torchvision transform巨坑

最近用到torchvision.transform.Normalize，发现不同版本对输入的改变不同，下面对比torchvision 0.2.1和0.4.2版本
0.2.1：

def normalize(tensor, mean, std):
    """Normalize a tensor image with mean and standard deviation.
    See ``Normalize`` for more details.
    Args:
        tensor (Tensor): Tensor image of size (C, H, W) to be normalized.
        mean (sequence): Sequence of means for each channel.
        std (sequence): Sequence of standard deviations for each channely.
    Returns:
        Tensor: Normalized Tensor image.
    """
    if not _is_tensor_image(tensor):
        raise TypeError('tensor is not a torch image.')
    # TODO: make efficient
    for t, m, s in zip(tensor, mean, std):
        t.sub_(m).div_(s)
    return tensor

0.4.2：

def normalize(tensor, mean, std, inplace=False):
    """Normalize a tensor image with mean and standard deviation.
    .. note::
        This transform acts out of place by default, i.e., it does not mutates the input tensor.
    See :class:`~torchvision.transforms.Normalize` for more details.
    Args:
        tensor (Tensor): Tensor image of size (C, H, W) to be normalized.
        mean (sequence): Sequence of means for each channel.
        std (sequence): Sequence of standard deviations for each channel.
        inplace(bool,optional): Bool to make this operation inplace.
    Returns:
        Tensor: Normalized Tensor image.
    """
    if not _is_tensor_image(tensor):
        raise TypeError('tensor is not a torch image.')

    if not inplace:
        tensor = tensor.clone()

    dtype = tensor.dtype
    mean = torch.as_tensor(mean, dtype=dtype, device=tensor.device)
    std = torch.as_tensor(std, dtype=dtype, device=tensor.device)
    tensor.sub_(mean[:, None, None]).div_(std[:, None, None])
    return tensor

可以看到0.2.1的normalize操作实际上是inplace操作，即会对输入输出同步改变，而0.4.2则提供了参数inplace，是否改变输入可自主选择。
在此提醒使用了0.2.1版本的normalize操作的各位，要注意输入的改变，如果不想改变则可先把输入的clone tensor传入到normalize中。