vuron/adept: Платформа для обучения и вывода нейронных сетей, с API приближенным к PyTorch.

tests/maxpool2d_tests.cpp

37 строк2 KB

kolkir

MaxPool2d cpu backend stub and test

09 фев 2025, 12:27

09 фев 2025, 12:27d0343a3

#include <adept/nn/maxpool2d.hpp>

#include "catch.hpp"

using namespace adept;

TEST_CASE("MaxPool2d construction", "[maxpool2d]") {
  MaxPool2d pool(MaxPool2dOptions(3));
}

TEST_CASE("MaxPool2d im3x3", "[maxpool2d]") {
  std::vector<float32_t> im_data = {1, 2, 3, 4, 5, 6, 7, 8, 9};

  auto im = Variable(Tensor::from_blob(
      im_data.data(),
      {.shape = {1, 1, 3, 3}, .device = device_t::CPU, .dtype = to_dtype<float32_t>()}));

  MaxPool2d pool(MaxPool2dOptions(2).with_stride(1));
  auto res = pool(im);

  REQUIRE_THAT(res.data().at<float32_t>({0, 0, 0, 0}), Catch::WithinRel(5.f, 0.001f));
  REQUIRE_THAT(res.data().at<float32_t>({0, 0, 0, 1}), Catch::WithinRel(6.f, 0.001f));
  REQUIRE_THAT(res.data().at<float32_t>({0, 0, 1, 0}), Catch::WithinRel(8.f, 0.001f));
  REQUIRE_THAT(res.data().at<float32_t>({0, 0, 1, 1}), Catch::WithinRel(9.f, 0.001f));

  mean(res).backward();

  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 0, 0}), Catch::WithinRel(0.0f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 0, 1}), Catch::WithinRel(0.0f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 0, 2}), Catch::WithinRel(0.0f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 1, 0}), Catch::WithinRel(0.0f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 1, 1}), Catch::WithinRel(0.25f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 1, 2}), Catch::WithinRel(0.25f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 2, 0}), Catch::WithinRel(0.0f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 2, 1}), Catch::WithinRel(0.25f, 0.001f));
  REQUIRE_THAT(im.grad().at<float32_t>({0, 0, 2, 2}), Catch::WithinRel(0.25f, 0.001f));
}