'std::bad_alloc' error while trying the demo

[acaglayan]

trafficstars

Hi, while trying the below command:

python demo_folder.py --demo-folder demo_folder --save-folder demo_save

I had the following error:

terminate called after throwing an instance of 'std::bad_alloc' what(): std::bad_alloc Aborted

greatly appreciate any idea to fix the problem. thanks.

[xinge008]

Can you show more details about your environment seting? such as spconv, pytorch, etc;

From the error, it seems a mistake for the installation of spconv.

[acaglayan]

Hi, I'm using the below packages on a debian 10.11 system:

cudatoolkit 11.3.1 cumm-cu113 0.2.6 cython 0.29.24
numpy 1.21.2
pillow 8.4.0
python 3.7.11
python-dateutil 2.8.2
pytorch 1.10.0
pytorch-mutex 1.0
spconv-cu113 2.1.19
torch-scatter 2.0.9
torchaudio 0.10.0
torchvision 0.11.1
yaml 0.2.5

I'm trying to use the trained model on a custom dataset. I think my error is related to that. I couldn't find info on how to apply a trained model to a custom data. So I tried to modify config files for the new data but couldn't make it run. thank you again.