"Killed" at the epock 5 when setting 123 at line 135.

AngeLouCN / DC-UNet

We proposed a novel U-Net-based model -- DC-UNet to do medical image segmentation.

325 stars 20 forks source link

"Killed" at the epock 5 when setting 123 at line 135. #2

Closed tphankr closed 3 years ago

tphankr commented 3 years ago

Allow me ask, When we train with the CVC_clinicDB dataset and I set 123 at line 156 in the below picture ( for i in range (123): ; the original code is "for i in range (10):") and then python main.py, the code allways is killed at epoch 5. . Why I setting 123 because I want to plot 123 picture in the validation dataset ( 20% of 612) If we set 10, the code is good and plot 10 pictures.

Thank you, very much.

AngeLouCN commented 3 years ago

Hi, we set the figure size at line 158. If you put 123 at line 156, it can not show 123 images in this small size. So, I suggest you keep 10 at line 156. You can predict all results (123 images) after finishing training.