Runtime Assertion Error

[hwuebben]

Traceback (most recent call last):
File "ssim_fused.py", line 33, in
ssim_value = fused_ssim(predicted_image, gt_image)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File ".venv\Lib\site-packages\fused_ssim_init_.py", line 40, in fused_ssim
map = FusedSSIMMap.apply(C1, C2, img1, img2, padding, train)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File ".venv\Lib\site-packages\torch\autograd\function.py", line 575, in apply
return super().apply(*args, **kwargs) # type: ignore[misc]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File ".venv\Lib\site-packages\fused_ssim_init_.py", line 11, in forward
ssim_map, dm_dmu1, dm_dsigma1_sq, dm_dsigma12 = fusedssim(C1, C2, img1, img2, train)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: t == DeviceType::CUDA INTERNAL ASSERT FAILED at ".venv/Lib/site-packages/torch/include\c10/cuda/impl/CUDAGuardImpl.h":28, please report a bug to PyTorch.

Greetings,

Can't get past this point.
I use win11. installed torch (2.5.1+cu124) and cuda (12.6) according to docs. Everything seems to run fine.

torch.cuda.is_available()
torch.version.cuda

shows True and 12.4

Any ideas?
Thank you