Generating edge maps not working

[Derdiry]

python main.py --model pidinet --config carv4 --sa --dil --resume --iter-size 24 -j 4 --gpu 0 --epochs 20 --lr 0.005 --lr-type multistep --lr-steps 10-16 --wd 1e-4 --savedir output --datadir input --dataset BSDS
resulting :
initialization done
conv weights: lr 0.005000, wd 0.000100 bn weights: lr 0.005000, wd 0.000010 relu weights: lr 0.005000, wd 0.000000
cuda is not used, the running might be slow
Threshold for ground truth: 76.800000 on BSDS_VOC
Threshold for ground truth: 76.800000 on BSDS_VOC
Traceback (most recent call last):
File "C:\Users\Al Gadi\Downloads\pidinet-master\pidinet-master\main.py", line 418, in
main(f)
File "C:\Users\Al Gadi\Downloads\pidinet-master\pidinet-master\main.py", line 228, in main
tr_avg_loss = train(
File "C:\Users\Al Gadi\Downloads\pidinet-master\pidinet-master\main.py", line 280, in train
loss += cross_entropy_loss_RCF(o, label, args.lmbda)
File "C:\Users\Al Gadi\Downloads\pidinet-master\pidinet-master\utils.py", line 150, in cross_entropy_loss_RCF
cost = F.binary_cross_entropy(
File "C:\Users\Al Gadi\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\functional.py", line 3122, in binary_cross_entropy
return torch._C._nn.binary_cross_entropy(input, target, weight, reduction_enum)
RuntimeError: all elements of target should be between 0 and 1

[LeiXuAI]

if you degrade your pytorch to version 1.XX, it will work.