Tensorflow c++ 实践及各种坑
在这篇文章中：
Tensorflow当前官网仅包含python、C、Java、Go的发布包，并无C++ release包，并且tensorflow官网也注明了并不保证除python以外库的稳定性，在功能方面python也是最完善的。众所周知，python在开发效率、易用性上有着巨大的优势，但作为一个解释性语言，在性能方面还是存在比较大的缺陷，在各类AI服务化过程中，采用python作为模型快速构建工具，使用高级语言(如C++，java)作为服务化程序实现是大势所趋。本文重点介绍tensorflow C++服务化过程中实现方式及遇到的各种问题。
实现方案
对于tensorflow c++库的使用，有两种方法：
(1) 最佳方式当然是直接用C++构建graph，但是当前c++tensorflow库并不像python api那样full-featured。可参照builds a small graph in c++ here, C++ tensorflow api中还包含cpu和gpu的数字内核实现的类，可用以添加新的op。可参照https://www.tensorflow.org/extend/adding_an_op
(2) 常用的方式，c++调用python生成好的graph。本文主要介绍该方案。
实现步骤
(1) 编译tensorflow源码C++ so(2) 模型训练输出结果(3) 模型固化(4) 模型加载及运行(5) 运行问题
(1) 源码编译
环境要求： 公司tlinux2.2版本， GCC版本 >= 4.8.5安装组件： protobuf 3.3.0 bazel 0.5.0 python 2.7 java8机器要求： 4GB内存
a. 安装java8
yum install java b. 安装protobuf 3.3.0
下载https://github.com/google/protobuf/archive/v3.3.0.zip
./configure && make && make install c. 安装bazel
download https://github.com/bazelbuild/bazel/releases sh bazel-0.5.0-installer-linux-x86_64.sh d. 编译源码
最好采用最新release版本：https://github.com/tensorflow/tensorflow/releases
bazel build //tensorflow:libtensorflow_cc.so 编译过程中可能遇到的问题：问题一： fatal error: unsupported/Eigen/CXX11/Tensor: No such file or directory
安装Eigen3.3或以上版本问题二： java.io.IOException: Cannot run program "patch"
yum install patch 问题三： 内存不够
(2) 模型训练与输出
模型训练输出可参照改用例去实践https://blog.metaflow.fr/tensorflow-saving-restoring-and-mixing-multiple-models-c4c94d5d7125， google上也很多，模型训练保存好得到下面文件：
(3) 模型固化
模型固化方式有三种：
a. freeze_graph 工具
bazel build tensorflow/python/tools:freeze_graph && bazel-bin/tensorflow/python/tools/freeze_graph --input_graph=graph.pb --input_checkpoint=checkpoint --output_graph=./frozen_graph.pb --output_node_names=output/output/scores b. 利用freeze_graph.py工具
# We save out the graph to disk, and then call the const conversion # routine. checkpoint_state_name = "checkpoint" input_graph_name = "graph.pb" output_graph_name = "frozen_graph.pb" input_graph_path = os.path.join(FLAGS.model_dir, input_graph_name) input_saver_def_path = "" input_binary = False input_checkpoint_path = os.path.join(FLAGS.checkpoint_dir, 'saved_checkpoint') + "-0" # Note that we this normally should be only "output_node"!!! output_node_names = "output/output/scores" restore_op_name = "save/restore_all" filename_tensor_name = "save/Const:0" output_graph_path = os.path.join(FLAGS.model_dir, output_graph_name) clear_devices = False freeze_graph.freeze_graph(input_graph_path, input_saver_def_path, input_binary, input_checkpoint_path, output_node_names, restore_op_name, filename_tensor_name, output_graph_path, clear_devices) c. 利用tensorflow python
import os, argparse import tensorflow as tf from tensorflow.python.framework import graph_util dir = os.path.dirname(os.path.realpath(__file__)) def freeze_graph(model_folder): # We retrieve our checkpoint fullpath checkpoint = tf.train.get_checkpoint_state(model_folder) input_checkpoint = checkpoint.model_checkpoint_path # We precise the file fullname of our freezed graph absolute_model_folder = "/".join(input_checkpoint.split('/')[:-1]) output_graph = absolute_model_folder + "/frozen_model.pb" print output_graph # Before exporting our graph, we need to precise what is our output node # This is how TF decides what part of the Graph he has to keep and what part it can dump # NOTE: this variable is plural, because you can have multiple output nodes output_node_names = "output/output/scores" # We clear devices to allow TensorFlow to control on which device it will load operations clear_devices = True # We import the meta graph and retrieve a Saver saver = tf.train.import_meta_graph(input_checkpoint + '.meta', clear_devices=clear_devices) # We retrieve the protobuf graph definition graph = tf.get_default_graph() input_graph_def = graph.as_graph_def() # fix batch norm nodes for node in input_graph_def.node: if node.op == 'RefSwitch': node.op = 'Switch' for index in xrange(len(node.input)): if 'moving_' in node.input[index]: node.input[index] = node.input[index] + '/read' elif node.op == 'AssignSub': node.op = 'Sub' if 'use_locking' in node.attr: