楼主你好，载入模型出错，提示不支持，然后app直接挂掉
cxy200927099 opened this issue · comments
这是我app运行的log，不太明白是什么意思，忘楼主指导指导
12-01 11:41:05.717 6028-6058/com.yangwenbo.caffesimple D/mali_winsys: EGLint new_window_surface(egl_winsys_display*, void*, EGLSurface, EGLConfig, egl_winsys_surface**, egl_color_buffer_format*, EGLBoolean) returns 0x3000
12-01 11:41:08.875 6028-6028/com.yangwenbo.caffesimple I/PlaceActivity: init caffe.....
12-01 11:41:08.903 6028-6028/com.yangwenbo.caffesimple I/PlaceActivity: onCreate: modelFile:/storage/emulated/0/caffe/shufflenet_deploy_80.protobin
12-01 11:41:08.903 6028-6028/com.yangwenbo.caffesimple I/PlaceActivity: onCreate: weightFile:/storage/emulated/0/caffe/shuffleNet_80_4_iter_12000.caffemodel
12-01 11:41:08.911 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/util/upgrade_proto.cpp:67] Attempting to upgrade input file specified using deprecated input fields: /storage/emulated/0/caffe/shufflenet_deploy_80.protobin
12-01 11:41:08.912 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/util/upgrade_proto.cpp:70] Successfully upgraded file specified using deprecated input fields.
12-01 11:41:08.912 6028-6028/com.yangwenbo.caffesimple I/caffe: WARNING /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/util/upgrade_proto.cpp:72] Note that future Caffe releases will only support input layers and not input fields.
12-01 11:41:08.912 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/util/upgrade_proto.cpp:77] Attempting to upgrade batch norm layers using deprecated params: /storage/emulated/0/caffe/shufflenet_deploy_80.protobin
12-01 11:41:08.912 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/util/upgrade_proto.cpp:80] Successfully upgraded batch norm layers using deprecated params.
12-01 11:41:08.916 6028-6040/com.yangwenbo.caffesimple W/art: Suspending all threads took: 7.877ms
12-01 11:41:08.927 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer input
12-01 11:41:08.928 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] input -> data
12-01 11:41:08.928 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer conv1
12-01 11:41:08.928 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] conv1 -> conv1
12-01 11:41:08.930 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer conv1_bn
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer conv1_scale
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer conv1_scale
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer conv1_relu
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer pool1
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] pool1 -> pool1
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer pool1_pool1_0_split
12-01 11:41:08.931 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] pool1_pool1_0_split -> pool1_pool1_0_split_0
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] pool1_pool1_0_split -> pool1_pool1_0_split_1
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_match_conv
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] resx1_match_conv -> resx1_match_conv
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv1
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/src/caffe/net.cpp:394] resx1_conv1 -> resx1_conv1
12-01 11:41:08.932 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv1_bn
12-01 11:41:08.933 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv1_scale
12-01 11:41:08.933 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv1_scale
12-01 11:41:08.933 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv1_relu
12-01 11:41:08.933 6028-6028/com.yangwenbo.caffesimple I/caffe: INFO /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:77] Creating layer resx1_conv2
12-01 11:41:08.933 6028-6028/com.yangwenbo.caffesimple I/caffe: ERROR /Users/chenxingyi/work/machineLearning/code/caffe-mobile/include/caffe/layer_factory.hpp:81] (registry.count(type)) == (1) Unknown layer type: ConvolutionDepthwise (known types: AbsVal, ArgMax, BNLL, BatchNorm, BatchReindex, Bias, Concat, Convolution, Crop, Deconvolution, Dropout, DummyData, ELU, Eltwise, Embed, Exp, Filter, Flatten, Im2col, InnerProduct, Input, LRN, LSTM, Log, MVN, MemoryData, PReLU, Parameter, Pooling, Power, RNN, ReLU, Reduction, Reshape, SPP, Scale, Sigmoid, Silence, Slice, Softmax, Split, TanH, Threshold, Tile)
--------- beginning of crash
12-01 11:41:08.934 6028-6028/com.yangwenbo.caffesimple A/libc: Fatal signal 11 (SIGSEGV), code 1, fault addr 0x0 in tid 6028 (nbo.caffesimple)
[ 12-01 11:41:08.935 4064: 4064 W/ ]
debuggerd: handling request: pid=6028 uid=10044 gid=10044 tid=6028
@cxy200927099 ConvolutionDepthwise 不在 Caffe 源码里，应该是你自定义的层吧。你需要把这个层相关的代码和参数 proto 加到 caffe-mobile 里。
嗯，加入了，可以编过了，可以关闭了