How to avoid log zero in EntropicConfusion ?

[QtacierP]

Thanks for your great paper and related codes. However, I still met some problems in my own implementation with Keras. How to initialize weights better? I mean that if I use my default setting from your PyTorch codes, the feature map in EntropicConfusion hasn't been normalized, and the Log will get a NAN problem. I think I should add an extra normalization?