1. 安装.deb文件，用sudo gdebi XXX.deb
sudo apt-get install xxx
2. 需要配置系统路径：
LD_LIBRARY_PATH=.../lib:LD_LIBRARY_PATH
export LD_LIBRARY_PATH
3. ln -sf xxx xxx 软链接
4. Training LeNet on MNIST with Caffe：
wangxiao@wangxiao-Aspire-VN7-591G:~/Downloads/caffe-master/examples/mnist$ cd
wangxiao@wangxiao-Aspire-VN7-591G:~$ sh /home/wangxiao/Downloads/caffe-master/examples/mnist/create_mnist.sh
Creating lmdb...
/home/wangxiao/Downloads/caffe-master/examples/mnist/create_mnist.sh: 16: /home/wangxiao/Downloads/caffe-master/examples/mnist/create_mnist.sh: build/examples/mnist/convert_mnist_data.bin: not found
/home/wangxiao/Downloads/caffe-master/examples/mnist/create_mnist.sh: 18: /home/wangxiao/Downloads/caffe-master/examples/mnist/create_mnist.sh: build/examples/mnist/convert_mnist_data.bin: not found
Done.
然而，convert_mnist_data.bin 就在那里哇。。。
于是乎：百度一下： 链接：http://blog.sciencenet.cn/blog-1583812-843207.html
然后就成功了 。。。
5. 2015-10-13 12:39:29 (22.2 KB/s) - ‘caffe_ilsvrc12.tar.gz.1’ saved [17858008/17858008]
Unzipping...
gzip: stdin: unexpected end of file
tar: Unexpected EOF in archive
tar: Unexpected EOF in archive
tar: Error is not recoverable: exiting now
Done.
6. Linux怎么安装.tar.gz软件
tar.gz 是用tar命令将源码文件打包成一个文件，然后用zip命令压缩成一个压缩包，所以在安装前要解压缩，命令如下：tar -xvzf test.tar.gz完成后会生成一个解压缩后的目录，进入该目录，查看readme文件或者install文件，里面有详细的安装命令。一般的安装步骤是./configure 回车make 回车make install 回车make clean 回车如果configure没有通过，说明你的系统环境没有达到要求，该软件所依赖的文件没有被安装，需要根据提示信息来具体分析。以上的各个命令只是最常用的使用方式，还有比较实用的参数没有介绍，可以自己查看帮助文档。
7. 在cifa10这个数据集上又出现了类似的问题：
wangxiao@wangxiao-Aspire-VN7-591G:~/Downloads/caffe-master/examples/cifar10$ sh create_cifar10.sh
Creating lmdb...
create_cifar10.sh: 12: create_cifar10.sh: ./build/examples/cifar10/convert_cifar_data.bin: not found
Computing image mean...
create_cifar10.sh: 16: create_cifar10.sh: ./build/tools/compute_image_mean: not found
Done.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
换了一种输入方式，可以了，然而出现了另一种错误：
wangxiao@wangxiao-Aspire-VN7-591G:~/Downloads/caffe-master$ sh ./examples/cifar10/create_cifar10.sh
Creating lmdb...
F1013 15:08:01.317740 12620 convert_cifar_data.cpp:58] Check failed: data_file Unable to open train file #2
*** Check failure stack trace: ***
@ 0x7fb181f96daa (unknown)
@ 0x7fb181f96ce4 (unknown)
@ 0x7fb181f966e6 (unknown)
@ 0x7fb181f99687 (unknown)
@ 0x402fc7 convert_dataset()
@ 0x401b8f main
@ 0x7fb1813c6ec5 (unknown)
@ 0x401c49 (unknown)
@ (nil) (unknown)
Aborted (core dumped)
Computing image mean...
terminate called after throwing an instance of ‘std::bad_alloc‘
what(): std::bad_alloc
Aborted (core dumped)
Done.
wangxiao@wangxiao-Aspire-VN7-591G:~$ cd /home/wangxiao/Downloads/caffe-master/
wangxiao@wangxiao-Aspire-VN7-591G:~/Downloads/caffe-master$ sh ./examples/cifar10/create_cifar10.sh
Creating lmdb...
./examples/cifar10/create_cifar10.sh: 12: ./examples/cifar10/create_cifar10.sh: ./build/examples/cifar10/convert_cifar_data.bin: not found
Computing image mean...
./examples/cifar10/create_cifar10.sh: 16: ./examples/cifar10/create_cifar10.sh:
./build/tools/compute_image_mean: not found
Done.
继续Baidu啊，搜到了，然并卵。。。
原文：http://www.cnblogs.com/wangxiaocvpr/p/4875618.html