真相是真--- ubuntu18 安装caffe

编译安装caffe太太太痛苦

ubuntu18Xcaffe用心良苦,为我们社畜遮风挡雨

sudo apt install caffe-cuda

就完事了,caffe指令已经自动加到系统环境变量

下载

git clone git://github.com/BVLC/caffe.git

 

测试

运行example（后两个sh文件得改一下，caffe所有指令的路径删了）

sudo sh data/mnist/get_mnist.sh
sudo ./examples/mnist/create_mnist.sh
sudo ./examples/mnist/train_lenet.sh

输出：

I0503 17:42:58.220419 16278 caffe.cpp:218] Using GPUs 0
I0503 17:42:58.258078 16278 caffe.cpp:223] GPU 0: GeForce GTX 1080 Ti
I0503 17:42:58.428989 16278 solver.cpp:44] Initializing solver from parameters: 
test_iter: 100
test_interval: 500
base_lr: 0.01
display: 100
max_iter: 10000

。。。。。
I0503 17:43:15.374331 16278 sgd_solver.cpp:105] Iteration 2800, lr = 0.00830984
I0503 17:43:15.419070 16287 data_layer.cpp:73] Restarting data prefetching from start.
I0503 17:43:15.912174 16278 solver.cpp:218] Iteration 2900 (185.935 iter/s, 0.537823s/100 iters), loss = 0.0165816
I0503 17:43:15.912200 16278 solver.cpp:237]     Train net output #0: loss = 0.0165815 (* 1 = 0.0165815 loss) 
。。。。。