PyTorch error on camvid predictions (size mismatch)

Reagan1311 / DABNet

Depth-wise Asymmetric Bottleneck for Real-time Semantic Segmentation (BMVC2019)

https://github.com/Reagan1311/DABNet

MIT License

140 stars 36 forks source link

PyTorch error on camvid predictions (size mismatch) #2

Closed niskov closed 5 years ago

niskov commented 5 years ago

When running the following command:

python predict.py --checkpoint DABNet_camvid.pth

I received the following error:

RuntimeError: Error(s) in loading state_dict for DABNet:

    size mismatch for classifier.0.conv.weight: copying a param with shape torch.Size([11, 259, 1, 1]) from checkpoint, the shape in current model is torch.Size([19, 259, 1, 1]).

Reagan1311 commented 5 years ago

predict.py is mainly for cityscapes test set (output the predicted image and submit to the official webpage for evaluation), if you want to test the result on camvid. You should run with following command: python test.py --dataset camvid --checkpoint DABNet_camvid.pth