Can someone guide me to examples where ranking losses such as “torch.nn.MarginRankingLoss” have been used to perform complete training of a model? I see small examples as on the official page Official MarginRankingLoss page, but I am still wondering how it can be used for complete model training.
If possible I would also like to know if there are other losses specifically made towards ranking and implemented in pytorch.