally a scheduler argument groups are added to the parser as ‘AUTOMATIC’.

lightning_template.utils.cli.recursive_instantate_class(config)#
class lightning_template.utils.cli.Trainer(num_folds: int | None = None, *args, **kwargs)#

Bases: lightning.pytorch.Trainer

num_folds = None#
Contents