qinnzou/DeepCrack

RuntimeError: CUDA out of memory

ousim97 opened this issue · 3 comments

Hi,

I have an issue in the test.py code with the pretrained model. Everytime I run it I got the following error:
RuntimeError: CUDA out of memory. Tried to allocate 120.00 MiB (GPU 0; 4.00 GiB total capacity; 2.48 GiB already allocated; 105.14 MiB free; 2.51 GiB reserved in total by PyTorch)

Note that I have reduced the batch size as suggested in other topics but I still get the same eroor. Do you have a solution please?

Thanks in advance

An nvidia 1060 or up gpu may satisfy the requirement.

ME TOO. Seems 4GB is not enough cause the batchsize has been set to 1

large GPU memory is required.