a = torch.tensor(2.0, requires_grad=True) # we set requires_grad=True to let ... We need to zero the grad variable since the backward() # call accumulates ...
確定! 回上一頁