DoublePrecisionPlugin¶

class lightning.pytorch.plugins.precision.DoublePrecisionPlugin[소스]¶

기반 클래스: lightning.pytorch.plugins.precision.precision_plugin.PrecisionPlugin

Plugin for training with double (torch.float64) precision.

connect(model, optimizers, lr_schedulers)[소스]¶

Converts the model to double precision and wraps it in a LightningDoublePrecisionModule to convert incoming floating point data to double (torch.float64) precision.

Does not alter optimizers or lr_schedulers.

반환 형식: Tuple[Module, List[Optimizer], List[Any]]

forward_context()[소스]¶

A context manager to change the default tensor type.

See: torch.set_default_tensor_type()

반환 형식: Generator[None, None, None]