training problem

[jack155861]

hi

when i trained the model, it stopped in T_Net
it shows the error

and I print, i get the different shape
print(s4_.shape)
#torch.Size([8, 96, 75, 75])
print(s4.shape)
#torch.Size([8, 96, 76, 76])

how to slove it ??

thanks

[rainsun1]

You should crop the image into size (256, 256) by "image, trimap, alpha = random_scale_and_creat_patch(image, trimap, alpha, patch_size)" as in the code.
Perhaps your image size is 800x600, and 600 cannot be devided by 16 scaled in T-Net.