FZfangzheng

icon company BIT icon location Beijing

Results 2 comments of FZfangzheng

Y'and Vy'

I also found this problem. The loss function in the paper has four parts, but only three are given in the code. L1(V’b, V’Y) is not given in the code.

Encountered known unsupported method torch.Tensor.add

Maybe x.add should be changed to torch.add