Caffe是一個清晰而高效的深度學習框架，其作者是博士畢業於UC Berkeley的賈揚清。Caffe是純粹的C++/CUDA架構，支持命令行、Python和MATLAB接口；可以在CPU和GPU直接無縫切換。我在MacbookPro（無NVIDIA顯卡）上大費周章地配置了Caffe的環境，並花了許多時間配置其python接口。
一、下載Caffe
github上的下載地址：https://github.com/BVLC/caffe
進入到下載后的路徑，並復制 Makefile.config.example 重命名為 Makefile.config (我電腦的用戶名是cuiqi，注意修改)
git clone https://github.com/BVLC/caffe.git cd /Users/cuiqi/Downloads/caffe-master && cp Makefile.config.example Makefile.config
二、安裝相關依賴
對於需要python接口的情況，需要以下依賴
1. CUDA
由於我的Mac沒有NVIDIA的GPU，所以只能使用CPU_ONLY模式，需要在 Makefile.config 中修改 CPU_ONLY := 1
2. BLAS via ATLAS, MKL, or OpenBLAS. # Basic Linear Algebra Subprograms，基礎線性代數程序集 3. Boost >= 1.55 # Deepdream是用Python接Caffe，因此還需要 boost.python 支持
brew install boost --with-python
brew install boost-python
4. OpenCV >= 2.4 including 3.0 5. protobuf, glog, gflagsbrew install protobuf
brew install glog
brew install gflags
6. IO libraries hdf5, leveldb, snappy, lmdb
brew install leveldb
brew install lmdb
brew tap homebrew/science
brew install homebrew/science/hdf5
# python driver for hdf5
pip install h5py
7. numpy for python
brew install numpy
三、修改Makefile.config中相應的路徑
如果要使用Anaconda的python環境可以在Makefile.config中取消相應的注釋，我試過這樣做，可是在CMAKE的時候並不奏效：
Python:
Interpreter : /usr/bin/python2.7 (ver. 2.7.10)
Libraries : /usr/lib/libpython2.7.dylib (ver 2.7.10)
NumPy : /System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/core/include (ver 1.8.0rc1)
所以我放棄了使用Anaconda，改使用系統自帶的python。
PYTHON_INCLUDE := /usr/include/python2.7
由於我使用了Homebrew安裝了numpy，所以我在makefile.config中修改相應的numpy路徑
PYTHON_INCLUDE := /usr/include/python2.7 \
/usr/local/Cellar/numpy/1.11.2/lib/python2.7/site-packages/numpy/core/include/numpy/core/include
我的Makefile.config文件：
## Refer to http://caffe.berkeleyvision.org/installation.html # Contributions simplifying and improving our build system are welcome! # cuDNN acceleration switch (uncomment to build with cuDNN). # USE_CUDNN := 1 # CPU-only switch (uncomment to build without GPU support). CPU_ONLY := 1 # uncomment to disable IO dependencies and corresponding data layers # USE_OPENCV := 0 # USE_LEVELDB := 0 # USE_LMDB := 0 # uncomment to allow MDB_NOLOCK when reading LMDB files (only if necessary) # You should not set this flag if you will be reading LMDBs with any # possibility of simultaneous read and write # ALLOW_LMDB_NOLOCK := 1 # Uncomment if you're using OpenCV 3 # OPENCV_VERSION := 3 # To customize your choice of compiler, uncomment and set the following. # N.B. the default for Linux is g++ and the default for OSX is clang++ # CUSTOM_CXX := g++ # CUDA directory contains bin/ and lib/ directories that we need. CUDA_DIR := /usr/local/cuda # On Ubuntu 14.04, if cuda tools are installed via # "sudo apt-get install nvidia-cuda-toolkit" then use this instead: # CUDA_DIR := /usr # CUDA architecture setting: going with all of them. # For CUDA < 6.0, comment the *_50 lines for compatibility. CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \ -gencode arch=compute_20,code=sm_21 \ -gencode arch=compute_30,code=sm_30 \ -gencode arch=compute_35,code=sm_35 \ -gencode arch=compute_50,code=sm_50 \ -gencode arch=compute_50,code=compute_50 # BLAS choice: # atlas for ATLAS (default) # mkl for MKL # open for OpenBlas BLAS := atlas # Custom (MKL/ATLAS/OpenBLAS) include and lib directories. # Leave commented to accept the defaults for your choice of BLAS # (which should work)! # BLAS_INCLUDE := /path/to/your/blas # BLAS_LIB := /path/to/your/blas # Homebrew puts openblas in a directory that is not on the standard search path # BLAS_INCLUDE := $(shell brew --prefix openblas)/include # BLAS_LIB := $(shell brew --prefix openblas)/lib # This is required only if you will compile the matlab interface. # MATLAB directory should contain the mex binary in /bin. # MATLAB_DIR := /usr/local # MATLAB_DIR := /Applications/MATLAB_R2012b.app # NOTE: this is required only if you will compile the python interface. # We need to be able to find Python.h and numpy/arrayobject.h. PYTHON_INCLUDE := /usr/include/python2.7 \ /usr/local/Cellar/numpy/1.11.2/lib/python2.7/site-packages/numpy/core/include/numpy/core/include # Anaconda Python distribution is quite popular. Include path: # Verify anaconda location, sometimes it's in root. # ANACONDA_HOME := $(HOME)/anaconda # PYTHON_INCLUDE := $(ANACONDA_HOME)/include \ # $(ANACONDA_HOME)/include/python2.7 \ # $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include \ # Uncomment to use Python 3 (default is Python 2) # PYTHON_LIBRARIES := boost_python3 python3.5m # PYTHON_INCLUDE := /usr/include/python3.5m \ # /usr/lib/python3.5/dist-packages/numpy/core/include # We need to be able to find libpythonX.X.so or .dylib. PYTHON_LIB := /usr/lib # PYTHON_LIB := $(ANACONDA_HOME)/lib # Homebrew installs numpy in a non standard path (keg only) # PYTHON_INCLUDE += $(dir $(shell python -c 'import numpy.core; print(numpy.core.