Your mask is wrong. Try scores = scores.masked_fill(scores == 0, -np.inf). scores now looks like tensor([[1.4796, 1.2361, 1.2137, 0.9487, ... ... <看更多>

Complex Autograd for torch.masked_fill is currently supported but untested. ... anjali411 commented on Mar 8 •. edited by pytorch-probot bot ... ... <看更多>

aten::masked_fill. Function Schema: aten::masked_fill.Scalar(Tensor self, Tensor mask, Scalar value) -> (Tensor). Original PyTorch API:. ... <看更多>

pytorch 常用函数- daiwk-github博客- 作者:daiwk. ... 标签：pytorch ... masked_fill_(mask, value) : 在mask值为1的位置处用value填充。mask的元素个数需和本tensor ... ... <看更多>

... <看更多>