| from torch.nn.modules import Module | |
| from typing import Any, TypeVar | |
| T_co = TypeVar('T_co', covariant=True) | |
| class DistributedDataParallelCPU(Module[T_co]): | |
| module: Module[T_co] = ... | |
| needs_reduction: bool = ... | |
| def __init__(self, module: Module[T_co]) -> None: ... | |
| def sync_parameters(self) -> None: ... | |
| # TODO doublecheck that this does return a T_co and not a list of T_cos | |
| def forward(self, *inputs: Any, **kwargs: Any) -> T_co: ... |