search

Deeachain / Segmentation-Pytorch

Semantic Segmentation in Pytorch. Network include: FCN、FCN_ResNet、SegNet、UNet、BiSeNet、BiSeNetV2、PSPNet、DeepLabv3_plus、 HRNet、DDRNet
465 stars 69 forks source link

Error when run train.sh #7

Closed NguyenCongPhucBK closed 3 years ago

NguyenCongPhucBK commented 3 years ago

Hi, I train model in Google Colab and i bug: image Can you help me? Thanks you so much!

NguyenCongPhucBK commented 3 years ago

I installed package in requirements.txt, and I still bug in local server: Capture

NguyenCongPhucBK commented 3 years ago

I edit to fix bug: image

But, i have other bug when run train.sh: image

NguyenCongPhucBK commented 3 years ago

Can you help me, @Deeachain ? Thanks you!

Deeachain commented 3 years ago

Hi， This is not a code bug! Report error solution see https://github.com/Deeachain/Segmentation-Pytorch/issues/7#issuecomment-847500499 You annotated trainLoader and testLoader, It needs to be revised. image

NguyenCongPhucBK commented 3 years ago

Thanks @Deeachain , I train successful!. I close Issuse. I hope DDRNet 23-slim coming soon! Thank you so much!

Deeachain commented 3 years ago

I installed package in requirements.txt, and I still bug in local server: Capture This is not a code bug, you should set the '--nproc_per_node=2' in 'train.sh' to the number of graphics cards when the number of your lab devices changes (the default is dual cards)