基于MNIST数据集的PyTorch GAN项目报错：'module' object is not callable

在使用PyTorch构建生成对抗网络（GAN）并处理MNIST数据集时，遍历DataLoader过程中触发如下错误：

TypeError: 'module' object is not callable

相关代码片段 #

导入模块：

import torch
import torch.nn as nn
import torch.optim as optim
import torchvision
import torchvision.datasets as datasets
from torch.utils.data import DataLoader, Dataset
import torchvision.transforms as transforms
from torch.utils.tensorboard import SummaryWriter

数据集与DataLoader定义：

dataset = datasets.MNIST(root="dataset/",transform = transforms,download = True)
loader = DataLoader(dataset, batch_size = batch_size,shuffle = True)

遍历代码：

for epoch in range(num_epochs):
    for batch_idx, (real, _) in enumerate(loader):
        real = real.view(-1, 784).to(device)

完整报错栈 #

Traceback (most recent call last):
  File "C:\Users\utkar\PycharmProjects\simpleGAN\main.py", line 57, in <module>
    for batch_idx, (real, _) in enumerate(loader):
  File "C:\Users\utkar\Anaconda3\envs\deeplearning\lib\site-packages\torch\utils\data\dataloader.py", line 628, in __next__
    data = self._next_data()
  File "C:\Users\utkar\Anaconda3\envs\deeplearning\lib\site-packages\torch\utils\data\dataloader.py", line 671, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "C:\Users\utkar\Anaconda3\envs\deeplearning\lib\site-packages\torch\utils\data\_utils\fetch.py", line 58, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "C:\Users\utkar\Anaconda3\envs\deeplearning\lib\site-packages\torch\utils\data\_utils\fetch.py", line 58, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "C:\Users\utkar\Anaconda3\envs\deeplearning\lib\site-packages\torchvision\datasets\mnist.py", line 145, in __getitem__
    img = self.transform(img)
TypeError: 'module' object is not callable

错误原因 #

• 核心问题：创建MNIST数据集时，transform参数传入的是整个transforms模块，而非可调用的变换实例。数据集在获取样本时会执行self.transform(img)，尝试调用传入的对象，但模块本身无法被调用，因此触发错误。

修复方案 #

需要将具体的变换操作通过transforms.Compose()组合成可调用的变换流水线，最基础的是将图像转换为PyTorch Tensor。修改后的代码如下：

# 定义变换流水线：将图像转为Tensor
transform = transforms.Compose([
    transforms.ToTensor()
])
dataset = datasets.MNIST(root="dataset/", transform=transform, download=True)
loader = DataLoader(dataset, batch_size=batch_size, shuffle=True)

如果需要添加归一化等更多预处理操作，可直接在Compose中扩展：

transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize((0.5,), (0.5,))  # 对单通道MNIST图像做归一化
])

内容的提问来源于stack exchange，提问作者Cosmo