Hi, thanks for awesome repo!
I am having trouble with this traceback and was wondering if you could help me get over it
For some context, I tried running all your configs on the same data , only with smaller batch size (GPU limitations)
Traceback (most recent call last):
File "deeplio/train.py", line 69, in
trainer.run()
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 141, in run
self.train(epoch)
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 311, in train
self.post_train_iter()
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 460, in post_train_iter
imgs_remossion = make_grid(imgs_remossion, nrow=2)
File "/home/bernardo/.local/lib/python3.6/site-packages/torchvision/utils.py", line 85, in makegrid
.copy(tensor[k])
RuntimeError: The size of tensor a (3) must match the size of tensor b (4) at non-singleton dimension 0
Hi, thanks for awesome repo! I am having trouble with this traceback and was wondering if you could help me get over it For some context, I tried running all your configs on the same data , only with smaller batch size (GPU limitations)
Traceback (most recent call last): File "deeplio/train.py", line 69, in
trainer.run()
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 141, in run
self.train(epoch)
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 311, in train
self.post_train_iter()
File "/home/bernardo/DeepLIO/deeplio/models/trainer.py", line 460, in post_train_iter
imgs_remossion = make_grid(imgs_remossion, nrow=2)
File "/home/bernardo/.local/lib/python3.6/site-packages/torchvision/utils.py", line 85, in makegrid
.copy(tensor[k])
RuntimeError: The size of tensor a (3) must match the size of tensor b (4) at non-singleton dimension 0
Thanks in advance