All types object class AP=0

[YHao29]

I have a very weird problem. Here's what I have tried:

1. I install the environment on my server with NVIDIA A6000

2. Prepared the full version nuscenes dataset and established a link to /data/nuscenes

3. prepared the pretrained model like

4. run: torchpack dist-run -np 1 python tools/test.py configs/nuscenes/det/transfusion/secfpn/camera+lidar/swint_v0p075/convfuser.yaml pretrained/bevfusion-det.pth --eval bbox --out result/test.pkl

it goes well with no error, only two warnings:

1. NumbaDeprecationWarning: The 'nopython' keyword argument was not supplied to the 'numba.jit' decorator. The implicit default value for this argument is currently False, but it will be changed to True in Numba 0.59.0. See https://numba.readthedocs.io/en/stable/reference/deprecation.html#deprecation-of-object-mode-fall-back-behaviour-when-using-jit for details.
2. UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').

But the results are weird:

So are there any tips for me to fix it?