This series of articles by the @yhl_leo produced, reproduced please indicate the source.
Article Link: http://blog.csdn.net/yhl_leo/article/details/50961542
Spent a day, installed on the computer configuration of the Caffe deep learning framework, many of the online tutorials and guidance have expired, the middle of the time spent a bit, here the personal thought the simplest way to organize the following.
version 1
- Notebook: ThinkPad W541
- Ubuntu 14.04 (64-bit)
- Dual Graphics:
- Intel (R) HD Graphics Family
- NVIDIA Quadro k2100m
- CUDA 7.5
- CuDNN
2 Preparatory work
The system originally wanted to install version wily, has been in use, but helpless, the computer can not install the version, the choice of 14.04.
Install some basic dependencies:
$ sudo apt-get Install Libprotobuf-dev libleveldb-dev libsnappy -dev libopencv-dev libhdf5-serial -dev protobuf- Compiler $ sudo apt-get install -- no< Span class= "Hljs-attribute" >-install -recommends libboost-all -dev $ sudo apt-get install Libopenblas-dev liblapack-dev libatlas-base -dev
To install the NVIDIA graphics driver, I am uninstalling the existing version before installing and uninstalling the method:
sudo apt-get--purdg nvidia*
After restarting the login, you may be looping to fill in the password, can not enter the system, press Ctrl+Alt+F1
, login, install video driver:
$ sudo apt-get install nvidia-current
Or specify a version:
$ sudo add-apt-repository ppa:xorg-edgers/ppa$ sudo apt-get update$ sudo apt-get install nvidia-346
Computer is a dual graphics card, users recommend to install before the installation of some services to ensure success. When I installed it, I only banned one item:
sudo service lightdm stop
(After the system starts, do not log into the desktop, directly press Ctrl+Alt+F1
, if you play the cloud server, should be very familiar with, then enter the system user name and login password, enter the above instructions, back to the graphical interface method is to press Ctrl+Alt+F7
, remember this operation, 14.04 version of the appearance of the password can not enter the desktop is more common, I installed Sogou pinyin have appeared ... )
3 CUDA
Many of the online installation of 6.x,7.0 version of the method is very cumbersome, personal selection of a relatively simple installation method.
First download the installation files on the official website (the link has been provided earlier):
I downloaded the deb
local installation file, and after the download is complete, follow the instructions in the documentation to install:
-i cuda-repo-ubuntu1404-7-5-local_7.5-18_amd64.deb$ sudo apt-get update$ sudo apt-get install cuda
Download the below Installation Guide for Linux
, which has a set of environment variables:
$ PATH=/usr/local/cuda-7.5/bin:$PATH$ LD_LIBRARY_PATH=/usr/local/cuda-7.5/lib64:$LD_LIBRARY_PATH
The configuration of other installation files is also available in the documentation.
4 Caffe
Download Caffe:
$ https://github.com/BVLC/caffe.git
If you don't have Git installed, read the blog: Ubuntu git installs and uses.
Then compile the Caffe:
cp Makefile.config.example Makefile.config$ make all
Some tutorials also configure OPENCV, which can be read in the blog: Ubuntu configuration OpenCV.
5 CuDNN
The CUDNN is a library of GPU-accelerated compute deep neural networks (download link, provided previously).
My download file is:cudnn-7.0-linux-x64-v4.0-prod.tgz
In the terminal, switch to the folder where the file is located, enter the following command:
$ sudo tar xvf cudnn-7.0-linux-x64-v4.0-prod.tgz$ CD cuda/include$ sudo cp *.h/usr/local/include/$ Cd.. /lib64$ sudo cp lib*/usr/local/lib/$ Cd/usr/local/lib$ sudo chmod +r libcudnn.so.4.0.4$ sudo ln-sf libcudnn.so.4.0.4Libcudnn.so.4$ sudo ln-sf libcudnn.so.4Libcudnn.so$ sudo ldconfig
Then switch to the Caffe root directory and Makefile.config
remove the USE_CUDNN before the line #
:
cdsudo vi Makefile.config
To recompile after saving:
sudosudosudo make test sudo
I compile the last step of the time, there are such errors:
libcudart.so.7.5openfilefileordirectory
There's a bunch of introductions on the web, how to set the environment variables (we've already set them up), so we can check if we've added environment variables:
echo$PATHecho$LD_LIBRARY_PATH
As you can see, the environment variables have been added well.
The workaround is to copy some files to the /usr/local/lib
folder:
$ sudoCP/usr/local/cuda-7.5/lib64/libcudart. so. 7. 5/usr/local/lib/libcudart. so. 7. 5&& sudo ldconfig$ sudoCP/usr/local/cuda-7.5/lib64/libcublas. so. 7. 5/usr/local/lib/libcublas. so. 7. 5&& sudo ldconfig$ sudoCP/usr/local/cuda-7.5/lib64/libcurand. so. 7. 5/usr/local/lib/libcurand. so. 7. 5&& sudo ldconfig
Again, try the sudo make runtest
command that appears as follows:
At this point, the Caffe installation is complete.
from : Personal blog site: Yhl ' s blog
Ubuntu14.04 Installing CUDA7.5 + Caffe + CuDNN