mmdetection 环境配置

1. 配置清华源镜像通道

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/

# 搜索时显示通道地址

conda config --set show_channel_urls yes

2. 新建环境并安装指定版本的依赖包

# python3.6 && pytorch1.6
conda -n create mmdet python=3.6
# pytorch=1.7 在coco上eval时报错
conda install pytorch=1.6.0 cudatoolkit=10.2 torchvision torchaudio


# python3.8 && pytorch1.8
conda -n create mmdet python=3.8
conda install pytorch=1.8.0 cudatoolkit=10.2 torchvision torchaudio

 

安装mmcv-full 

# pytorch1.6
pip install mmcv-full==1.2.2 -f https://download.openmmlab.com/mmcv/dist/cu102/torch1.6.0/index.html

# pytorch1.8
pip install mmcv-full==1.3.16 -f https://download.openmmlab.com/mmcv/dist/cu102/torch1.8.0/index.html

可编辑模式安装mmdection

pip install -v -e .

3. 检查安装