Pytorch实现gan

最新推荐文章于 2024-08-20 22:08:31 发布

柚子柚子蜜

最新推荐文章于 2024-08-20 22:08:31 发布

阅读量587

点赞数 1

文章标签： MNIST数据集下载失败输入输出匹配资源路径 loss函数

本文链接：https://blog.csdn.net/weixin_44129933/article/details/116145274

版权

问题一：数据集下载失败

解决方案：

通过百度网盘手动下载文件，网盘链接密码：locq。将文件拷贝到\raw下（无需解压）
在这里插入图片描述
mnist数据集下载的代码如下：

mnist = datasets.MNIST(
    root='./data/', 
    train=True, 
    transform=img_transform, 
    download=True
)

修改mnist.py
（改之前）

resources = [
        ("http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz", "f68b3c2dcbeaaa9fbdd348bbdeb94873"),
        ("http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz", "d53e105ee54ea40749a09fcbcd1e9432"),
        ("http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz", "9fb629c4189551a2d022fa330f9573f3"),
        ("http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz", "ec29112dd5afa0611ce80d1b7f02629c")
     ]

改之后：换成如下自己的mnist下载存放的路径

 resources = [
        ("file:///E:\\pycharm-workspace\\untitled\\data\\MNIST\\raw\\train-images-idx3-ubyte.gz",None),
        ("file:///E:\\pycharm-workspace\\untitled\\data\\MNIST\\raw\\train-labels-idx1-ubyte.gz",None),
        ("file:///E:\\pycharm-workspace\\untitled\\data\\MNIST\\raw\\t10k-images-idx3-ubyte.gz",None),
        ("file:///E:\\pycharm-workspace\\untitled\\data\\MNIST\\raw\\t10k-labels-idx1-ubyte.gz",None)
    ]

问题二：输入输出不匹配

ValueError: Using a target size (torch.Size([128])) that is different to the input size (torch.Size([128, 1])) is deprecated. Please ensure they have the same size.

解决

 real_label = Variable(torch.ones(num_img)).cuda()  # 定义真实的图片label为1  
 fake_label = Variable(torch.zeros(num_img)).cuda()  # 定义假的图片的label为0  
 real_out = D(real_img)  # 将真实图片放入判别器中
 # print(real_out.size()) 128*1
 # print(real_label.size())  128
 定位到损失函数：即real_out([128,1])和real_label([128])不匹配
 d_loss_real = criterion(real_out, real_label)  # 得到真实图片的loss