Why is it that when I run pointpillar, I need to ‘converting 3D prediction to kitti format’ at val, which is fast the first time, but slow the second time, and takes almost an hour the second time. Why is this?

[HoyinLau]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• 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

basic envirnment according to the docs

Reproduces the problem - code sample

only use the original setting

Reproduces the problem - command or script

only use the original setting

Reproduces the problem - error message

only use the original setting

Additional information

No response