AttributeError: module 'torch._C' has no attribute '_cuda_setDevice'

[byungjinku]

Traceback (most recent call last):
File "c:\Users\byung\Desktop\AtlasNet-master\train.py", line 14, in
torch.cuda.set_device(opt.multi_gpu[0])
File "C:\Users\byung\anaconda3\Lib\site-packages\torch\cuda_init_.py", line 350, in set_device
torch._C._cuda_setDevice(device)
^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: module 'torch._C' has no attribute '_cuda_setDevice'

I tried to run your demo but I had this issue

[ThibaultGROUEIX]

Hi @byungjinku

Thanks for trying the demo

It looks like your torch version might not be GPU enabled from a quick google search ; pytorch/pytorch#33655 (comment)

best regards,
Thibault

[byungjinku]

this code supoort only cuda 10.1 ?

we can't use on cuda 11.8?

[ThibaultGROUEIX]

No it should support both.

[byungjinku]

/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/cuda/init.py:104: UserWarning:
NVIDIA GeForce RTX 4080 with CUDA capability sm_89 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 compute_37.
If you want to use the NVIDIA GeForce RTX 4080 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name))
Jitting Chamfer 3D
Traceback (most recent call last):
File "train.py", line 16, in
import training.trainer as trainer
File "/home/bjku/AtlasNet/training/trainer.py", line 13, in
from training.trainer_loss import TrainerLoss
File "/home/bjku/AtlasNet/training/trainer_loss.py", line 2, in
import auxiliary.ChamferDistancePytorch.chamfer3D.dist_chamfer_3D as dist_chamfer_3D
File "/home/bjku/AtlasNet/auxiliary/ChamferDistancePytorch/chamfer3D/dist_chamfer_3D.py", line 15, in
"/".join(os.path.abspath(file).split('/')[:-1] + ["chamfer3D.cu"]),
File "/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 997, in load
keep_intermediates=keep_intermediates)
File "/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1202, in _jit_compile
with_cuda=with_cuda)
File "/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1293, in _write_ninja_file_and_build_library
with_cuda=with_cuda)
File "/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1635, in _write_ninja_file_to_build_library
cuda_flags = common_cflags + COMMON_NVCC_FLAGS + _get_cuda_arch_flags()
File "/home/bjku/miniconda3/envs/atlasnet/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1438, in _get_cuda_arch_flags
raise ValueError("Unknown CUDA arch ({}) or GPU not supported".format(arch))
ValueError: Unknown CUDA arch (8.9) or GPU not supported

I'tried to resoleve but i couldn't