[Benchmark] Benchmark cpp for YOLOv5 (#1260)

* 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 --------- Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com>
2025-10-04 16:22:57 +08:00 · 2023-02-08 15:49:28 +08:00
parent 9cd00ad4c5
commit 66f059b65b
25 changed files with 432 additions and 56 deletions
--- a/benchmark/cpp/flags.h
+++ b/benchmark/cpp/flags.h
@@ -0,0 +1,99 @@
+// Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#pragma once
+
+#include "gflags/gflags.h"
+#include "fastdeploy/utils/perf.h"
+
+DEFINE_string(model, "", "Directory of the inference model.");
+DEFINE_string(image, "", "Path of the image file.");
+DEFINE_string(device, "cpu",
+              "Type of inference device, support 'cpu' or 'gpu'.");
+DEFINE_int32(device_id, 0, "device(gpu) id.");
+DEFINE_int32(warmup, 200, "Number of warmup for profiling.");
+DEFINE_int32(repeat, 1000, "Number of repeats for profiling.");
+DEFINE_string(profile_mode, "runtime", "runtime or end2end.");
+DEFINE_string(backend, "default",
+              "The inference runtime backend, support: ['default', 'ort', "
+              "'paddle', 'ov', 'trt', 'paddle_trt']");
+DEFINE_int32(cpu_thread_nums, 8, "Set numbers of cpu thread.");
+DEFINE_bool(
+    include_h2d_d2h, false, "Whether run profiling with h2d and d2h.");
+DEFINE_bool(
+    use_fp16, false,
+    "Whether to use FP16 mode, only support 'trt' and 'paddle_trt' backend");
+DEFINE_bool(
+    collect_memory_info, false, "Whether to collect memory info");
+DEFINE_int32(dump_period, 100, "How often to collect memory info.");
+
+void PrintUsage() {
+  std::cout << "Usage: infer_demo --model model_path --image img_path --device "
+               "[cpu|gpu] --backend "
+               "[default|ort|paddle|ov|trt|paddle_trt] "
+               "--use_fp16 false"
+            << std::endl;
+  std::cout << "Default value of device: cpu" << std::endl;
+  std::cout << "Default value of backend: default" << std::endl;
+  std::cout << "Default value of use_fp16: false" << std::endl;
+}
+
+bool CreateRuntimeOption(fastdeploy::RuntimeOption* option) {
+  if (FLAGS_device == "gpu") {
+    option->UseGpu();
+    if (FLAGS_backend == "ort") {
+      option->UseOrtBackend();
+    } else if (FLAGS_backend == "paddle") {
+      option->UsePaddleInferBackend();
+    } else if (FLAGS_backend == "trt" || FLAGS_backend == "paddle_trt") {
+      option->UseTrtBackend();
+      option->SetTrtInputShape("input", {1, 3, 112, 112});
+      if (FLAGS_backend == "paddle_trt") {
+        option->EnablePaddleToTrt();
+      }
+      if (FLAGS_use_fp16) {
+        option->EnableTrtFP16();
+      }
+    } else if (FLAGS_backend == "default") {
+      return true;
+    } else {
+      std::cout << "While inference with GPU, only support "
+                   "default/ort/paddle/trt/paddle_trt now, "
+                << FLAGS_backend << " is not supported." << std::endl;
+      return false;
+    }
+  } else if (FLAGS_device == "cpu") {
+    option->SetCpuThreadNum(FLAGS_cpu_thread_nums);
+    if (FLAGS_backend == "ort") {
+      option->UseOrtBackend();
+    } else if (FLAGS_backend == "ov") {
+      option->UseOpenVINOBackend();
+    } else if (FLAGS_backend == "paddle") {
+      option->UsePaddleInferBackend();
+    } else if (FLAGS_backend == "default") {
+      return true;
+    } else {
+      std::cout << "While inference with CPU, only support "
+                   "default/ort/ov/paddle now, "
+                << FLAGS_backend << " is not supported." << std::endl;
+      return false;
+    }
+  } else {
+    std::cerr << "Only support device CPU/GPU now, " << FLAGS_device
+              << " is not supported." << std::endl;
+    return false;
+  }
+
+  return true;
+}