torch.autograd.function.FunctionCtx.mark_dirty

FunctionCtx.mark_dirty(*args)

标记给定的张量已在原地操作中被修改。

此方法最多应调用一次，可在 setup_context() 或 forward() 方法中调用，且所有参数都应是输入。

在调用 forward() 时，在原地修改过的每个张量都应该传递给此函数，以确保检查的正确性。无论在修改之前或之后调用此函数，都没有关系。

示例：

>>> class Inplace(Function):
>>>     @staticmethod
>>>     def forward(ctx, x):
>>>         x_npy = x.numpy() # x_npy shares storage with x
>>>         x_npy += 1
>>>         ctx.mark_dirty(x)
>>>         return x
>>>
>>>     @staticmethod
>>>     @once_differentiable
>>>     def backward(ctx, grad_output):
>>>         return grad_output
>>>
>>> a = torch.tensor(1., requires_grad=True, dtype=torch.double).clone()
>>> b = a * a
>>> Inplace.apply(a)  # This would lead to wrong gradients!
>>>                   # but the engine would not know unless we mark_dirty
>>> b.backward() # RuntimeError: one of the variables needed for gradient
>>>              # computation has been modified by an inplace operation