GAN模型——pytorch实现

 论文传送门：Generative Adversarial Nets

class Discriminator(nn.Module):  # 定义判别器
    def __init__(self, img_size=(28, 28)):  # 初始化方法
        super(Discriminator, self).__init__()  # 继承初始化方法
        self.img_size = img_size  # 图片尺寸，默认单通道灰度图

        self.linear1 = nn.Linear(self.img_size[0] * self.img_size[1], 512)  # linear映射
        self.linear2 = nn.Linear(512, 256)  # linear映射
        self.linear3 = nn.Linear(256, 1)  # linear映射
        self.leakyrelu = nn.LeakyReLU(0.2, inplace=True)  # leakyrelu激活函数
        self.sigmoid = nn.Sigmoid()  # sigmoid激活函数，将输出压缩至（0，1）

    def forward(self, x):  # 前传函数
        x = torch.flatten(x, 1)  # 输入图片从三维压缩至一维特征向量，(n,1,28,28)-->(n,784)
        x = self.linear1(x)  # linear映射，(n,784)-->(n,512)