: Given transposed=1, weight of size 1024 2048 4 4, expected input[8, 2048, 10, 10] to have 1024 channels, but got 2048 channels instead

IanTaehoonYoo / semantic-segmentation-pytorch

Pytorch implementation of FCN, UNet, PSPNet, and various encoder models.

MIT License

86 stars 19 forks source link

: Given transposed=1, weight of size 1024 2048 4 4, expected input[8, 2048, 10, 10] to have 1024 channels, but got 2048 channels instead #3

Closed jiafengshen closed 4 years ago

IanTaehoonYoo commented 4 years ago

Hi, @jiafengshen Could you give me some information you used the model and input data?

Thanks Best regards,

IanTaehoonYoo commented 4 years ago

I found that problem when testing the model 'unet_resnet34' and I fixed it. I update the project, so you should update it. use the command: pip install --upgrade seg-torch

Thanks