编译caffe on Ubuntu16.04

1. 确保系统内已经存在Cuda8.0

方法：在、usr/local/cuda/下找到sample，进入目录，键入

make all

编译完成后，找到device_query这个执行文件并运行。如果Cuda安装成功，会显示GPU相关信息

2. 安装依赖包

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev

3. 安装openblas

sudo apt-get install openblas-dev

4. 安装opencv

sudo apt-get install libopencv-dev

5. 下载caffe

git clone https://github.com/BVLC/caffe

6. 修改Makefile.config文件

USE_CUDNN := 1
BLAS := open
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/hdf5/serial /usr/lib/x86_64-linux-gnu/hdf5/serial

7. 编译caffe

make all -j?
make runtest -j?

8. 安装python依赖库

for req in $(cat requirements.txt); do sudo pip install $req; done

9. 编译python接口

sudo vim ~/.bashrc

添加

export PYTHONPATH=/usr/caffe/python:$PYTHONPATH

然后

source ~/.bashrc

最后

make pycaffe -j?

试试看

python
>>>import caffe