I am using a dataset with 121 classes, I followed the guidelines given at #8003 , however, I did not find any functional code to replace PALETTE = None.
I also have an old version of mmdetection on another PC and the guide at #8003 helped me successfully to replace PALETTE = None.
Now I have a bigger GPU and I installed a newer version of mmdetection,
I want to display bounding boxes with test.py but facing this issue "The length of the palette should not be less than num_classes."
Below are the details of the dataset and installation environment.
My command line: CUDA_VISIBLE_DEVICES=0 python tools/test.py configs/yolact/yolact_r50_1xb8-55e_coco.py work_dirs/yolact_r50_1xb8-55e_coco/epoch_55.pth --show-dir work_dirs
I have setup mmdetection with this guide https://mmdetection.readthedocs.io/en/latest/get_started.html I think I have the latest version of mmdetection with mmengine, we can see the installation
pip install -U openmim mim install mmengine mim install "mmcv>=2.0.0"
I am using a dataset with 121 classes, I followed the guidelines given at #8003 , however, I did not find any functional code to replace PALETTE = None.
I also have an old version of mmdetection on another PC and the guide at #8003 helped me successfully to replace PALETTE = None.
Now I have a bigger GPU and I installed a newer version of mmdetection,
Is there any possibility to solve this problem.
Thank you