[Benchmark] Add SaveBenchmarkResult func for benchmark (#1442)

* add GPL lisence

* add GPL-3.0 lisence

* add GPL-3.0 lisence

* add GPL-3.0 lisence

* support yolov8

* add pybind for yolov8

* add yolov8 readme

* add cpp benchmark

* add cpu and gpu mem

* public part split

* add runtime mode

* fixed bugs

* add cpu_thread_nums

* deal with comments

* deal with comments

* deal with comments

* rm useless code

* add FASTDEPLOY_DECL

* add FASTDEPLOY_DECL

* fixed for windows

* mv rss to pss

* mv rss to pss

* Update utils.cc

* use thread to collect mem

* Add ResourceUsageMonitor

* rm useless code

* fixed bug

* fixed typo

* update ResourceUsageMonitor

* fixed bug

* fixed bug

* add note for ResourceUsageMonitor

* deal with comments

* add macros

* deal with comments

* deal with comments

* deal with comments

* re-lint

* rm pmap and use mem api

* rm pmap and use mem api

* add mem api

* Add PrintBenchmarkInfo func

* Add PrintBenchmarkInfo func

* Add PrintBenchmarkInfo func

* deal with comments

* fixed enable_paddle_to_trt

* add log for paddle_trt

* support ppcls benchmark

* use new trt option api

* update benchmark info

* simplify benchmark.cc

* simplify benchmark.cc

* deal with comments

* Add ppseg && ppocr benchmark

* add OCR rec img

* add ocr benchmark

* fixed trt shape

* add trt shape

* resolve conflict

* add ENABLE_BENCHMARK define

* Add ClassifyDiff

* Add Resize for ClassifyResult

* deal with comments

* add convert info script

* resolve conflict

* Add SaveBenchmarkResult func

* fixed bug

* fixed bug

* fixed bug

---------

Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com>

benchmark/cpp/CMakeLists.txt

@@ -36,4 +36,7 @@ else()
   target_link_libraries(benchmark_ppocr_cls ${FASTDEPLOY_LIBS} gflags)
   target_link_libraries(benchmark_ppocr_rec ${FASTDEPLOY_LIBS} gflags)
 endif()
-install_fastdeploy_libraries(${CMAKE_CURRENT_BINARY_DIR})
+# only for Android ADB test
+if(ANDROID)
+  install_fastdeploy_libraries(${CMAKE_CURRENT_BINARY_DIR})
+endif()

benchmark/cpp/flags.h

@@ -44,6 +44,7 @@ DEFINE_bool(
 DEFINE_bool(
     collect_memory_info, false, "Whether to collect memory info");
 DEFINE_int32(sampling_interval, 50, "How often to collect memory info(ms).");
+DEFINE_string(result_path, "benchmark.txt", "Path of benchmark result file.");
 
 static void PrintUsage() {
   std::cout << "Usage: infer_demo --model model_path --image img_path --device "
@@ -92,6 +93,9 @@ static void PrintBenchmarkInfo() {
        << "ms" << std::endl;
   }
   std::cout << ss.str() << std::endl;
+  // Save benchmark info
+  fastdeploy::benchmark::ResultManager::SaveBenchmarkResult(ss.str(),
+                                                  FLAGS_result_path);
 #endif
   return;
 }

benchmark/cpp/macros.h

@@ -23,6 +23,8 @@
     return 0;                                                               \
   }                                                                         \
   auto __im__ = cv::imread(FLAGS_image);                                    \
+  std::stringstream __ss__;                                                 \
+  __ss__.precision(6);                                                      \
   fastdeploy::benchmark::ResourceUsageMonitor __resource_moniter__(         \
       FLAGS_sampling_interval, FLAGS_device_id);                            \
   if (FLAGS_collect_memory_info) {                                          \
@@ -35,6 +37,7 @@
     }                                                                       \
     double __profile_time__ = MODEL_NAME.GetProfileTime() * 1000;           \
     std::cout << "Runtime(ms): " << __profile_time__ << "ms." << std::endl; \
+    __ss__ << "Runtime(ms): " << __profile_time__ << "ms." << std::endl;    \
   } else {                                                                  \
     std::cout << "Warmup " << FLAGS_warmup << " times..." << std::endl;     \
     for (int __i__ = 0; __i__ < FLAGS_warmup; __i__++) {                    \
@@ -56,14 +59,20 @@
     __tc__.End();                                                           \
     double __end2end__ = __tc__.Duration() / FLAGS_repeat * 1000;           \
     std::cout << "End2End(ms): " << __end2end__ << "ms." << std::endl;      \
+    __ss__ << "End2End(ms): " << __end2end__ << "ms." << std::endl;         \
   }                                                                         \
   if (FLAGS_collect_memory_info) {                                          \
     float __cpu_mem__ = __resource_moniter__.GetMaxCpuMem();                \
     float __gpu_mem__ = __resource_moniter__.GetMaxGpuMem();                \
     float __gpu_util__ = __resource_moniter__.GetMaxGpuUtil();              \
     std::cout << "cpu_rss_mb: " << __cpu_mem__ << "MB." << std::endl;       \
+    __ss__ << "cpu_rss_mb: " << __cpu_mem__ << "MB." << std::endl;          \
     std::cout << "gpu_rss_mb: " << __gpu_mem__ << "MB." << std::endl;       \
+    __ss__ << "gpu_rss_mb: " << __gpu_mem__ << "MB." << std::endl;          \
     std::cout << "gpu_util: " << __gpu_util__ << std::endl;                 \
+    __ss__ << "gpu_rss_mb: " << __gpu_mem__ << "MB." << std::endl;          \
     __resource_moniter__.Stop();                                            \
   }                                                                         \
+  fastdeploy::benchmark::ResultManager::SaveBenchmarkResult(__ss__.str(),   \
+                                          FLAGS_result_path);               \
 }

fastdeploy/benchmark/utils.cc

@@ -350,6 +350,21 @@ TensorDiff ResultManager::CalculateDiffStatis(const FDTensor& lhs,
   }
 }
 
+void ResultManager::SaveBenchmarkResult(const std::string& res,
+                                        const std::string& path) {
+  if (path.empty()) {
+    FDERROR << "Benchmark data path can not be empty!" << std::endl;
+    return;
+  }
+  auto openmode = std::ios::app;
+  std::ofstream fs(path, openmode);
+  if (!fs.is_open()) {
+    FDERROR << "Fail to open result file: " << path << std::endl;
+  }
+  fs << res;
+  fs.close();
+}
+
 #if defined(ENABLE_VISION)
 bool ResultManager::SaveDetectionResult(const vision::DetectionResult& res,
                                         const std::string& path) {

fastdeploy/benchmark/utils.h

@@ -138,6 +138,9 @@ struct FASTDEPLOY_DECL ResultManager {
   /// Calculate diff value between two FDTensor results.
   static TensorDiff CalculateDiffStatis(const FDTensor& lhs,
                                         const FDTensor& rhs);
+  /// Save Benchmark data
+  static void SaveBenchmarkResult(const std::string& res,
+                                  const std::string& path);
 #if defined(ENABLE_VISION)
   /// Save & Load functions for basic results.
   static bool SaveDetectionResult(const vision::DetectionResult& res,