yaohuimin

Results 3 comments of yaohuimin

Hyperparameters for ImageNet training?

Hello @kiimmm, Would you please release your MoCo version code? I'm trying to reimplement it but got stuck in the MoCo code. Thanks.

MoCo code

@HobbitLong Hello there, I'm curious about this implementation of MoCo version, because the accuracy is also high. Why didn't you choose the MoCo version as the hardware requirements is pretty...

fix a bug, which has the little probability of producing nan in the loss

> > This is the simplest way to prevent calculating log(0). And this is nesseceray when the embedding dim is large. When feature representations have very high dimensions, the maximum...