del node.attr['use_locking'] # We start a session and restore the graph weights with tf.Session() as sess: saver.restore(sess, input_checkpoint) # We use a built-in TF helper to export variables to constants output_graph_def = graph_util.convert_variables_to_constants( sess, # The session is used to retrieve the weights input_graph_def, # The graph_def is used to retrieve the nodes output_node_names.split(",") # The output node names are used to select the usefull nodes ) # Finally we serialize and dump the output graph to the filesystem with tf.gfile.GFile(output_graph, "wb") as f: f.write(output_graph_def.SerializeToString()) print("%d ops in the final graph." % len(output_graph_def.node)) if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument("--model_folder", type=str, help="Model folder to export") args = parser.parse_args() freeze_graph(args.model_folder) 坑 BatchNorm bug
在具体实际项目，用方式一与方式二将生成的模型利用tensorflow c++ api加载，报以上错误，采用tensorflow python加载模型报同样错：
原因是模型中用到了BatchNorm，修复方式如上面c中给出的方案
(4) 模型加载及运行
构建输入输出
模型输入输出主要就是构造输入输出矩阵，相比python的numpy库，tensorflow提供的Tensor和Eigen::Tensor还是非常难用的，特别是动态矩阵创建，如果你的编译器支持C++14，可以用xTensor库，和numpy一样强大，并且用法机器类似。如果是C++11版本就好好看看eigen库和tensorflow::Tensor文档吧。例举集中简单的用法：
矩阵赋值：
tensorflow::Tensor four_dim_plane(DT_FLOAT, tensorflow::TensorShape({1, MODEL_X_AXIS_LEN, MODEL_Y_AXIS_LEN, fourth_dim_size})); auto plane_tensor = four_dim_plane.tensor<float, 4>(); for (uint32_t k = 0; k < array_plane.size(); ++k) { for (uint32_t j = 0; j < MODEL_Y_AXIS_LEN; ++j) { for (uint32_t i = 0; i < MODEL_X_AXIS_LEN; ++i) { plane_tensor(0, i, j, k) = array_plane[k](i, j); } } } SOFTMAX:
Eigen::Tensor<float, 1> ModelApp::TensorSoftMax(const Eigen::Tensor<float, 1>& tensor) { Eigen::Tensor<float, 0> max = tensor.maximum(); auto e_x = (tensor - tensor.constant(max())).exp(); Eigen::Tensor<float, 0> e_x_sum = e_x.sum(); return e_x / e_x_sum(); } 模型加载及session初始化：
int32_t ModelApp::Init(const std::string& graph_file, Logger *logger) { auto status = NewSession(SessionOptions(), &m_session); if (!status.ok()) { LOG_ERR(logger, "New session failed! %s", status.ToString().c_str()); return Error::ERR_FAILED_NEW_TENSORFLOW_SESSION; } GraphDef graph_def; status = ReadBinaryProto(Env::Default(), graph_file, &graph_def); if (!status.ok()) { LOG_ERR(logger, "Read binary proto failed! %s", status.ToString().c_str()); return Error::ERR_FAILED_READ_BINARY_PROTO; } status = m_session->Create(graph_def); if (!status.ok()) { LOG_ERR(logger, "Session create failed! %s", status.ToString().c_str()); return Error::ERR_FAILED_CREATE_TENSORFLOW_SESSION; } return Error::Success; } 运行：
0.10以上的tensorflow库是线程安全的，因此可多线程调用predict
int32_t ModelApp::Predict(const Action& action, std::vector<int>* info, Logger *logger) { ... auto tensor_x = m_writer->Generate(action, logger); Tensor phase_train(DT_BOOL, TensorShape()); phase_train.scalar<bool>()() = false; std::vector<std::pair<std::string, Tensor>> inputs = { {"input_x", tensor_x}, {"phase_train", phase_train} }; std::vector<Tensor> result; auto status = m_session->Run(inputs, {"output/output/scores"}, {}, &result); if (!status.ok()) { LOG_ERR(logger, "Session run failed! %s", status.ToString().c_str()); return Error::ERR_FAILED_TENSORFLOW_EXECUTION; } ... auto scores = result[0].flat<float>() ; ... return Error::SUCCESS; } (5) 运行问题
问题一：运行告警
2017-08-16 14:11:14.393295: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2017-08-16 14:11:14.393324: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2017-08-16 14:11:14.393331: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2017-08-16 14:11:14.393338: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations. 是因为在编译tensorflow so库的时候没有把这些CPU加速指令编译进去，因此可以在编译的时候加入加速指令，在没有GPU条件下，加入这些库实测可以将CPU计算提高10%左右。
bazel build -c opt --copt=-mavx --copt=-mfma --copt=-mfpmath=both --copt=-msse4.2 -k //tensorflow:libtensorflow_cc.so 需要注意的是并不是所有CPU都支持这些指令，一定要实机测试，以免abort。
问题二: C++ libtensorflow和python tensorflow混用
为验证C++加载模型调用的准确性，利用swig将c++ api封装成了python库供python调用，在同时import tensorflow as tf和import封装好的python swig接口时，core dump
该问题tensorflow官方并不打算解决
