RuntimeError: cuda runtime error (8)

Accelerated Computing CUDA CUDA Programming and Performance
1

Hello ,I am a novice ,just beginning to learn how to use the CUDA tools.When I run a piece of code, the console raised such a message, and I do not know how to fix it, could you help me?

Error message:
RuntimeError: cuda runtime error (8) : invalid device function at c:\users\administrator\downloads\new-builder\win-wheel\pytorch\aten\src\thc\generic/THCTensorMathPointwise.cu:265

Environment：
operation system : win10; python_version ：3.6.5 ; cuda_version : 8.0.61; pytorch_version: 0.4.0; GPU: GeForce GTX 760; GPU driver_version: 398.82

x = torch.Tensor( [[ 1.,  1.,  1.],
                   [ 1.,  2.,  1.],
                   [ 1.,  3.,  1.],
                   [ 1.,  4.,  1.]])   
y = torch.rand(4,3)

if torch.cuda.is_available() :
    x = x.cuda()
    y = y.cuda()
    x + y
2

It means your pytorch was compiled with a set of architecture flags for GPU code that don’t match your GPU.

Your GPU is a compute capability 3.0 GPU.

[url]https://discuss.pytorch.org/t/pytorch-no-longer-supports-this-gpu-because-it-is-too-old/13803[/url]