Hello, why is it taking a long time when I run the eval.py file, my GPU doesn't seem to be working, but my CPU is fully loaded。what shoud I do?thanks!

Hello,

One question, is there a way to continue training from a saved epoch / checkpoint?

Because I am training the model using the Google Colab GPUs and every so often the execution is interrupted, and I have to start the training again from 0.

Thank you in advance

Hello!

I am trying to reproduce the results of the paper.

When I run eval.py I get the following error: FileNotFoundError: [Errno 2] No such file or directory: 'tmp/LSNet_thin_diffFPN/checkpoint_epoch_100.pt'

I already created the ./tmp directory and the ./tmp/LSNet_thin_diffFPN directory that are not created in the repository, but what to do with "checkpoint_epoch_100.pt"?

I created a virtual environment with all the versions of the libraries corresponding to the date of October 21, 2022, which is the date of the code. That is, I am using python 3.7, pytorch 1.10.0, scikit-image 0.19.0, scikit-learn 1.0.2, torchvision 0.11.0, torchaudio 0.10.0, tensorboardx 2.4.1, opencv-python 4.5.4.58

Thank you in advance!

the loss named bce_loss，but infact you used FOCAL_loss when calculating BCE_loss.Is this the result of negligence？

I deleted the function of profile() and its results in the project. Will this operation change the results?Thank you.

Instead of outputting the binary image of the test set, it is directly compared with the standard result to obtain the accuracy. Can you tell me how to output the binary image of the test set？ Thank you.

When I try "python train.py", it appeared "torch.nn.modules.module.ModuleAttributeError: 'AdaptiveAvgPool2d' object has no attribute 'total_ops'". Is this a question of my torch version？

Our version：
Python 3.8.13

torch.version
'1.7.1+cu101'
torchvision.version
'0.8.2+cu101'