杂碎笔记

torchvision.transforms

normalize = transforms.Normalize(
    mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]
)

该transform除了做normalize之外，会变换输入图片的输出，输入假设是（256，256，3），输出则是（3，256，256）。除此之外，训练做了normalize测试一定要做。

UnicodeDecodeError: ‘ascii’ codec can’t decode byte 0xc0 in position 0: ordinal not in range(128)

python2训练的模型在python3上测试带来的错误解决办法是

    from functools import partial
    import pickle
    pickle.load = partial(pickle.load, encoding="latin1")
    pickle.Unpickler = partial(pickle.Unpickler, encoding="latin1")
    checkpoint = torch.load(model_best.pth.tar', map_location=lambda storage, loc: storage, pickle_module=pickle)