JetsonNano国内环境配置

因为网络原因，Nvidia的资源在国内很慢，这里提供解决方案。

apt

不难理解，类比ubuntu换源，替换/etc/apt/sources.list
注意：Jetson处理器为aarch64，建议用架构齐全的中科大源。

deb https://mirrors.ustc.edu.cn/ubuntu-ports/ bionic main restricted universe multiverse
deb https://mirrors.ustc.edu.cn/ubuntu-ports/ bionic-updates main restricted universe multiverse
deb https://mirrors.ustc.edu.cn/ubuntu-ports/ bionic-backports main restricted universe multiverse
deb https://mirrors.ustc.edu.cn/ubuntu-ports/ bionic-security main restricted universe multiverse

保存更新sudo apt update

pip

深度学习依赖pip，和apt一样换成国内源

sudo apt install python3-pip
pip3 install -i https://pypi.douban.com/simple/ -U pip
pip3 config set global.index-url https://pypi.douban.com/simple/

torch

PyTorch比较特殊，基础pip源无，手动安装whl
官方匹配如下

PyTorch v1.0 - torchvision v0.2.2
PyTorch v1.1 - torchvision v0.3.0
PyTorch v1.2 - torchvision v0.4.0
PyTorch v1.3 - torchvision v0.4.2
PyTorch v1.4 - torchvision v0.5.0
PyTorch v1.5 - torchvision v0.6.0
PyTorch v1.6 - torchvision v0.7.0
PyTorch v1.7 - torchvision v0.8.1

这里选用1.6 + 0.7

安装torch

我传了一份在国内网盘（注意torch1.5开始不再支持python2）
链接: https://pan.baidu.com/s/1Qz7qkXSOw7tlpRP-kO2V0w
提取码: ur6t

sudo apt install python3-pip libopenblas-base libopenmpi-dev 
pip3 install Cython
pip3 install numpy torch-1.7.0-cp36-cp36m-linux_aarch64.whl

安装vision

同步迁移至国内gitee镜像加速

sudo apt install libjpeg-dev zlib1g-dev libpython3-dev libavcodec-dev libavformat-dev libswscale-dev
git clone --branch v0.7.0 https://gitee.com/azureology/torchvision.git
cd torchvision
export BUILD_VERSION=0.7.0
python3 setup.py install --user

参考

https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-7-0-now-available/72048