Closed damvantai closed 6 years ago
I have not test the code on mnist. I think you can try the following method to make arcface
work probably on mnist
.
batch norm
layer, but the mnist
dataset only contain 10 category and the variance in each category is not large enough to fit so many batch norm
layers, so it will cause the nan
. You should to reduce the number of batch norm
layers. i am using small neural network, but loss error? nan https://github.com/damvantai/tensorflow-sphereface-asoftmax/blob/master/test_mnist.py
i use sphereface loss is right but arcface loss ? error
i meet same problem on mnist(resnet 50, arcface loss), have you solved this problem? @damvantai
when i use arcface loss in mnist, my error loss nan ? do you test your code accuracy ?