neuralop.training.callbacks.CheckpointCallback

class neuralop.training.callbacks.CheckpointCallback(save_dir: Path | str, save_best: str | None = None, save_interval: int = 1, save_optimizer: bool = False, save_scheduler: bool = False, save_regularizer: bool = False, resume_from_dir: Path | str | None = None)[source]

Methods

on_epoch_end(*args, **kwargs)	Save state to dir if all conditions are met
on_val_epoch_end(*args, **kwargs)	Update state dict with errors

on_epoch_start
on_init_end
on_train_start
on_val_epoch_start

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

Update state dict with errors

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

Save state to dir if all conditions are met