tensorflow: python save model, cpp load model & weight

load_mode.cpp

#include "tensorflow/core/public/session.h"
#include "tensorflow/core/platform/env.h"

using namespace tensorflow;
int main(int argc, char* argv[]) {
  // Initialize a tensorflow session
  Session* session;
  Status status = NewSession(SessionOptions(), &session);
  if (!status.ok()) {
    std::cout << status.ToString() << "\n";
    return 1;
  }

  GraphDef graph_def;
  std::string graph_path = std::string(argv[1]) + ".pb";
  status = ReadBinaryProto(Env::Default(), graph_path.c_str(), &graph_def);
  std::cout << "0" << std::endl;
  if (!status.ok()) {
    std::cout << status.ToString() << "\n";
    return 1;
  }

  // Add the graph to the session
  std::cout << "1" << std::endl;
  status = session->Create(graph_def);
  if (!status.ok()) {
    std::cout << status.ToString() << "\n";
    return 1;
  }

  Tensor fpath(DT_STRING, TensorShape({1, 1}));
  //fpath.matrix<std::string>()(0,0) = "/root/models/model.ckpt";
  fpath.matrix<std::string>()(0,0) = argv[1];
  std::vector<std::pair<string, tensorflow::Tensor>> inputs = {
    //{"save/Const:0", fpath},
    {"save/Const", fpath},
  };

  status = session->Run(inputs, {}, {"save/restore_all"}, nullptr);
  std::cout << "2" << std::endl;
  if (!status.ok()) {
    std::cout << status.ToString() << "\n";
    return 1;
  }

  std::vector<tensorflow::Tensor> outputs;
  
  status = session->Run({}, {"c"}, {}, &outputs);
  std::cout << "3" << std::endl;
  if (!status.ok()) {
    std::cout << status.ToString() << "\n";
    return 1;
  }

  // Grab the first output (we only evaluated one graph node: "c")
  // and convert the node to a scalar representation.
  auto output_c = outputs[0].scalar<float>();

  // (There are similar methods for vectors and matrices here:
  // https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/public/tensor.h)

  // Print the results
  std::cout << "4" << std::endl;
  std::cout << outputs[0].DebugString() << "\n"; // Tensor<type: float shape: [] values: 30>
  std::cout << output_c() << "\n"; // 30

  // Free any resources used by the session
  session->Close();
  return 0;
}

save_model.py

  def save_model(self, model_path):
      saver = tf.train.Saver()
      saver.save(self.sess, model_path, meta_graph_suffix="meta", write_meta_graph=True)
      saver_def = saver.as_saver_def()
      print "save_def: filename_tensor_name, restore_op_name, restore_op_name"
      print saver_def.filename_tensor_name
      print saver_def.restore_op_name
      print saver_def.save_tensor_name
      
      model_dir = os.path.dirname(model_path)
      model_name = os.path.basename(model_path)
      ## below call must after previous code, otherwise you could get below error message
      ## Not found: FeedInputs: unable to find feed output save/Const:0
      tf.train.write_graph(self.sess.graph_def, model_dir, model_name + ".pb", as_text=False)