下载好相应的文件,一步步的执行如下命令就行:

### python2.7-opencv-2.4.11-pytorch-0.2.0-tensorflow-gpu-base-1.3.0
0. chmod +x Anaconda2-5.0.0-Linux-x86_64.sh
1. ./Anaconda2-5.0.0-Linux-x86_64.sh
2. conda install opencv-2.4.11-nppy27_0.tar.bz2
3. pip install torch-0.2.0.post3-cp27-cp27mu-manylinux1_x86_64.whl
4. pip install torchvision-0.1.8-py2.py3-none-any.whl
5. conda install tensorflow-gpu-base-1.3.0-py27cuda8.0cudnn6.0_0.tar.bz2
6. conda install protobuf-3.0.0-py27_1.tar.bz2
7. conda install backports.weakref-1.0rc1-py27_0.tar.bz2
8. conda install mock-2.0.0-py27_0.tar.bz2
9. conda install pbr-1.8.0-py27_0.tar.bz2
10.  conda install tensorflow-tensorboard-0.1.5-py27_0.tar.bz2

实验室的集群式服务器使用的是CentOS然后每个人一个单独的账户,都不是管理员用户,然后因为校园网的原因,非管理员用户联网还存在问题,所以研究了很长的时间,最后发现anaconda是个神器,还是用这个东西爽!!!!!!