nn.KLDivLoss,nn.CrossEntropyLoss,nn.MSELoss,Focal_Loss
KLloss:https://blog.csdn.net/qq_50001789/article/details/128974654https://pytorch.org/docs/stable/nn.html1.nn.L1Loss1.1公式L1Loss:计算预测x和目标y之间的平均绝对值误差MAE,即L1损失:loss=1n∑i=1,...n∣xi−yi∣loss=\frac{1}{n}\sum