Anaconda+Rtx1080TI+Centos+yolov5

1. 安装Anaconda

wget -c https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2020.02-Linux-x86_64.sh

bash Anaconda3-2020.02-Linux-x86_64.sh

conda --version # conda 4.8.2

2. 使用anaconda，创建新的虚拟环境

配置国内镜像源
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/conda-forge/

设置搜索时显示通道地址

conda config --set show_channel_urls yes

conda create -n yolov5_env python=3.8 (conda remove -n yolov5_env --all)
conda env list
conda activate yolov5_env

3. 安装CUDA10.1

conda install cudatoolkit=10.1 -n yolov5_env -c https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/linux-64/

conda install cudnn=7.6.5 -n yolov5_env -c https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/linux-64/

4. 安装pytorch

conda install pytorch torchvision cudatoolkit=10.1 -c pytorch

5. 下载yolov5 code from github

git clone https://github.com/ultralytics/yolov5.git

cd yolov5

6.安装依赖库

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple -r requirements.txt

7.运行测试文件

python detect.py