先下载了老师百度网盘的课件中的caffe-ssd的代码，编译安装好了ssd的pycaffe。
按照老师教的打包好了lmdb数据。
没有自己修改 examples/ssd/wider_face/ssd_pascal.py，因为看视频发现ssd_pascal.py需要修改的地方都已经改好了，已经可以直接运行了。
于是直接在caffe-ssd目录下面运行：
[root@localhost caffe-ssd]# python examples/ssd/wider_face/ssd_pascal.py
结果：

I0616 11:57:13.472048 14297 net.cpp:228] conv6_1_relu does not need backward computation.
I0616 11:57:13.472051 14297 net.cpp:228] conv6_1 does not need backward computation.
I0616 11:57:13.472055 14297 net.cpp:228] conv5_3_relu5_3_0_split does not need backward computation.
I0616 11:57:13.472059 14297 net.cpp:228] relu5_3 does not need backward computation.
I0616 11:57:13.472062 14297 net.cpp:228] conv5_3 does not need backward computation.
I0616 11:57:13.472067 14297 net.cpp:228] relu5_1 does not need backward computation.
I0616 11:57:13.472070 14297 net.cpp:228] conv5_1 does not need backward computation.
I0616 11:57:13.472074 14297 net.cpp:228] pool4 does not need backward computation.
I0616 11:57:13.472079 14297 net.cpp:228] conv4_3_relu4_3_0_split does not need backward computation.
I0616 11:57:13.472082 14297 net.cpp:228] relu4_3 does not need backward computation.
I0616 11:57:13.472086 14297 net.cpp:228] conv4_3 does not need backward computation.
I0616 11:57:13.472090 14297 net.cpp:228] relu4_2 does not need backward computation.
I0616 11:57:13.472093 14297 net.cpp:228] conv4_1 does not need backward computation.
I0616 11:57:13.472097 14297 net.cpp:228] pool3 does not need backward computation.
I0616 11:57:13.472101 14297 net.cpp:228] relu3_3 does not need backward computation.
I0616 11:57:13.472105 14297 net.cpp:228] conv3_3 does not need backward computation.
I0616 11:57:13.472108 14297 net.cpp:228] relu3_1 does not need backward computation.
I0616 11:57:13.472112 14297 net.cpp:228] conv3_1 does not need backward computation.
I0616 11:57:13.472115 14297 net.cpp:228] pool2 does not need backward computation.
I0616 11:57:13.472120 14297 net.cpp:228] relu2_2 does not need backward computation.
I0616 11:57:13.472123 14297 net.cpp:228] conv2_2 does not need backward computation.
I0616 11:57:13.472126 14297 net.cpp:228] relu2_1 does not need backward computation.
I0616 11:57:13.472131 14297 net.cpp:228] conv2_1 does not need backward computation.
I0616 11:57:13.472134 14297 net.cpp:228] pool1 does not need backward computation.
I0616 11:57:13.472137 14297 net.cpp:228] relu1_2 does not need backward computation.
I0616 11:57:13.472141 14297 net.cpp:228] conv1_2 does not need backward computation.
I0616 11:57:13.472144 14297 net.cpp:228] relu1_1 does not need backward computation.
I0616 11:57:13.472148 14297 net.cpp:228] conv1_1 does not need backward computation.
I0616 11:57:13.472152 14297 net.cpp:228] data_data_0_split does not need backward computation.
I0616 11:57:13.472157 14297 net.cpp:228] data does not need backward computation.
I0616 11:57:13.472159 14297 net.cpp:270] This network produces output detection_eval
I0616 11:57:13.472240 14297 net.cpp:283] Network initialization done.
I0616 11:57:13.472580 14297 solver.cpp:75] Solver scaffolding done.
I0616 11:57:13.475674 14297 caffe.cpp:241] Resuming from models/VGGNet/wider_face/SSD_tiny_300x300/VGG_tiny_wider_face_SSD_tiny_300x300_iter_27.solverstate
I0616 11:57:13.540910 14297 sgd_solver.cpp:356] SGDSolver: restoring history
I0616 11:57:13.566229 14297 caffe.cpp:251] Starting Optimization
I0616 11:57:13.566277 14297 solver.cpp:294] Solving VGG_tiny_wider_face_SSD_tiny_300x300_train
I0616 11:57:13.566282 14297 solver.cpp:295] Learning Rate Policy: multistep
I0616 11:57:13.569295 14297 blocking_queue.cpp:50] Data layer prefetch queue empty

卡在这里一动不动。不知道为什么
想请教老师怎么寻找原因呀？