先导片：Jetson采用arm64架构

一、安装jetson5.1.2

安装好jetson自带cuda、cudnn和tensorRT

官方文档

1. 更换源

   sudo vi /etc/apt/sources.list.d/nvidia-l4t-apt-source.list
   

   

2. 更新源

   sudo apt install python3-pip
   

   sudo -H pip3 install -U pip
   

   sudo -H pip install jetson-stats
   

3. 安装jetson

   sudo apt install nvidia-jetpack
   

二、安装jtop

1. 安装

   sudo apt install python3-pip
   sudo -H pip3 install -U pip
   sudo -H pip install jetson-stats
   

2. 查看

   sudo jtop
   

   在这里插入代码片

三、配置CUDA和cuDNN

1. 配置环境变量

   vim ~/.bashrc
   

   增加以下内容

   export LD_LIBRARY_PATH=/usr/local/cuda/lib64
   export PATH=/usr/local/cuda/bin:$PATH
   export CUDA_HOME=/usr/local/cuda

2. 更新环境变量

   source ~/.bashrc
   

3. 复制文件到cuda目录

   cd /usr/include && sudo cp cudnn* /usr/local/cuda/include
   cd /usr/lib/aarch64-linux-gnu && sudo cp libcudnn* /usr/local/cuda/lib64
   

4. 查看CUDA

   nvcc -V
   

   

5. 修改文件权限

   sudo chmod 777 /usr/local/cuda/include/cudnn.h 
   sudo chmod 777 /usr/local/cuda/lib64/libcudnn*
   

6. 重新软链接

   #这里的8.6.0和8对应安装的cudnn版本号和首数字
   cd /usr/local/cuda/lib64
   
   sudo ln -sf libcudnn.so.8.6.0 libcudnn.so.8
   
   sudo ln -sf libcudnn_ops_train.so.8.6.0 libcudnn_ops_train.so.8
   sudo ln -sf libcudnn_ops_infer.so.8.6.0 libcudnn_ops_infer.so.8
   
   sudo ln -sf libcudnn_adv_train.so.8.6.0 libcudnn_adv_train.so.8
   sudo ln -sf libcudnn_adv_infer.so.8.6.0 libcudnn_adv_infer.so.8
   
   sudo ln -sf libcudnn_cnn_train.so.8.6.0 libcudnn_cnn_train.so.8
   sudo ln -sf libcudnn_cnn_infer.so.8.6.0 libcudnn_cnn_infer.so.8
   
   sudo ldconfig
   

   7. 查看cuDNN

   dpkg -l libcudnn8
   

   
   8. 报错libopenblas.so.0: cannot open shared object file: No such file or director
   

   sudo apt-get install libopenblas-dev
   

   

四、安装Pytorch

torch官方教程
torch下载文档
torch下载地址
torch2.0.0对应torchvision0.15.1

pip install torch-2.0.0+nv23.05-cp38-cp38-linux_aarch64.whl

torchvision下载

sudo apt-get install libjpeg-dev zlib1g-dev libpython3-dev libopenblas-dev libavcodec-dev libavformat-dev libswscale-dev

将解压后的文件夹命名为torchvision

export BUILD_VERSION=0.15.1

python3 setup.py install --user