EvaluationLoop¶

class pytorch_lightning.loops.dataloader.EvaluationLoop(verbose=True)[source]¶

Loops over all dataloaders for evaluation.

advance(*args, **kwargs)[source]¶

Performs evaluation on one single dataloader.

connect(epoch_loop)[source]¶

Connect the evaluation epoch loop with this loop.

on_advance_end()[source]¶

Hook to be called each time after advance is called.

on_advance_start(*args, **kwargs)[source]¶

Hook to be called each time before advance is called.

Accepts all arguments passed to :attr`run`.

on_run_end()[source]¶

Runs the _on_evaluation_epoch_end hook.

Return type: List[Dict[str, Union[Tensor, Dict[str, Tensor]]]]

on_run_start(*args, **kwargs)[source]¶

Runs the _on_evaluation_model_eval, _on_evaluation_start and _on_evaluation_epoch_start hooks.

The function to run when run() should be skipped, determined by the condition in skip.

Resets the internal state of the loop.

Use to release memory etc.

property dataloaders: Sequence[torch.utils.data.dataloader.DataLoader]¶

Returns the validation or test dataloaders.

property done: bool¶

Returns whether all dataloaders are processed or evaluation should be skipped altogether.

property num_dataloaders: int¶

Returns the total number of dataloaders.

property skip: bool¶

Returns whether the evaluation should be skipped.