__file__)'))/include # PYTHON_LIB += $(shell brew --prefix numpy)/lib # Uncomment to support layers written in Python (will link against Python libs) # WITH_PYTHON_LAYER := 1 # Whatever else you find you need goes here. INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib # If Homebrew is installed at a non standard location (for example your home directory) and you use it for general dependencies # INCLUDE_DIRS += $(shell brew --prefix)/include # LIBRARY_DIRS += $(shell brew --prefix)/lib # Uncomment to use `pkg-config` to specify OpenCV library paths. # (Usually not necessary -- OpenCV libraries are normally installed in one of the above $LIBRARY_DIRS.) # USE_PKG_CONFIG := 1 # N.B. both build and distribute dirs are cleared on `make clean` BUILD_DIR := build DISTRIBUTE_DIR := distribute # Uncomment for debugging. Does not work on OSX due to https://github.com/BVLC/caffe/issues/171 # DEBUG := 1 # The ID of the GPU that 'make runtest' will use to run unit tests. TEST_GPUID := 0 # enable pretty build (comment to see full commands) Q ?= @
四、編譯
使用cmake工具，將cmake的sourcecode設為caffe的路徑，在caffe目錄下建立build文件夾，並將cmake的build路徑設為該build的路徑，configure，generate。
此時打開終端工具進入build目錄下，make。
make all
make test
make runtest
make pycaffe
如果沒有出現錯誤，並全部通過后，此時make告一段落，但是並不能確定可以在python中import。
在終端上：
vi etc/profile
或
sudo vi /etc/profile
進入修改系統的環境變量，將 caffe/python/添加到Python系統路徑：
export PYTHONPATH=path to caffe/python:$PYTHONPATH
例如：
export PYTHONPATH=/Users/cuiqi/Downloads/caffe-master/python:$PYTHONPATH
終端輸入python，確認這個python環境是剛剛在makefile.config中設定的那個，可以 which python 確認。由於我沒有使用Anaconda，我的是
/usr/local/bin/python ，確認為剛剛在makefile.config中設定的路徑。
>>> import caffe
# 可能的錯誤
ImportError: No module named skimage.io
# 解決
pip install scikit-image
參考：https://github.com/rainyear/lolita/issues/10?utm_source=tuicool&utm_medium=referral
16.12.22更新：
還可能出現的錯誤：segment fault：11
在 ITNOSE 上一作者認為此錯誤很有可能是系統里也有多個版本的Python，並且編譯和運行的時候不是一個版本。
解決辦法：
當然就是要統一編譯與運行的python版本。終端鍵入 which python ，反饋給我/usr/local/bin/python，而CMAKE工具反饋給我的是系統自帶的python路徑。
1.對於我而言，我准備使用系統自帶的python。使用Homebrew工具brew install python后，python環境路徑為/usr/local/bin/python，而caffe編譯時使用了系統自帶的python路徑 /System/Library/Frameworks/Python.framework/Versions/2.7/bin/python。這個路徑可以使用otool反編譯工具檢出，例如：otool -L python/caffe/_caffe.so （_caffe.so在下載好的caffe的python/caffe下）。
2.修改bash_profile配置文件以在啟動系統時使用自帶python，打開終端，鍵入 vi ~/.bash_profile ，輸入：
# Setting PATH for Python 2.7
# The orginal version is saved in .bash_profile.pysave
PATH="/System/Library/Frameworks/Python.framework/Versions/2.7/bin:${PATH}"
export PATH
完成后輸入：wq（保存退出），終端鍵入 source ~/.bash_profile 或者注銷及重啟使修改生效。
3.再次終端鍵入 which python，返回 /System/Library/Frameworks/Python.framework/Versions/2.7/bin/python，說明修改生效。鍵入python，
>>> import caffe
無報錯，成功。
參考：
http://www.itnose.net/detail/6522348.html
https://www.zhihu.com/question/30941329
16.12.29更新：
可能出現的錯誤：TypeError: __init__() got an unexpected keyword argument 'syntax'；
參考dupuleng的解決方法，這是由於protobuf版本的問題造成的。
查看protobuf版本：
pip show protobuf
protobuf信息：
Name: protobuf
Version: 2.6.1
Summary: Protocol Buffers
Home-page: https://developers.google.com/protocol-buffers/
Author: protobuf@googlegroups.com
Author-email: protobuf@googlegroups.com
License: New BSD License
Location: /Users/cuiqi/anaconda/lib/python2.7/site-packages
Requires: setuptools
此版本（2.6.1）需要變更：
pip uninstall protobuf
pip install 'protobuf>=3.0.0a3'
pip show numpy protobuf
protobuf信息：
Name: protobuf
Version: 3.1.0.post1
Summary: Protocol Buffers
Home-page: https://developers.google.com/protocol-buffers/
Author: protobuf@googlegroups.com
Author-email: protobuf@googlegroups.com
License: New BSD License
Location: /Users/cuiqi/anaconda/lib/python2.7/site-packages
Requires: six, setuptools
完成
關於cmake：
經過我的測試發現修改makefile.config后，cmake的 configure之前還要手動修改python的相關路徑，否則cmake會使用mac的系統python環境，造成上述segment fault 11的情況。關鍵在於要勾選Advanced選項：
不勾選Advanced：
更改好之后再configure，generate，cmake就會選擇相應的環境路徑了。