Losses

Work in progress

Come back later!

using AbstractTrees, InferOpt, InteractiveUtils
AbstractTrees.children(x::Type) = subtypes(x)
print_tree(InferOpt.AbstractLossLayer)

AbstractLossLayer
├─ FenchelYoungLoss
├─ ImitationLoss
├─ SPOPlusLoss
└─ StructuredSVMLoss