PredictCallback

class pytorch_forecasting.models.base_model.PredictCallback(mode: str | Tuple[str, str] = 'prediction', return_index: bool = False, return_decoder_lengths: bool = False, return_y: bool = False, write_interval: Literal['batch', 'epoch', 'batch_and_epoch'] = 'batch', return_x: bool = False, mode_kwargs: Dict[str, Any] | None = None, output_dir: str | None = None, predict_kwargs: Dict[str, Any] | None = None)

Bases: BasePredictionWriter

Internally used callback to capture predictions and optionally write them to disk.

Inherited-members:

Methods

load_state_dict(state_dict)	Called when loading a checkpoint, implement to reload callback state given callback's state_dict.
on_after_backward(trainer, pl_module)	Called after loss.backward() and before optimizers are stepped.
on_before_backward(trainer, pl_module, loss)	Called before loss.backward().
on_before_optimizer_step(trainer, pl_module, ...)	Called before optimizer.step().
on_before_zero_grad(trainer, pl_module, ...)	Called before optimizer.zero_grad().
on_exception(trainer, pl_module, exception)	Called when any trainer execution is interrupted by an exception.
on_fit_end(trainer, pl_module)	Called when fit ends.
on_fit_start(trainer, pl_module)	Called when fit begins.
on_load_checkpoint(trainer, pl_module, ...)	Called when loading a model checkpoint, use to reload state.
on_predict_batch_end(trainer, pl_module, ...)	Called when the predict batch ends.
on_predict_batch_start(trainer, pl_module, ...)	Called when the predict batch begins.
on_predict_end(trainer, pl_module)	Called when predict ends.
on_predict_epoch_end(trainer, pl_module)	Called when the predict epoch ends.
on_predict_epoch_start(trainer, pl_module)	Called when the predict epoch begins.
on_predict_start(trainer, pl_module)	Called when the predict begins.
on_sanity_check_end(trainer, pl_module)	Called when the validation sanity check ends.
on_sanity_check_start(trainer, pl_module)	Called when the validation sanity check starts.
on_save_checkpoint(trainer, pl_module, ...)	Called when saving a checkpoint to give you a chance to store anything else you might want to save.
on_test_batch_end(trainer, pl_module, ...[, ...])	Called when the test batch ends.
on_test_batch_start(trainer, pl_module, ...)	Called when the test batch begins.
on_test_end(trainer, pl_module)	Called when the test ends.
on_test_epoch_end(trainer, pl_module)	Called when the test epoch ends.
on_test_epoch_start(trainer, pl_module)	Called when the test epoch begins.
on_test_start(trainer, pl_module)	Called when the test begins.
on_train_batch_end(trainer, pl_module, ...)	Called when the train batch ends.
on_train_batch_start(trainer, pl_module, ...)	Called when the train batch begins.
on_train_end(trainer, pl_module)	Called when the train ends.
on_train_epoch_end(trainer, pl_module)	Called when the train epoch ends.
on_train_epoch_start(trainer, pl_module)	Called when the train epoch begins.
on_train_start(trainer, pl_module)	Called when the train begins.
on_validation_batch_end(trainer, pl_module, ...)	Called when the validation batch ends.
on_validation_batch_start(trainer, ...[, ...])	Called when the validation batch begins.
on_validation_end(trainer, pl_module)	Called when the validation loop ends.
on_validation_epoch_end(trainer, pl_module)	Called when the val epoch ends.
on_validation_epoch_start(trainer, pl_module)	Called when the val epoch begins.
on_validation_start(trainer, pl_module)	Called when the validation loop begins.
setup(trainer, pl_module, stage)	Called when fit, validate, test, predict, or tune begins.
state_dict()	Called when saving a checkpoint, implement to generate callback's state_dict.
teardown(trainer, pl_module, stage)	Called when fit, validate, test, predict, or tune ends.
write_on_batch_end(trainer, pl_module, ...)	Override with the logic to write a single batch.
write_on_epoch_end(trainer, pl_module, ...)	Override with the logic to write all batches.

Attributes

result
state_key	Identifier for the state of the callback.

on_predict_batch_end(trainer: Trainer, pl_module: LightningModule, outputs: Any, batch: Any, batch_idx: int, dataloader_idx: int = 0) → None

Called when the predict batch ends.

on_predict_epoch_end(trainer: Trainer, pl_module: LightningModule) → None

Called when the predict epoch ends.

write_on_batch_end(trainer, pl_module, prediction, batch_indices, batch, batch_idx, dataloader_idx)

Override with the logic to write a single batch.

write_on_epoch_end(trainer, pl_module, predictions, batch_indices)

Override with the logic to write all batches.