bug show_results

[iceriver97]

Traceback (most recent call last):
  File "./tools/test.py", line 263, in <module>
    main()
  File "./tools/test.py", line 228, in main
    outputs = single_gpu_test(model, data_loader, args.show, args.show_dir)
  File "/home/Dataset/weiqian/mmdetection3d/mmdet3d/apis/test.py", line 53, in single_gpu_test
    out_dir=out_dir,
  File "/home/Dataset/weiqian/mmdetection3d/mmdet3d/models/detectors/mvx_two_stage.py", line 466, in show_results
    if isinstance(data['points'][0], DC):
KeyError: 'points'

[YoushaaMurhij]

I met the same error when I tried to launch non dist. test script.
python $(dirname "$0")/test.py $CONFIG $CHECKPOINT --eval bbox --show

[RoboticsYimingLi]

Is this problem solved? @YoushaaMurhij @iceriver97

[a2801603396]

I met the same error. Counld you tell me how to solved it?

[a2801603396]

If I don't add - show during testing, I won't report this error. But I also report this error during training. This error will be reported when training 1 epochs validation.

[RoboticsYimingLi]

Please use dist mode. Thanks!

[h0n1]

Please use dist mode. Thanks!

It seems that show mode is not supported when use non dist mode

if not distributed:
# assert False
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show, args.show_dir)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = custom_multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)