Open ani16 opened 4 years ago
CUDA 8.0 doesn't support Quadro K2200 GPU and Kepler architecture. It isn't related to Darknet.
Hi
Thanks for reply. please can you tell what cuda version should I use to train the yolo model with own dataset with this GPU. I.e K2200. Actually with cpu, training is very slow. Earlier I was having k40 GPU , training was coming very well with that.
With current GPU I am facing problem. Can you please suggest?
Thanks in advance.
On Wed, Oct 7, 2020, 10:16 PM Alexey notifications@github.com wrote:
CUDA 8.0 doesn't support Quadro K2200 GPU and Kepler architecture. It isn't related to Darknet.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/AlexeyAB/darknet/issues/6801#issuecomment-705060997, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFLNCW26EOFEMM7H6EXHNITSJSLPFANCNFSM4SHUGTSQ .
Hi
I am facing compiling error with cuda 8.0 on Quadro K2200 GPU.
How to resolve this. Shall I use yolov3 repo for the same?