[test] learning rate scheduler unittest

Add unittests for the constant and exponential learning rate schedulers.

Signed-off-by: Parichay Kapoor <pk.kapoor@samsung.com>

diff --git a/jni/Android.mk b/jni/Android.mk
index db923c9..8d3a761 100644 (file)
--- a/jni/Android.mk
+++ b/jni/Android.mk
@@ -196,6 +196,8 @@ NNTRAINER_SRCS := $(NNTRAINER_ROOT)/nntrainer/models/neuralnet.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/optimizers/optimizer_impl.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/optimizers/adam.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/optimizers/sgd.cpp \
+                  $(NNTRAINER_ROOT)/nntrainer/optimizers/lr_scheduler_constant.cpp \
+                  $(NNTRAINER_ROOT)/nntrainer/optimizers/lr_scheduler_exponential.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/utils/util_func.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/utils/ini_wrapper.cpp \
                   $(NNTRAINER_ROOT)/nntrainer/utils/profiler.cpp \

diff --git a/test/unittest/meson.build b/test/unittest/meson.build
index 389f5c1..90cab48 100644 (file)
--- a/test/unittest/meson.build
+++ b/test/unittest/meson.build
@@ -42,6 +42,7 @@ test_target = [
   ['unittest_base_properties', []],
   ['unittest_common_properties', []],
   ['unittest_nntrainer_tensor_pool', []],
+  ['unittest_nntrainer_lr_scheduler', []],
 ]
 
 if get_option('enable-profile')

diff --git a/test/unittest/unittest_nntrainer_lr_scheduler.cpp b/test/unittest/unittest_nntrainer_lr_scheduler.cpp
new file mode 100644 (file)
index 0000000..6064a37
--- /dev/null
+++ b/test/unittest/unittest_nntrainer_lr_scheduler.cpp
@@ -0,0 +1,140 @@
+// SPDX-License-Identifier: Apache-2.0
+/**
+ * Copyright (C) 2021 Parichay Kapoor <pk.kapoor@samsung.com>
+ *
+ * @file   unittest_nntrainer_lr_scheduler.cpp
+ * @date   09 December 2021
+ * @brief  This is unittests for learning rate schedulers
+ * @see    https://github.com/nnstreamer/nntrainer
+ * @author Parichay Kapoor <pk.kapoor@samsung.com>
+ * @bug    No known bugs except for NYI items
+ *
+ */
+
+#include <fstream>
+#include <gtest/gtest.h>
+
+#include <lr_scheduler.h>
+#include <lr_scheduler_constant.h>
+#include <lr_scheduler_exponential.h>
+#include <nntrainer_error.h>
+
+#include "nntrainer_test_util.h"
+
+/**
+ * @brief test constructing lr scheduler
+ *
+ */
+TEST(lr_constant, ctor_initializer_01_p) {
+  EXPECT_NO_THROW(nntrainer::createLearningRateScheduler<
+                  nntrainer::ConstantLearningRateScheduler>({}));
+}
+
+/**
+ * @brief test constructing lr scheduler
+ *
+ */
+TEST(lr_constant, ctor_initializer_02_p) {
+  auto &ac = nntrainer::AppContext::Global();
+  EXPECT_NO_THROW(
+    ac.createObject<nntrainer::LearningRateScheduler>("constant"));
+}
+
+/**
+ * @brief test constructing lr scheduler
+ *
+ */
+TEST(lr_constant, ctor_initializer_03_n) {
+  auto &ac = nntrainer::AppContext::Global();
+  EXPECT_ANY_THROW(ac.createObject<nntrainer::LearningRateScheduler>("random"));
+}
+
+/**
+ * @brief test set and get learning rate
+ *
+ */
+TEST(lr_constant, prop_01_n) {
+  auto &ac = nntrainer::AppContext::Global();
+  auto lr = ac.createObject<nntrainer::LearningRateScheduler>("constant");
+
+  /** fails as learning rate property is not set */
+  EXPECT_ANY_THROW(lr->getLearningRate(0));
+}
+
+/**
+ * @brief test set and get learning rate
+ *
+ */
+TEST(lr_constant, prop_02_n) {
+  auto &ac = nntrainer::AppContext::Global();
+  auto lr = ac.createObject<nntrainer::LearningRateScheduler>("constant");
+
+  EXPECT_ANY_THROW(lr->setProperty({"random=1.0"}));
+}
+
+/**
+ * @brief test set and get learning rate
+ *
+ */
+TEST(lr_constant, prop_03_p) {
+  auto &ac = nntrainer::AppContext::Global();
+  auto lr = ac.createObject<nntrainer::LearningRateScheduler>("constant");
+
+  EXPECT_NO_THROW(lr->setProperty({"learning_rate=1.0"}));
+
+  EXPECT_NO_THROW(lr->getLearningRate(0));
+  EXPECT_FLOAT_EQ(lr->getLearningRate(0), lr->getLearningRate(100));
+  EXPECT_FLOAT_EQ(lr->getLearningRate(10), 1.0f);
+}
+
+/**
+ * @brief test set and get learning rate
+ *
+ */
+TEST(lr_exponential, prop_01_n) {
+  auto &ac = nntrainer::AppContext::Global();
+  auto lr = ac.createObject<nntrainer::LearningRateScheduler>("exponential");
+
+  EXPECT_ANY_THROW(lr->getLearningRate(0));
+}
+
+/**
+ * @brief test set and get learning rate
+ *
+ */
+TEST(lr_exponential, prop_02_p) {
+  auto &ac = nntrainer::AppContext::Global();
+  auto lr = ac.createObject<nntrainer::LearningRateScheduler>("exponential");
+
+  EXPECT_NO_THROW(lr->setProperty({"learning_rate=1.0"}));
+  EXPECT_ANY_THROW(lr->getLearningRate(0));
+
+  EXPECT_NO_THROW(lr->setProperty({"decay_steps=1"}));
+  EXPECT_ANY_THROW(lr->getLearningRate(0));
+
+  EXPECT_NO_THROW(lr->setProperty({"decay_rate=0.9"}));
+  EXPECT_NO_THROW(lr->getLearningRate(0));
+
+  EXPECT_FLOAT_EQ(lr->getLearningRate(0), 1.0f);
+  EXPECT_FLOAT_EQ(lr->getLearningRate(1), 1.0 * std::pow(0.9f, 1));
+  EXPECT_FLOAT_EQ(lr->getLearningRate(3), 1.0 * std::pow(0.9f, 3));
+}
+
+int main(int argc, char **argv) {
+  int result = -1;
+
+  try {
+    testing::InitGoogleTest(&argc, argv);
+  } catch (...) {
+    std::cerr << "Error duing InitGoogleTest" << std::endl;
+    return 0;
+  }
+
+  try {
+    result = RUN_ALL_TESTS();
+  } catch (...) {
+    std::cerr << "Error duing RUN_ALL_TESTS()" << std::endl;
+  }
+
+  return result;
+}