CircleLoss icon indicating copy to clipboard operation
CircleLoss copied to clipboard

Published 20 hours ago • verified publisher icon TinyZeaMays

self.soft_plus(torch.logsumexp(logit_n, dim=0) + torch.logsumexp(logit_p, dim=0))

Open Heath-zyl opened this issue 4 years ago • 2 comments

Isn't it "loss = self.soft_plus(torch.logsumexp(logit_n, dim=0) * torch.logsumexp(logit_p, dim=0))" ?

Heath-zyl avatar Jan 20 '21 03:01 Heath-zyl

Yes, I also confused about it.

zzk2021 avatar Feb 09 '23 08:02 zzk2021

Cartesian product

zzk2021 avatar Feb 09 '23 08:02 zzk2021