oneflow.nn.Module.register_full_backward_hook

Module.register_full_backward_hook(hook: Callable[[oneflow.nn.modules.module.Module, Union[Tuple[oneflow.Tensor, …], oneflow.Tensor], Union[Tuple[oneflow.Tensor, …], oneflow.Tensor]], Union[None, oneflow.Tensor]])

Registers a backward hook on the module.

The hook will be called every time the gradients with respect to module inputs are computed. The hook should have the following signature:

hook(module, grad_input, grad_output) -> TensorTuple or None

The grad_input and grad_output are oneflow.TensorTuple that contain the gradients with respect to the inputs and outputs respectively. The hook should not modify its arguments, but it can optionally return a new gradient with respect to the input that will be used in place of grad_input in subsequent computations. grad_input will only correspond to the inputs given as positional arguments and all kwarg arguments are ignored. Entries in grad_input and grad_output will be None for all non-Tensor arguments.

For technical reasons, when this hook is applied to a Module, its forward function will receive a view of each Tensor passed to the Module. Similarly the caller will receive a view of each Tensor returned by the Module’s forward function.

Warning

Modifying inputs or outputs inplace is not allowed when using backward hooks and will raise an error.

Returns

a handle that can be used to remove the added hook by calling handle.remove()

Return type

oneflow.utils.hooks.RemovableHandle