赞
踩
解决: sudo -s
然后重新打开即可
sudo gedit /etc/apt/sources.list
改完刷新
sudo apt-get update
sudo apt-get upgrade
sudo apt-get install build-essential
这是一台是1650显卡的迷你电脑。 在ubuntu的软件仓库查看本机相关的显卡信息
卸载显卡驱动
$ sudo apt-get --purge remove nvidia*
$ sudo apt autoremove
查看驱动
ubuntu-drivers devices
然后查下 CUDA对应的版本号,安排470的驱动
安装这个
sudo apt install nvidia-driver-470-server
默认auto安装(不要用这个)
sudo ubuntu-drivers autoinstall
安装完重启后
检查安装状况
nvidia-smi
进链接下载CUDA(这里注意!版本对应tensorrt,因为我RT选的7.2.3.4 。所以版本号不能选择11.1.0的小版本否则在转序列化文件的时候就会出现类似提示。 )
[06/21/2022-10:23:36] [W] [TRT] TensorRT was linked against cuBLAS/cuBLAS LT 11.3.0 but loaded cuBLAS/cuBLAS LT 11.2.1
装错版本按下面方式卸载
cd /usr/local/cuda-xx.x/bin/
sudo ./cuda-uninstaller
sudo rm -rf /usr/local/cuda-xx.x
开始安装-下载11.1.1
如下命令安装
wget https://developer.download.nvidia.com/compute/cuda/11.1.1/local_installers/cuda_11.1.1_455.32.00_linux.run
sudo sh cuda_11.1.1_455.32.00_linux.run 执行安装
安装完成后
gedit ~/.bashrc (编辑在底下增加如下两行指向)
export PATH=/usr/local/cuda-11.1/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-11.1/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
source ~/.bashrc
即可
`
下载连接:https://developer.nvidia.com/rdp/cudnn-archive
将下载好的提取出CUDA包
tar -zxvf cudnn-11.2-linux-x64-v8.1.1.33.tgz
目标文件夹配置包拷贝到cuda底下
sudo cp cuda/include/* /usr/local/cuda/include/
sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64/
sudo chmod a+r /usr/local/cuda/lib64/libcudnn*
版本查看
7.x及以前在cudnn.h中:
cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
8.x后在cudnn_version.h中:
cat /usr/local/cuda/include/cudnn_version.h | grep CUDNN_MAJOR -A 2
后续CUDA相关软链接警告:
/sbin/ldconfig.real: /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_adv_infer.so.8 is not a symbolic link
复制如下贴上去。
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_adv_train.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_adv_train.so.8
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_adv_infer.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_adv_infer.so.8
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_ops_train.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_ops_train.so.8
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_cnn_infer.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn_cnn_infer.so.8
————————————————
sudo ln -sf /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn.so.8.1.1 /usr/local/cuda-11.1/targets/x86_64-linux/lib/libcudnn.so.8
运行下载的sh 文件,然后Enter下翻
bash Anaconda3-2022.05-Linux-x86_64.sh
sudo gedit ~/.bashrc
export PATH="/home/sc/anaconda3/bin:$PATH"(环境包bin地址)
source ~/.bashrc
创建python环境
conda create -n 环境名 python=版本号
如:conda create -n yolov5 python=3.8
conda info --env #查看现有环境
source activate yolov5 环境名 #进入该环境
conda deactivate #退出环境
conda remove -n your_env_name --all #删除环境
conda remove --name your_env_name package_name # 删除虚拟环境中的某个包:
pip install -i 镜像源地址 包名 国内常用源镜像地址: pip install -i 镜像源地址 包名 国内常用源镜像地址: 清华 https://pypi.tuna.tsinghua.edu.cn/simple 阿里云 http://mirrors.aliyun.com/pypi/simple/ 中国科技大学 https://pypi.mirrors.ustc.edu.cn/simple/ 华中理工大学 http://pypi.hustunique.com/ 山东理工大学 http://pypi.sdutlinux.org/ 豆瓣 http://pypi.douban.com/simple/
输入如下
pip install torch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 -i https://pypi.tuna.tsinghua.edu.cn/simple/
离线包安装
http://download.pytorch.org/whl/torch_stable.html
安装完成
下载:wget https://github.com/opencv/opencv/archive/3.4.5.zip
参考 opencv快速编译安装
装完试跑
对应CUDA 下载
下载地址:https://developer.nvidia.com/nvidia-tensorrt-7x-download
解压完后将路径放进 ~/. bashrc
vim ~/.bashrc
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/sc/TensorRT-7.2.3.4/lib
source ~/.bashrc
通常我们还会将配置文件复制到usr
sudo cp -r ./lib/* /usr/lib
sudo cp -r ./include/* /usr/include
下载对应yolov Rt包:https://github.com/wang-xinyu/tensorrtx.git
拷贝gen_wts.py 和权重文件到yolov包下执行
python gen_wts.py -w yolov5s.pt -o yolov5s.wts
转到rt包下 创建build
cmake ..
make
转.engine文件
sudo ./yolov5 -s yolov5s.wts yolov5s.engine s
sudo ./yolov5 -d yolov5s.engine ../samples
while True: ret, img = video_capture.read() if ret: image_raw, use_time = yolov5_wrapper.infer(img) # print('time->{:.2f}ms, saving into output/'.format(use_time * 1000)) #cv2.namedWindow('image',0) # print("image_raw: ",image_raw.shape) image_raw = cv2.resize(image_raw, (1280,720)) cv2.imshow('image', image_raw) cv2.waitKey(1) else: break cv2.destroyAllWindows() video_write.release() yolov5_wrapper.destroy()
运行
deepsort 没截图 就先这样。
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。