utils package

Submodules

utils.utils module

check_config(config: DictConfig) → None[source]
A couple of optional utilities, controlled by main config file.

  • check for required configs in the main config

  • disabling warnings

  • easier access to debug mode

  • forcing debug friendly configuration

  • forcing multi-gpu friendly configuration

  • setting seed for random number generators

  • setting up default csv logger

Parameters:

config (DictConfig) – the main hydra config

empty(*args, **kwargs)[source]

This function does nothing. It is used to disable logging of hyperparameters by Lightning loggers.

Parameters:

  • args –

  • kwargs –

finish(config: DictConfig, task: LightningModule, model: LightningModule, datamodule: LightningDataModule, trainer: Trainer, callbacks: List[Callback], logger: List[LightningLoggerBase]) → None[source]

Makes sure everything closed properly.

Parameters:

  • config (DictConfig) – Hydra config

  • task (pl.LightningModule) – Lightning task

  • model (pl.LightningModule) – Lightning model

  • datamodule (pl.LightningDataModule) – Lightning datamodule

  • trainer (pl.Trainer) – Lightning trainer

  • callbacks (List[pl.Callback]) – Lightning callbacks

  • logger (List[pl.loggers.LightningLoggerBase]) – Lightning logger

get_logger(name='utils.utils', level=20) → Logger[source]

Gets the Python logger of the system.

Parameters:

  • name (str) – name of the logger you want to get defaults to __name__

  • level – logging level defaults to logging.INFO

Returns:

Python logger

Return type:

logging.Logger

log_hyperparameters(config: DictConfig, model: LightningModule, trainer: Trainer) → None[source]

This method controls which parameters from Hydra config are saved by Lightning loggers.

Additionally, saves:

  • sizes of train, val, test dataset

  • number of trainable model parameters

Parameters:

  • config (DictConfig) – Hydra config

  • model (pl.LightningModule) – Lightning model

  • trainer (pl.Trainer) – Lightning trainer

print_config(config: DictConfig, fields: Sequence[str] = ('trainer', 'task', 'model', 'optimizer', 'datamodule', 'callbacks', 'loss', 'metric', 'logger', 'seed', 'train', 'test', 'predict'), resolve: bool = True, add_missing_fields: bool = True) → None[source]

Prints content of DictConfig using Rich library and its tree structure.

Parameters:

  • config (DictConfig) – Hydra config

  • fields (Sequence[str]) – Determines which main fields from config will be printed and in what order.

  • resolve (bool) – Whether to resolve reference fields of DictConfig.

  • add_missing_fields (bool) – Whether to add missing fields from config to fields.

Module contents