!25040 [MS][LITE][TOD] Fix Memory leak in unified API code example

Merge pull request !25040 from ehaleva/export

mindspore/lite/examples/train_lenet/model/prepare_model.sh

@@ -5,10 +5,10 @@ if [[ -z ${EXPORT} ]]; then
     rm -f lenet_tod.mindir
   if [ -n "$2" ]; then
     DOCKER_IMG=$2
-    docker run -w $PWD --runtime=nvidia -v /home/$USER:/home/$USER --privileged=true ${DOCKER_IMG} /bin/bash -c "PYTHONPATH=../../../../../model_zoo/official/cv/lenet/src python lenet_export.py '$1'; chmod 444 lenet_tod.mindir; rm -rf __pycache__"
+    docker run -w $PWD --runtime=nvidia -v /home/$USER:/home/$USER --privileged=true ${DOCKER_IMG} /bin/bash -c "PYTHONPATH=../../../../../tests/perf_test python lenet_export.py '$1'; chmod 444 lenet_tod.mindir; rm -rf __pycache__"
   else
     echo "MindSpore docker was not provided, attempting to run locally"
-    PYTHONPATH=../../../../../model_zoo/official/cv/lenet/src python lenet_export.py $1
+    PYTHONPATH=../../../../../tests/perf_test python lenet_export.py $1
   fi
 fi
 

mindspore/lite/examples/unified_api/model/prepare_model.sh

@@ -5,10 +5,10 @@ if [[ -z ${EXPORT} ]]; then
     rm -f lenet_tod.mindir
   if [ -n "$2" ]; then
     DOCKER_IMG=$2
-    docker run -w $PWD --runtime=nvidia -v /home/$USER:/home/$USER --privileged=true ${DOCKER_IMG} /bin/bash -c "PYTHONPATH=../../../../../model_zoo/official/cv/lenet/src python lenet_export.py '$1'; chmod 444 lenet_tod.mindir; rm -rf __pycache__"
+    docker run -w $PWD --runtime=nvidia -v /home/$USER:/home/$USER --privileged=true ${DOCKER_IMG} /bin/bash -c "PYTHONPATH=../../../../../tests/perf_test python lenet_export.py '$1'; chmod 444 lenet_tod.mindir; rm -rf __pycache__"
   else
     echo "MindSpore docker was not provided, attempting to run locally"
-    PYTHONPATH=../../../../../model_zoo/official/cv/lenet/src python lenet_export.py $1
+    PYTHONPATH=../../../../../tests/perf_test python lenet_export.py $1
   fi
 fi
 

mindspore/lite/examples/unified_api/src/net_runner.cc

@@ -109,39 +109,15 @@ class Measurement : public mindspore::TrainCallBack {
   unsigned int epochs_;
 };
 
-// Definition of verbose callback function after forwarding operator.
-bool after_callback(const std::vector<mindspore::MSTensor> &after_inputs,
-                    const std::vector<mindspore::MSTensor> &after_outputs,
-                    const mindspore::MSCallBackParam &call_param) {
-  printf("%s\n", call_param.node_name.c_str());
-  for (size_t i = 0; i < after_inputs.size(); i++) {
-    int num2p = (after_inputs.at(i).ElementNum());
-    printf("in%zu(%d): ", i, num2p);
-    if (num2p > kPrintNum) num2p = kPrintNum;
-    if (after_inputs.at(i).DataType() == mindspore::DataType::kNumberTypeInt32) {
-      auto d = reinterpret_cast<const int *>(after_inputs.at(i).Data().get());
-      for (int j = 0; j < num2p; j++) printf("%d, ", d[j]);
-    } else {
-      auto d = reinterpret_cast<const float *>(after_inputs.at(i).Data().get());
-      for (int j = 0; j < num2p; j++) printf("%f, ", d[j]);
-    }
-    printf("\n");
+NetRunner::~NetRunner() {
+  if (model_ != nullptr) {
+    delete model_;
   }
-  for (size_t i = 0; i < after_outputs.size(); i++) {
-    auto d = reinterpret_cast<const float *>(after_inputs.at(i).Data().get());
-    int num2p = (after_outputs.at(i).ElementNum());
-    printf("ou%zu(%d): ", i, num2p);
-    if (num2p > kElem2Print) {
-      num2p = kElem2Print;
-    }
-    for (int j = 0; j < num2p; j++) printf("%f, ", d[j]);
-    printf("\n");
+  if (graph_ != nullptr) {
+    delete graph_;
   }
-  return true;
 }
 
