近期由于复现论文实验,在服务器上安装caffe框架,首先根据
http://caffe.berkeleyvision.org/installation.html
网址来了解caffe编译所选要的前提环境:
Caffe has several dependencies:
- CUDA is required for GPU mode.
- library version 7+ and the latest driver version are recommended, but 6.* is fine too
- 5.5, and 5.0 are compatible but considered legacy
- BLAS via ATLAS, MKL, or OpenBLAS.
- Boost >= 1.55
protobuf
,glog
,gflags
,hdf5
Optional dependencies:
- OpenCV >= 2.4 including 3.0
- IO libraries:
lmdb
,leveldb
(note: leveldb requiressnappy
) - cuDNN for GPU acceleration (v6)
Pycaffe and Matcaffe interfaces have their own natural needs.
- For Python Caffe:
Python 2.7
orPython 3.3+
,numpy (>= 1.7)
, boost-providedboost.python
- For MATLAB Caffe: MATLAB with the
mex
compiler.
第一步就是安装必须依赖,
在安装之前服务器上已经有一些依赖。如CUDA8.0,opencv3.2.cuDNN.
所以我安装BLAS,基本线性代数子程序(BLAS)是一个规范,它规定了一组用于执行常见线性代数运算的低级例程,
这里我选择了ATLAS,
https://blog.youkuaiyun.com/weixin_28949825/article/details/79426732
这个教程里有blas、lapack、atlas在Ubuntu上的安装。
然后安装Boost,Boost provides free peer-reviewed portable C++ source libraries.
https://blog.youkuaiyun.com/a379039233/article/details/80493167
先下载,再编译。
最后在python环境中装protobuf
, glog
, gflags
, hdf5
pip3 install protobuf
pip install glog
安装gflags
https://blog.youkuaiyun.com/weixin_44354586/article/details/89399763
安装hdf5 :HDF5是一个独特的技术套件,可以管理非常大和复杂的数据收集。
下载 hdf5-1.12 ,其他版本链接里自己查找
https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.12/hdf5-1.12.0/src/
安装:https://www.jianshu.com/p/eb93fa5748f0