Yes.
- logits →
nn.BCEWithLogitsLoss
- logits → sigmoid →
nn.BCELoss
That would be the standard definition, but you can basically define True/False, Positive/Negative as you wish.
Yes.
nn.BCEWithLogitsLoss
nn.BCELoss
That would be the standard definition, but you can basically define True/False, Positive/Negative as you wish.