Cross entropy in SMP. Loss of semantic segmentation models

hi all
How can I use the cross-entropy loss in SMP.losses?