Hang Zhang

The VGG model I used was converted from Caffe, which was trained without std normalization.

Sorry guys. This was an early project when PyTorch was just released. The code is not runnable any more. Please try other forks or try other versions: https://github.com/zhanghang1989/MSG-Net https://github.com/zhanghang1989/MXNet-Gluon-Style-Transfer

谢谢支持！这个功能不是必要的，不过大家可以借鉴你的代码

This looks like a bug with PyTorch. I will convert the vgg weights later to avoid this issue.

Great. please keep me posted.

Not sure if you've already solved the issue. Detectron2 ResNet is caffe style, which is slightly different architecture from the TorchVision version.

This is a really old and unmaintained repo. Please use more recent projects, e.g. [mmsegmentation](https://github.com/open-mmlab/mmsegmentation)

Is the docker in torch-encoding working? https://hangzhang.org/PyTorch-Encoding/notes/compile.html#using-docker

For issue related to lang-seg, it would be great to reach out to the original authors

我没有你们遇到的问题，你们用的最新的版本吗？ ```python In [1]: from resnest.torch import resnest50 ...: net = resnest50(pretrained=True) Downloading: "https://github.com/zhanghang1989/ResNeSt/releases/download/weights_step1/resnest50-528c19ca.pth" to /Users/hang.zhang/.cache/torch/hub/checkpoints/resnest50-528c19ca.pth 100%|█████████████████████████████████████████████████████████████████████████████████████| 105M/105M [00:06