Yanqi Ma

I have also got this problem during run `torchpack dist-run -np 1 python tools/test.py configs/nuscenes/seg/fusion-bev256d2-lss.yaml pretrained/bevfusion-seg.pth --eval map`. I don't know how to evaluate the pretrained model for the segmentation...

No, I haven't found a solution yet.

same problem. The "cu"-setting in "code-runner.executorMap" doesn't work.

I met the same issue. Are there any solutions?