我相信存储在trainloader.dataset.data或.target中的数据不会被随机打乱,只有当DataLoader被作为生成器或迭代器调用时,数据才会被随机打乱。
你可以通过多次执行next(iter(trainloader))来进行检验,一次没有随机打乱,一次有随机打乱,这两种情况应该会给出不同的结果。
import torch
import torchvision
transform = torchvision.transforms.Compose([
torchvision.transforms.ToTensor(),
])
MNIST_dataset = torchvision.datasets.MNIST('~/Desktop/intern/',download = True, train = False,
transform = transform)
dataLoader = torch.utils.data.DataLoader(MNIST_dataset,
batch_size = 128,
shuffle = False,
num_workers = 10)
target = dataLoader.dataset.targets
MNIST_dataset = torchvision.datasets.MNIST('~/Desktop/intern/',download = True, train = False,
transform = transform)
dataLoader_shuffled= torch.utils.data.DataLoader(MNIST_dataset,
batch_size = 128,
shuffle = True,
num_workers = 10)
target_shuffled = dataLoader_shuffled.dataset.targets
print(target == target_shuffled)
_, target = next(iter(dataLoader));
_, target_shuffled = next(iter(dataLoader_shuffled))
print(target == target_shuffled)
这将会给出:
tensor([True, True, True, ..., True, True, True])
tensor([False, False, False, False, False, False, False, False, False, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, False, False, False, False, False, True,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, True, False, False, False, False, False,
False, True, False, False, False, False, False, False, False, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, True, True, False, False, False, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, False, False, True, False, False, True, False,
False, False, False, False, False, False, False, False])
然而,存储在数据和目标中的数据和标签是固定的列表,由于您试图直接访问它们,它们不会被洗牌。