前言
CAFFE（快速特征嵌入的卷积结构，Convolutional Architecture for Fast Feature Embedding）是一个深度学习框架，最初开发于加利福尼亚大学伯克利分校。
我们按照官方的说明网页来介绍如何在Ubuntu 16.04 LTS系统上手工编译安装该软件。
安装依赖包
一般依赖项
1
2
| 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
|
特殊依赖软件
安装
下载caffe
由于我安装caffe的目的是跑SSD代码，所以选择的是weiliu89的fork分支进行安装。
1
2
3
| $ git clone https://github.com/weiliu89/caffe.git
$ cd caffe
$ git checkout ssd
|
编译
我们先尝试编译一下代码，然后在遇到问题的过程中逐步解决问题。
1
2
3
| # Modify Makefile.config according to your Caffe installation.
$ cp Makefile.config.example Makefile.config
$ make -j8
|
此时，可能会出现错误一：
.build_release/src/caffe/proto/caffe.pb.h:9:42: fatal error: google/protobuf/stubs/common.h: No such file or directory
错误一解决方案：
1
2
3
| $ sudo apt-get install libprotobuf-dev protobuf-compiler
# 继续编译
$ make -j8
|
此时，可能会出现错误二：
./include/caffe/common.hpp:5:27: fatal error: gflags/gflags.h: No such file or directory
错误二解决方案：
1
2
3
| $ sudo apt-get install libprotobuf-dev protobuf-compiler
# 继续编译
$ make -j8
|
此时，可能会出现错误三：
./include/caffe/common.hpp:6:26: fatal error: glog/logging.h: No such file or directory
错误三解决方案：
1
2
3
| $ sudo apt-get install libgoogle-glog-dev
# 继续编译
$ make -j8
|
此时，可能会出现错误四：
./include/caffe/util/device_alternate.hpp:34:23: fatal error: cublas_v2.h: No such file or directory
错误四产生原因：
没有把cuda的头文件、库的路径放置到caffe的Makefile.config中
错误四解决方案：
在Makefile.config文件中添加下列两行，具体位置可放在# If Homebrew is installed at a non standard location (for example your home directory) and you use it for general dependencies这句话下面。
1
2
| INCLUDE_DIRS += /usr/local/cuda-8.0/include
LIBRARY_DIRS += /usr/local/cuda-8.0/lib64
|
修改完后，继续编译make -j8
此时，可能会出现错误五：
./include/caffe/util/mkl_alternate.hpp:14:19: fatal error: cblas.h: No such file or directory
错误五解决方案：
1
2
3
| $ sudo apt-get install libopenblas-dev
# 继续编译
$ make -j8
|
此时，可能会出现错误六：
src/caffe/layers/hdf5_data_layer.cpp:13:18: fatal error: hdf5.h: No such file or directory
错误六解决方案：
第一步，先安装下面的依赖包
1
| $ sudo apt install libhdf5-serial-dev
|
第二步，编辑Makefile.config文件
原文件为：
1
2
3
| # Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib
|
现在修改为：
1
2
3
| # Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial/
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial/
|
第三步，编辑Makefile文件
原文件为：
1
| LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_hl hdf5
|
现修改为：
1
| LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial
|
修改完后，继续编译make -j8
此时，可能会出现错误七：
./include/caffe/util/db_leveldb.hpp:7:24: fatal error: leveldb/db.h: No such file or directory
错误七解决方案：
1
2
3
| $ sudo apt-get install libleveldb-dev
# 继续编译
$ make -j8
|
此时，可能会出现错误八：
./include/caffe/util/db_lmdb.hpp:8:18: fatal error: lmdb.h: No such file or directory
错误八解决方案：
1
2
3
| $ sudo apt-get install liblmdb-dev
# 继续编译
$ make -j8
|
此时，可能会出现错误九：
/usr/bin/ld: cannot find -lsnappy
错误九解决方案：
1
2
3
| $ sudo apt-get install libsnappy-dev
# 继续编译
$ make -j8
|
这时候，基本上就可以成功编译了。
接着，继续执行下列命令：
1
2
3
| $ make py
$ make test -j8
$ make runtest -j8
|
此时，可能会出现错误十：
.build_release/tools/caffe: error while loading shared libraries: libcudart.so.8.0: cannot open shared object file: No such file or directory
错误十解决方案：
在~/.bashrc文件中添加
1
| export LIBRARY_PATH=$LIBRARY_PATH:/usr/local/cuda-8.0/lib64
|
之后再执行make runtest -j8，就会顺利地跑测试程序，不出意外的话，可得到如下结果：
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
| $ make runtest -j8
.build_release/tools/caffe
caffe: command line brew
usage: caffe <command> <args>
commands:
train train or finetune a model
test score a model
device_query show GPU diagnostic information
time benchmark model execution time
Flags from tools/caffe.cpp:
-gpu (Optional; run in GPU mode on given device IDs separated by ','.Use
'-gpu all' to run on all available GPUs. The effective training batch
size is multiplied by the number of devices.) type: string default: ""
-iterations (The number of iterations to run.) type: int32 default: 50
-level (Optional; network level.) type: int32 default: 0
-model (The model definition protocol buffer text file.) type: string
default: ""
-phase (Optional; network phase (TRAIN or TEST). Only used for 'time'.)
type: string default: ""
-sighup_effect (Optional; action to take when a SIGHUP signal is received:
snapshot, stop or none.) type: string default: "snapshot"
-sigint_effect (Optional; action to take when a SIGINT signal is received:
snapshot, stop or none.) type: string default: "stop"
-snapshot (Optional; the snapshot solver state to resume training.)
type: string default: ""
-solver (The solver definition protocol buffer text file.) type: string
default: ""
-stage (Optional; network stages (not to be confused with phase), separated
by ','.) type: string default: ""
-weights (Optional; the pretrained weights to initialize finetuning,
separated by ','. Cannot be set simultaneously with snapshot.)
type: string default: ""
.build_release/test/test_all.testbin 0 --gtest_shuffle
Cuda number of devices: 1
Setting to use device 0
Current device id: 0
Current device name: GeForce GTX 1060 3GB
Note: Randomizing tests' orders with a seed of 17807 .
[==========] Running 2301 tests from 299 test cases.
[----------] Global test environment set-up.
[----------] 8 tests from LRNLayerTest/3, where TypeParam = caffe::GPUDevice<double>
[ RUN ] LRNLayerTest/3.TestSetupWithinChannel
[ OK ] LRNLayerTest/3.TestSetupWithinChannel (1471 ms)
[ RUN ] LRNLayerTest/3.TestSetupAcrossChannels
...
[----------] Global test environment tear-down
[==========] 2301 tests from 299 test cases ran. (372301 ms total)
[ PASSED ] 2301 tests.
|
最后的PASSED表示全部测试都通过了，说明安装成功。
参考资料
修订历史
