Jetson Nano 主板YoloV5 环境部署

1. cuda 安装

wget https://github.com/conda-forge/miniforge/releases/download/22.9.0-1/Mambaforge-22.9.0-1-Linux-aarch64.sh

sh Miniforge3-Linux-aarch64.sh
source ~/.bashrc

配置国内镜像源：

conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/msys2/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/menpo/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/bioconda/
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/r
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/pro
conda config --add channels http://mirrors.tuna.tsinghua.edu.cn/anaconda//cloud/simpleitk
conda config --set show_channel_urls yes

2. 如果环境变量未配置需手动配置：

# >>> conda initialize >>>
# !! Contents within this block are managed by 'conda init' !!
__conda_setup="$('/home/nvidia/mambaforge/bin/conda' 'shell.bash' 'hook' 2> /dev/null)"
if [ $? -eq 0 ]; then
    eval "$__conda_setup"
else
    if [ -f "/home/nvidia/mambaforge/etc/profile.d/conda.sh" ]; then
        . "/home/nvidia/mambaforge/etc/profile.d/conda.sh"
    else
        export PATH="/home/nvidia/mambaforge/bin:$PATH"
    fi
fi
unset __conda_setup

if [ -f "/home/nvidia/mambaforge/etc/profile.d/mamba.sh" ]; then
    . "/home/nvidia/mambaforge/etc/profile.d/mamba.sh"
fi
# <<< conda initialize <<<

2. 安装torch

2. torch 安装
基于 python3.6 安装方式(以下方式为 1.7.0 为例):
wget https://nvidia.box.com/shared/static/p57jwntv436lfrd78inwl7iml6p13fzh.whl -O
torch-1.7.0-cp36-cp36m-linux_aarch64.whl
sudo apt-get install python3-pip libopenblas-base libopenmpi-dev
pip3 install Cython -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install numpy torch-1.7.0-cp36-cp36m-linux_aarch64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple

3. torchvision 安装

sudo apt-get install libjpeg-dev zlib1g-dev libpython3-dev libavcodec-dev libavformat-dev
libswscale-dev
git clone --branch v0.x.0 https://github.com/pytorch/vision torchvision
cd torchvision
export BUILD_VERSION=0.x.0 # (填写 torchvision 安装的版本号)
python3 setup.py install

vim ~/.bashrc
#添加
export OPENBLAS_CORETYPE=ARMV8
source ~/.bashrc

4. 验证

import torch
print(torch.__version__)
print('CUDA available: ' + str(torch.cuda.is_available()))
print('cuDNN version: ' + str(torch.backends.cudnn.version()))
a = torch.cuda.FloatTensor(2).zero_()
print('Tensor a = ' + str(a))
b = torch.randn(2).cuda()
print('Tensor b = ' + str(b))
c = a + b
print('Tensor c = ' + str(c))
import torchvision
print(torchvision.__version__)

5. 安装依赖

先打开requirements.txt注释：torchvision、 torch 2行。

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