yolov9
yolov9 copied to clipboard
Abnormal segmentation training for YOLOv9
When training with segment-train_dual.py, the loss values of val are all 0; And the last result of mAP or recall are always unstable
我也是,请问怎样解决?
可能是cuda环境的配置问题,看看电脑是cuda多少,再看看自己配置的cuda是多少,我就是这个问题修改了