MMDetection安装问题

问题一：Install MMDetection报错

pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/{
    cu_version}/{
    torch_version}/index.html

本人环境
cu_version10.1 （输入nvcc -V命令,即可查询到CUDA版本号。）
torch_version1.10.2

mmcv-full is only compiled on PyTorch 1.x.0 because the compatibility
usually holds between 1.x.0 and 1.x.1. If your PyTorch version is
1.x.1, you can install mmcv-full compiled with PyTorch 1.x.0 and it usually works well.
https://github.com/open-mmlab/mmcv#installation

 pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu101/torch1.10.0/index.html

出现报错

查看了一下 PyTorch 和 CUDA 版本组合，发现没有我们对应的版本，降低torch版本，将1.10.2降为1.8.1
https://github.com/open-mmlab/mmcv/blob/master/README_zh-CN.md

整体步骤：

#1.Prepare environment
conda create -n openmmlab python=3.7 -y
conda activate openmmlab
conda install pytorch=1.8.1 cudatoolkit=10.1 torchvision -c pytorch
#2.Install mmcv-full.
 pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu101/torch1.8.0/index.html
#3.Install MMDetection.
git clone https://github.com/open-mmlab/mmdetection.git
cd mmdetection
pip install -r requirements/build.txt
pip install -v -e .  # or "python setup.py develop"

问题二 fatal: unable to access…

在git下载项目时遇到问题：

fatal: unable to access 'https://github.com/open-mmlab/mmdetection.git/': gnutls_handshake() failed: A TLS packet with unexpected length was received.

解决方案：
将https改为git重新执行

git clone git://github.com/open-mmlab/mmdetection.git

问题三:ImportError: cannot import name ‘PILLOW_VERSION’ from ‘PIL’

方法2.根据报错的最后一行提示，打开function.py文件，使用from PIL import Image, ImageOps, ImageEnhance, version 替换文件中from PIL import Image, ImageOps, ImageEnhance,PILLOW_VERSION这句。

参考解决方案：https://blog.csdn.net/Lee_lg/article/details/103901632

当cuda 11
https://pytorch.org/get-started/previous-versions/

conda install pytorch==1.7.0 torchvision==0.8.0 torchaudio==0.7.0 cudatoolkit=11.0 -c pytorch

镜像问题https://blog.csdn.net/weixin_49304494/article/details/122622134