MNIST，Torchvision中的输出和广播形状不匹配

Question

在Torchvision中使用MNIST数据集时出现以下错误

RuntimeError: output with shape [1, 28, 28] doesn't match the broadcast shape [3, 28, 28]

这是我的代码：

import torch
from torchvision import datasets, transforms

transform = transforms.Compose([transforms.ToTensor(),
                            transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
                          ])
trainset = datasets.MNIST('~/.pytorch/MNIST_data/', download=True, train=True, transform=transform)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)
images, labels = next(iter(trainloader))

Answer 1

该错误是由于数据集上的颜色与灰度的关系，数据集是灰度的。

我通过将transform更改为

transform = transforms.Compose([transforms.ToTensor(),
  transforms.Normalize((0.5,), (0.5,))
])