Problem in running yolov8 on gpu with pytorch cuda()

Robotics & Edge Computing Jetson & Embedded Systems Jetson Orin Nano
1

hi, im currently using orin nano and i have got cuda 11.4 and pytorch 2.0.1 installed, but checking with

print(torch.cuda.is_available())

still returns False. what should i do?
thanks for the response, please help…

3

Hi,

Please install PyTorch with the below instructions to get the GPU support:

Thanks.

4

thanks for the reply, it works and the GPU compatible torch is already installed. But i’m currently getting into new problem that my torchvision version (that installed along with ultralytics) is not compatible with the torch that i installed recently. The error sounds like this

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
torchvision 0.17.1 requires torch==2.2.1, but you have torch 2.0.0+nv23.5 which is incompatible.

Please give any guidance, thank you…

5

Hi,

For torch 2.0.0+nv23.5, please use 0.14.1 TorchVision.
The compiling instructions can be found below:

Thanks.

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.