1. def __rotate(img):
    2.     probability = random.random()
    3.     if (probability < 0.5):
    4.         angles = [90, 180, 270]
    5.         angle = random.choice(angles)
    6.         return TF.rotate(img, angle)
    7.     else:
    8.         return img
    11. train_tf = transforms.Compose([
    12.         transforms.Resize(256),
    13.         transforms.Lambda(lambda img: __rotate(img)),
    14.         transforms.CenterCrop(224),
    15.         transforms.ToTensor(),
    16.         transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    17.     ])
    19.  train_dataloader = DataLoader(train_set, opt.batch_size, sampler=sampler, num_workers=4)

    如上代码在windows环境下报如题的错误,linux下没有问题,查询文档得知是多线程相关在windows不支持,初始化参数改为num_workers=0就不会报错了,但是加载数据速度会变慢