问题：
caffe-windows-ms\NugetPackages\boost.1.59.0.0\lib\native\include\boost/python/detail/wrap_python.hpp(50): fatal error C1083: 无法打开包括文件: “pyconfig.h”: No such file or directory (编译源文件 ..\..\src\caffe\layer_factory.cpp)
解决：
caffe-windows-ms\windows/CommonSettings.props
<PythonSupport>true</PythonSupport> 改为。
<PythonSupport>false</PythonSupport>
1>------ 已启动全部重新生成: 项目: libcaffe, 配置: Release x64 ------
1> ProtoCompile.cmd : Create proto temp directory "caffe-windows-ms\windows\..\src\caffe\proto\temp"1> ProtoCompile.cmd : Generating "caffe-windows-ms\windows\..\src\caffe\proto\temp\caffe.pb.h" and "caffe-windows-ms\windows\..\src\caffe\proto\temp\caffe.pb.cc"
1> 系统找不到指定的路径。
1> ProtoCompile.cmd : Create proto include directory
1> 子目录或文件 caffe-windows-ms\windows\..\include\caffe\proto 已经存在。
1> ProtoCompile.cmd : Compare newly compiled caffe.pb.h with existing one
1> FC: 无法打开 CAFFE-WINDOWS-MS\WINDOWS\..\SRC\CAFFE\PROTO\TEMP\CAFFE.PB.H - 找不到此类文件或文件夹
1>
1> ProtoCompile.cmd : Move newly generated caffe.pb.h to "caffe-windows-ms\windows\..\include\caffe\proto\caffe.pb.h"
1> ProtoCompile.cmd : and caffe.pb.cc to "caffe-windows-ms\windows\..\src\caffe\proto\caffe.pb.cc"
1> 系统找不到指定的文件。
1> 系统找不到指定的文件。
1> blob.cpp
1> common.cpp
1> data_transformer.cpp
1> internal_thread.cpp
1> layer.cpp
1> absval_layer.cpp
1> accuracy_layer.cpp
1> arccos_layer.cpp
1> argmax_layer.cpp
1> axpy_layer.cpp
1> base_conv_layer.cpp
1> base_data_layer.cpp
1> batch_contrastive_loss_layer.cpp
1> batch_norm_layer.cpp
1> batch_reindex_layer.cpp
1> bias_layer.cpp
1> bnll_layer.cpp
1> bn_layer.cpp
1> center_loss_layer.cpp
1> channel_scale_layer.cpp
1> concat_layer.cpp
1> confidence_penalty_loss_layer.cpp
1> contrastive_loss_layer.cpp
1> conv_layer.cpp
1> correlation_loss_layer.cpp
1> cosine_layer.cpp
1> crop_layer.cpp
1> cudnn_bn_layer.cpp
1> cudnn_conv_layer.cpp
1> cudnn_lcn_layer.cpp
1> cudnn_lrn_layer.cpp
1> cudnn_pooling_layer.cpp
1> cudnn_relu_layer.cpp
1> cudnn_sigmoid_layer.cpp
1> cudnn_softmax_layer.cpp
1> cudnn_tanh_layer.cpp
1> data_layer.cpp
1> deconv_layer.cpp
1> dropout_layer.cpp
1> dummy_data_layer.cpp
1> eltwise_layer.cpp
1> elu_layer.cpp
1> embed_layer.cpp
1> euclidean_loss_layer.cpp
1> exp_layer.cpp
1> fcn_accuracy_layer.cpp
1> filter_layer.cpp
1> flatten_layer.cpp
1> flip_layer.cpp
1> focal_loss_layer.cpp
1> general_constrastive_loss_layer.cpp
1> general_triplet_loss_layer.cpp
1> global_lse_margin_layer.cpp
1> gram_layer.cpp
1> hdf5_data_layer.cpp
1> hdf5_output_layer.cpp
1> hinge_loss_layer.cpp
1> hotspot_layer.cpp
1> im2col_layer.cpp
1> image_data_layer.cpp
1> infogain_loss_layer.cpp
1> inner_distance_layer.cpp
1> inner_product_layer.cpp
1> input_layer.cpp
1> insanity_layer.cpp
1> label_specific_add.cpp
1> largemargin_inner_product_layer.cpp
1> local_layer.cpp
1> log_layer.cpp
1> loss_layer.cpp
1> lrn_layer.cpp
1> lstm_layer.cpp
1> lstm_unit_layer.cpp
1> margin_inner_product_layer.cpp
1> memory_data_layer.cpp
1> moving_normalize_layer.cpp
1> multinomial_logistic_loss_layer.cpp
1> multi_label_image_data_layer.cpp
1> mvn_layer.cpp
1> nca_loss_layer.cpp
1> neuron_layer.cpp
1> noise_layer.cpp
1> normalize_layer.cpp
1> pairwise_layer.cpp
1> parameter_layer.cpp
1> permute_layer.cpp
1> pooling_layer.cpp
1> power_layer.cpp
1> predict_box_layer.cpp
1> prelu_layer.cpp
1> proposal_layer.cpp
1> psroi_pooling_layer.cpp
1> recurrent_layer.cpp
1> reduction_layer.cpp
1> relu_layer.cpp
1> reshape_layer.cpp
1> resize_layer.cpp
1> roi_pooling_layer.cpp
1> scale_layer.cpp
1> sigmoid_cross_entropy_loss_layer.cpp
1> sigmoid_layer.cpp
1> silence_layer.cpp
1> slice_layer.cpp
1> smooth_L1_loss_layer.cpp
1> softmax_layer.cpp
1> softmax_loss_layer.cpp
1> soft_contrastive_loss_layer.cpp
1> soft_truncation_layer.cpp
1> split_layer.cpp
1> spp_layer.cpp
1> subregion_layer.cpp
1> tanh_layer.cpp
1> threshold_layer.cpp
1> tile_layer.cpp
1> transformer_layer.cpp
1> triplet_loss_layer.cpp
1> truncation_layer.cpp
1> window_data_layer.cpp
1> layer_factory.cpp
1> net.cpp
1> parallel.cpp
1> caffe.pb.cc
1> solver.cpp
1> adadelta_solver.cpp
1> adagrad_solver.cpp
1> adam_solver.cpp
1> nesterov_solver.cpp
1> rmsprop_solver.cpp
1> sgd_solver.cpp
1> syncedmem.cpp
1> benchmark.cpp
1> blocking_queue.cpp
1> cudnn.cpp
1> db.cpp
1> db_leveldb.cpp
1> db_lmdb.cpp
1> hdf5.cpp
1> im2col.cpp
1> insert_splits.cpp
1> io.cpp
1> math_functions.cpp
1> nms.cpp
1> signal_handler.cpp
1> upgrade_proto.cpp
1> util_img.cpp
1> libcaffe.vcxproj -> caffe-windows-ms\windows\..\Build\x64\Release\libcaffe.lib
========== 全部重新生成: 成功 1 个，失败 0 个，跳过 0 个 ==========