Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add test for eta and mitigate float error. (#7446)

Browse Source 
* Add eta test.
* Don't skip test.
This commit is contained in:
Jiaming Yuan
2021-11-18 20:42:48 +08:00
committed by GitHub
parent 7cfb310eb4
commit 9fb4338964
2 changed files with 65 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
tests/cpp/test_serialization.cc
View File

@@ -149,6 +149,7 @@ void TestLearnerSerialization(Args args, FeatureMap const& fmap, std::shared_ptr
Json m_0 = Json::Load(StringView{continued_model.c_str(), continued_model.size()});
Json m_1 = Json::Load(StringView{model_at_2kiter.c_str(), model_at_2kiter.size()});
CompareJSON(m_0, m_1);
}
@@ -610,7 +611,6 @@ TEST_F(MultiClassesSerializationTest, CPUCoordDescent) {
#if defined(XGBOOST_USE_CUDA)
TEST_F(MultiClassesSerializationTest, GpuHist) {
GTEST_SKIP() << "This test is broken for CUDA 11.0 + Windows combination, skipping";
TestLearnerSerialization({{"booster", "gbtree"},
{"num_class", std::to_string(kClasses)},
{"seed", "0"},
@@ -620,6 +620,9 @@ TEST_F(MultiClassesSerializationTest, GpuHist) {
// different result (1e-7) with CPU predictor for some
// entries.
{"predictor", "gpu_predictor"},
// Mitigate the difference caused by hardware fused multiply
// add to tree weight during update prediction cache.
{"learning_rate", "1.0"},
{"tree_method", "gpu_hist"}},
fmap_, p_dmat_);
@@ -630,7 +633,8 @@ TEST_F(MultiClassesSerializationTest, GpuHist) {
{"max_depth", std::to_string(kClasses)},
// GPU_Hist has higher floating point error. 1e-6 doesn't work
// after num_parallel_tree goes to 4
{"num_parallel_tree", "3"},
{"num_parallel_tree", "4"},
{"learning_rate", "1.0"},
{"tree_method", "gpu_hist"}},
fmap_, p_dmat_);
@@ -638,6 +642,7 @@ TEST_F(MultiClassesSerializationTest, GpuHist) {
{"num_class", std::to_string(kClasses)},
{"seed", "0"},
{"nthread", "1"},
{"learning_rate", "1.0"},
{"max_depth", std::to_string(kClasses)},
{"tree_method", "gpu_hist"}},
fmap_, p_dmat_);