自己定义 l2函数和l1函数的具体数值 pytorch 查看权重参数

l1正则化的方法

regularization_loss = 0
        regularization_loss1 = 0
        for param in net.parameters():
            regularization_loss += torch.sum(abs(param))
        
    
        #ww=net.parameters#net.weight
        loss =loss
        +0.25*regularization_loss1+0.5*regularization_loss

l2正则化的方法

regularization_loss = 0
        regularization_loss1 = 0
        for param in net.parameters():
          
            regularization_loss += torch.sum(abs(param*param))
    
        #ww=net.parameters#net.weight
        loss =loss
        +0.25*regularization_loss1+0.5*regularization_loss