About nuscenes datasets: test in ./projects/BEVFusion

Prerequisite

[X] I have searched Issues and Discussions but cannot get the expected help.
[X] I have read the FAQ documentation but cannot get the expected help.
[X] The bug has not been fixed in the latest version (dev-1.x) or latest version (dev-1.0).

Task

I'm using the official example scripts/configs for the officially supported tasks/models/datasets.

Branch

main branch https://github.com/open-mmlab/mmdetection3d

Environment

Reproduces the problem - code sample

Reproduces the problem - command or script

python tools/test.py /home/lxt/mmdetection3d/projects/BEVFusion/configs/bevfusion_lidar-cam_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py /home/lxt/mmdetection3d/checkpoints/bevfusion_lidar-cam_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d-5239b1af.pth --show --show-dir ./result/nuscenes --task multi-modality_det

Reproduces the problem - error message

When I used the above command to infer bevfusion on the Nuscenes test set, the following error occurred

Afterwards, I changed the task to multi-view_det,The project can operate normally and obtain Map and NDS. But -- show did not work and did not show prediction results. There is also no file generation in ./result/nuscenes

What surprised me was that Bevfusion is a multimodal 3D detection model, but its chosen task is multi-modality_det, but it can operate normally, although it cannot produce predicted PKL results. I am certain that my dataset is complete, but errors may occur when choosing multimodal tasks.

Can you please explain to me which task should be selected when I test bevfusion? And why executing the above instructions cannot visualize and generate prediction results?

Thanks!

Additional information

No response

open-mmlab / mmdetection3d