"AssertionError: Torch not compiled with CUDA enabled"

I get an error “AssertionError: Torch not compiled with CUDA enabled”. I’m using GTX950 gpu. CUDA was available 1 week ago, but not anymore. And I can’t solve this problem. I could not find the solution to the problem in any title related to this subject. Please help. Thank you in advance

Check if you’ve updated any drivers, if the GPU is still recognized by nvidia-smi, or if you’ve accidentally installed the CPU version of PyTorch.

Hi, Since I am using Win10, I did nvcc - version control and the result is as follows. Also the pytorch version isn’t the cpu version.

nvcc: NVIDIA ® Cuda compiler driver
Copyright © 2005-2020 NVIDIA Corporation
Built on Wed_Jul_22_19:09:35_Pacific_Daylight_Time_2020
Cuda compilation tools, release 11.0, V11.0.221
Build cuda_11.0_bu.relgpu_drvr445TC445_37.28845127_0

Are you able to compile and/or run any other CUDA applications?

Sorry, I might have misunderstood. Saying “CUDA application” are you talking about a .cu file or any nvcc command usage? By the way i can access GPU and use it, using “numba” library.