PyCenterNet2

1. V100
2. conda create -n myenv python=3.8
3. conda activate myenv
4. conda install pytorch=1.10 cudatoolkit=11.3 -c pytorch
5. pip install -U numpy==1.23.5
6. pip install mmcv-full==1.4.0 -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.10/index.html
7. python setup.py develop
8. pip install torchvision==0.11.3 -i https://pypi.tuna.tsinghua.edu.cn/simple
9. pip install huggingface_hub safetensors -i https://pypi.tuna.tsinghua.edu.cn/simple
10. pip install matplotlib -i https://pypi.tuna.tsinghua.edu.cn/simple
    參考連結：https://blog.csdn.net/qq_36622589/article/details/124355564