search

owruby / APE-GAN

Pytorch Implementation of APE-GAN
14 stars 1 forks source link

Hello!There is an error occured:RuntimeError: CUDA out of memory. Tried to allocate 32.00 MiB (GPU 0; 2.00 GiB total capacity; 757.17 MiB already allocated; 24.89 MiB free; 44.83 MiB cached) #1

Open zhjikoshlizhzc opened 5 years ago

zhjikoshlizhzc commented 5 years ago

File "generate.py", line 137, in main(args) File "generate.py", line 111, in main x_adv = fgsm(model, x, t, loss_func, eps) File "D:\Graduate\APE-GAN-master\utils.py", line 14, in fgsm loss.backward(retain_graph=True) File "C:\Users\acer\Anaconda3\lib\site-packages\torch\tensor.py", line 102, in backward torch.autograd.backward(self, gradient, retain_graph, create_graph) File "C:\Users\acer\Anaconda3\lib\site-packages\torch\autograd__init__.py", line 90, in backward allow_unreachable=True) # allow_unreachable flag RuntimeError: CUDA out of memory. Tried to allocate 32.00 MiB (GPU 0; 2.00 GiB total capacity; 757.17 MiB already allocated; 24.89 MiB free; 44.83 MiB cached) Do you have methods to solve it?Thanks!

ritalsq commented 5 years ago

Hi senior， here are some questions，I think i need your help！ 图片 would you have some answers？