cuda,cudnn,Anaconda,OpenCV,Darknet,caffe安装文档
what(): OpenCV(3.4.2) /home/arthas/opencv/modules/highgui/src/window.cpp:632: error: (-2:Unspecified error) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Carbon support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function 'cvShowImage'
需要将opencv卸载并安装相应依赖库后从新安装opencv
$ cd opencv/build
$ sudo make uninstall
$ cd ..
$ sudo rm -r build
$ sudo rm -r /usr/local/include/opencv2 /usr/local/include/opencv /usr/include/opencv /usr/include/opencv2 /usr/local/share/opencv /usr/local/share/OpenCV /usr/share/opencv /usr/share/OpenCV /usr/local/bin/opencv* /usr/local/lib/libopencv*
$ sudo apt-get --purge remove opencv-doc opencv-data python-opencv
$ sudo apt-get install gcc g++ cmake pkg-config build-essential
$ sudo apt-get install libgtk2.0-dev libavcodec-dev libavformat-dev libtiff5-dev libswscale-dev libjasper-dev
$ sudo apt-get install pkg-config
pip提示You are using pip version 10.0.1, however version 18.0 is available.You should consider upgrading via the 'pip install --upgrade pip' command.
$ pip install --upgrade pip
$ pip install msgpack-python
$ pip install msgpack
darknet报错：./darknet: error while loading shared libraries: libopencv_highgui.so.3.4: cannot open shared object file: No such file or directory
$ cd /usr/local/cuda-9.2/
$ sudo ldconfig
==> WARNING: A newer version of conda exists. <==
current version: 4.5.4
latest version: 4.5.11
使用命令升级
$ conda update -n base conda
$ conda create -n <虚拟环境名> python=<指定python版本>
$ conda remove -n <虚拟环境名> --all
$ conda env list
$ source activate <虚拟环境名>
$ activate <虚拟环境名>
$ source deactivate
$ deactivate
$ conda search <查找包的名字>
$ cd ~/caffe/python
$ python
>>> import caffe
>>> caffe.set_mode_gpu() # 设置为gpu模式
$ cd caffe/build
$ make uninstall
$ cd ..
$ rm -rf build
然后进入到caffe/python中执行
$ for req in $(cat requirements.txt); do pip install $req; done
安装一些依赖包之后进行一下操作
$ cd caffe
$ mkdir build
$ cd build
$ cmake ..
$ make all -j8
执行完以上命令后先别执行make install,先回到caffe根路径
$ make pycaffe
$ make pytest
import caffe遇到报错：raise ValueError, "Can't create weekday with n == 0"执行
$ pip install python-dateutil --upgrade
caffe.set_mode_gpu()报错AttributeError: module 'caffe' has no attribute 'set_mode_gpu'
$ cd ~/caffe/python
$ python
>>> import caffe
>>> caffe.set_mode_gpu()
继续执行make pytest,遇到ImportError: No module named 'pydotplus'和ImportError: No module named 'pydot'执行
$ pip install pydotplus
若有其他包未安装报错同上面方法一样找到相应的包安装，直到所有包安装完成后在执行
$ make install
$ make runtest
建议在虚拟环境下安装
访问pytorch官网选择系统相应的版本选项后使用命令进行安装
建议在虚拟环境下安装
使用pip进行安装的很有可能会报error while loading shared libraries: libopencv_highgui.so.x.x: cannot open shared object file: No such file or directory的错，并且去/usr/local/cuda-9.2目录下使用sudo ldconfig命令也不一定有用，使用cond可以很好的避免这个问题
先查看可安装的版本
$ conda install tensorflow-gpu==<需要安装的版本号>
先查看可安装的版本
$ conda install tensorflow-cpu==<需要安装的版本号>
$ nvidia-smi
$ watch -n 1 nvidia-smi
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent