RuntimeError happened when run test.py to check coco dataset AP

[Erissonleo]

My platform is win10, gtx1060, cuda10, pytorch1.1.0, and I could run demo.py, but when I run python test.py ctdet --dataset coco --exp_id coco_dla --keep_res --load_model ../models/ctdet_coco_dla_2x.pth it happens

RuntimeError: cuda runtime error (71) : operation not supported at C:\w\1\s\tmp_conda_3.6_035809\conda\conda-bld\pytorch_1556683229598\work\torch/csrc/generic/StorageSharing.cpp:245

and I can't solve it. Could anyone help me? Thanks.

[xingyizhou]

I have no idea.

[howdidthatwork]

My platform is win10, gtx1060, cuda10, pytorch1.1.0, and I could run demo.py, but when I run python test.py ctdet --dataset coco --exp_id coco_dla --keep_res --load_model ../models/ctdet_coco_dla_2x.pth it happens

RuntimeError: cuda runtime error (71) : operation not supported at C:\w\1\s\tmp_conda_3.6_035809\conda\conda-bld\pytorch_1556683229598\work\torch/csrc/generic/StorageSharing.cpp:245

and I can't solve it. Could anyone help me? Thanks.

I have the same issue

[Fansgithub2019]

I disable the multi-thread dataloader by --not_prefetch_test to solve the problem.refer to https://github.com/xingyizhou/CenterNet/issues/126#issuecomment-499606727