Deep learning with Caffe: installation

• Download the caffe library.
• Install the prerequisites:
• CUDA 1.55
• BLAS (I installed the atlas)
• sudo apt-get install libatlas-base-dev
• OpenCV
• Boost
• glog, gflags, protobuf, leveldb, snappy, hdf5

• sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev 

  libboost-all-dev libhdf5-serial-dev
  

   
• numpy (>=1.7)
• aptitude search numpy
• sudo aptitude install python-numpy
• (Install aptitude if you don't have)
• google logging library

• wget https://google-glog.googlecode.com/files/glog-0.3.3.tar.gz
  tar zxvf glog-0.3.3.tar.gz
  ./configure
  make && make install 

• Compile
•  cp Makefile.config.example Makefile.config 
• # Adjust Makefile.config (for example, if using Anaconda Python) 
• make all 
• make test 
• make runtest
•  If there is no gpu then gputest will fail. I got the following core dump since I don't have the gpu 
• 
•  Without the CUDA enabled compilation you will be able to see the 'runtest' output as follows:
• 
•