Is it possible to use the nn.MultiMarginLoss() for classfication where i have 5 classes?
Im a bit confused about the difference between multi label and multi classes… Multi label is if an x have multiple y values? My x values only have one y value.
Any clarification is much appreciated.
Regards