-NetRunner::~NetRunner() {}
-
 void NetRunner::InitAndFigureInputs() {
   auto context = std::make_shared<mindspore::Context>();
   auto cpu_context = std::make_shared<mindspore::CPUDeviceInfo>();
@@ -149,6 +125,8 @@ void NetRunner::InitAndFigureInputs() {
   context->MutableDeviceInfo().push_back(cpu_context);
 
   graph_ = new mindspore::Graph();
+  MS_ASSERT(graph_ != nullptr);
+
   auto status = mindspore::Serialization::Load(ms_file_, mindspore::kMindIR, graph_);
   if (status != mindspore::kSuccess) {
     std::cout << "Error " << status << " during serialization of graph " << ms_file_;
@@ -161,6 +139,8 @@ void NetRunner::InitAndFigureInputs() {
   }
 
   model_ = new mindspore::Model();
+  MS_ASSERT(model_ != nullptr);
+
   status = model_->Build(mindspore::GraphCell(*graph_), context, cfg);
   if (status != mindspore::kSuccess) {
     std::cout << "Error " << status << " during build of model " << ms_file_;
@@ -168,6 +148,7 @@ void NetRunner::InitAndFigureInputs() {
   }
 
   acc_metrics_ = std::shared_ptr<AccuracyMetrics>(new AccuracyMetrics);
+  MS_ASSERT(acc_metrics_ != nullptr);
   model_->InitMetrics({acc_metrics_.get()});
 
   auto inputs = model_->GetInputs();

mindspore/lite/src/cxx_api/callback/ckpt_saver.cc

@@ -34,6 +34,7 @@ CkptSaver::CkptSaver(int save_every_n, const std::vector<char> &filename_prefix)
 CkptSaver::~CkptSaver() {
   if (callback_impl_ == nullptr) {
     MS_LOG(ERROR) << "Callback implement is null.";
+    return;
   }
   auto internal_call_back = callback_impl_->GetInternalCallback();
   if (internal_call_back != nullptr) {

mindspore/lite/src/cxx_api/callback/loss_monitor.cc

@@ -33,6 +33,7 @@ LossMonitor::LossMonitor(int print_every_n_steps) {
 LossMonitor::~LossMonitor() {
   if (callback_impl_ == nullptr) {
     MS_LOG(ERROR) << "Callback implement is null.";
+    return;
   }
   auto internal_call_back = callback_impl_->GetInternalCallback();
   if (internal_call_back != nullptr) {

mindspore/lite/src/cxx_api/callback/train_accuracy.cc

@@ -35,6 +35,7 @@ TrainAccuracy::TrainAccuracy(int print_every_n, int accuracy_metrics, const std:
 TrainAccuracy::~TrainAccuracy() {
   if (callback_impl_ == nullptr) {
     MS_LOG(ERROR) << "Callback implement is null.";
+    return;
   }
   auto internal_call_back = callback_impl_->GetInternalCallback();
   if (internal_call_back != nullptr) {

mindspore/lite/src/cxx_api/metrics/accuracy.cc

@@ -35,6 +35,7 @@ AccuracyMetrics::AccuracyMetrics(int accuracy_metrics, const std::vector<int> &i
 AccuracyMetrics::~AccuracyMetrics() {
   if (metrics_impl_ == nullptr) {
     MS_LOG(ERROR) << "Metrics implement is null.";
+    return;
   }
   auto internal_metrics = metrics_impl_->GetInternalMetrics();
   if (internal_metrics != nullptr) {