calculate t was wrong

[Aruen24]

with torch.no_grad():
self.t = target_logit.mean() * 0.01 + (1 - 0.01) * self.t
but your paper is: t(k) = αr(k) + (1 - α)t(k-1);
where t0 = 0, α is the momentum parameter and set to0.99.

[QuasarLight]

i got this problem，did you solve that？

[powermano]

I think the code is right

self.t = target_logit.mean() * 0.01 + (1 - 0.01) * self.t