安装Caffe 1.首先安装Ubuntu16.04。 2.完成后,打开命令行控制台,输入以下命令,安装基本依赖(General dependencies)。 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 注意:建议一次安装一个软件包,便于弄清是那个未安装成功,以节省时间。 3.安装ATLAS,输入下述命令: sudo apt-get install libatlas-base-dev 4.安装剩余依赖: sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev 5.下载Caffe: sudo apt-get install git git clone https://github.com/BVLC/caffe.git 6.修改Makefile.config: cp Makefile.config.example Makefile.config gedit Makefile.config 找到#CPU_ONLY := 1,取消注释(我们设置为CPU模式) 找到 # Whatever else you find you need goes here. INCLUDE_DIRS := (PYTHON_INCLUDE) /usr/local/include LIBRARY_DIRS := (PYTHON_LIB) /usr/local/lib /usr/lib 修改为 # 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/x86_64-linux-gnu/hdf5/serial 7.进行编译 make all make test make runtest 8.耐心等待编译完成
测试MNIST 1. ~/caffe/data/mnist下执行 ./get_mnist.sh ,下载并解压4个文件。 2. 执行examples/mnist/create_mnist.sh 脚本,转换数据格式,将mnist data转化为可用的lmdb格式的文件:~/caffe/examples/mnist目录下多出mnist_test_lmdb和mnist_train_lmdb。 3. 使用~/caffe/examples/mnist中的lenet,参数为 lenet_train_test.prototxt,执行./train_lenet.sh 训练。发现报错;“Using GPUs 0 Cannot use GPU in CPU-only Caffe: check mode. ”, 修改 caffe根目录下CMakeLists.txt: caffe_option(CPU_ONLY “Build Caffe without CUDA support” ON) 修改/caffe/examples/mnist/lenet_solver.prototxt中 solver_mode: CPU 4.返回caffe根目录,执行./examples/mnist/train_lenet.sh 这里要迭代10000次,需要等一会。 5.训练结束后,caffe根目录下执行:./build/tools/caffe.bin test -model=examples/mnist/lenet_train_test.prototxt -weights=examples/mnist/lenet_iter_10000.caffemodel 可以看到精度高于0.98
参考:http://blog.csdn.net/forest_world/article/details/51376554
新闻热点
疑难解答