RuntimeError: CUDA unknown error - this may be due to an incorrectly set up environment, e.g. $ python3 -c 'import torch torch.rand(3).cuda()'įile "/home/user/dev/miniconda3/lib/python3.8/site-packages/torch/cuda/_init_.py", line 172, in _lazy_init Return torch._C._cuda_getDeviceCount() > 0 Setting the available devices to be zero. changing env variable CUDA_VISIBLE_DEVICES after program start. home/user/dev/miniconda3/lib/python3.8/site-packages/torch/cuda/_init_.py:52: UserWarning: CUDA initialization: CUDA unknown error - this may be due to an incorrectly set up environment, e.g. Please let me know if you have any suggestions on troubleshooting this issue.įollowing results are from the RTX3090 machine Miniconda env $ python3 -c 'import torch print(_available())' nvidia-driver: 450.80, from Debian testing.nvidia-cuda-toolkit: 11.0.3-2, from Debian testing.nvidia-driver: 455.38, from Debian experimental.However, I do not know the commonalities between the two machines The same error also occurs in a separate (new) machine with Quadro RTX 5000, leading me to speculate this could be a setup error. However, I started getting errors when trying to put variables into GPU with.
CUDA DIM3 INITIALIZATION DRIVERS
I recently obtained a RTX3090, and had to make appropriate updates on nvidia drivers for Ampere architecture support.
I was running Pytorch without issues using GTX 1080 Ti. See latest post for CUDA error: all CUDA-capable devices are busy or unavailable Resolved issue
0 kommentar(er)
