pytorch, for the cross_entropy function, What if the input labels are all ignore

Home > Net > pytorch, for the cross_entropy function, What if the input labels are all ignore_index?

pytorch, for the cross_entropy function, What if the input labels are all ignore_index?

Time：08-08

for the nn.CrossEntropyLoss(), What if the input labels are all ignore_index ? I got a 'nan' loss, how to fix it ? thanks!

detail of my code is following: