Closed Bonsen closed 6 years ago
In Cross-entropy training, loss = L.xloss(out, labels, ignore=255)?
So how to set the classes weights?
See #13
In Cross-entropy training, loss = L.xloss(out, labels, ignore=255)?