A problem in net.py

[xiefan-guo]

Thanks for your awesome reproduce work!

But, I think you have a inadvertent error in net.py.

output_pre = (output - output_bias) / mask_sum + output_bias

I think it should be as follows:

output_pre = (output - output_bias) / mask_sum * (self.kernel_size * self.kernel_size * self.in_channels) + output_bias