Skip to content

Commit

Permalink
add tests
Browse files Browse the repository at this point in the history
  • Loading branch information
Dmitry Razdoburdin committed Jul 24, 2024
1 parent 663388f commit 8743b00
Show file tree
Hide file tree
Showing 3 changed files with 107 additions and 7 deletions.
5 changes: 2 additions & 3 deletions plugin/sycl/tree/hist_updater.cc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -190,7 +190,7 @@ void HistUpdater<GradientSumT>::EvaluateAndApplySplits(
std::vector<ExpandEntry> nodes_for_apply_split;
AddSplitsToTree(gmat, p_tree, num_leaves, depth,
&nodes_for_apply_split, temp_qexpand_depth);
ApplySplit(nodes_for_apply_split, gmat, hist_, p_tree);
ApplySplit(nodes_for_apply_split, gmat, p_tree);
}

// Split nodes to 2 sets depending on amount of rows in each node
Expand Down Expand Up @@ -306,7 +306,7 @@ void HistUpdater<GradientSumT>::ExpandWithLossGuide(
right_leaf_weight, e.best.loss_chg, e.stats.GetHess(),
e.best.left_sum.GetHess(), e.best.right_sum.GetHess());

this->ApplySplit({candidate}, gmat, hist_, p_tree);
this->ApplySplit({candidate}, gmat, p_tree);

const int cleft = (*p_tree)[nid].LeftChild();
const int cright = (*p_tree)[nid].RightChild();
Expand Down Expand Up @@ -807,7 +807,6 @@ template <typename GradientSumT>
void HistUpdater<GradientSumT>::ApplySplit(
const std::vector<ExpandEntry> nodes,
const common::GHistIndexMatrix& gmat,
const common::HistCollection<GradientSumT, MemoryType::on_device>& hist,
RegTree* p_tree) {
using CommonRowPartitioner = xgboost::tree::CommonRowPartitioner;
builder_monitor_.Start("ApplySplit");
Expand Down
1 change: 0 additions & 1 deletion plugin/sycl/tree/hist_updater.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -119,7 +119,6 @@ class HistUpdater {

void ApplySplit(std::vector<ExpandEntry> nodes,
const common::GHistIndexMatrix& gmat,
const common::HistCollection<GradientSumT, MemoryType::on_device>& hist,
RegTree* p_tree);

void AddSplitsToRowSet(const std::vector<ExpandEntry>& nodes, RegTree* p_tree);
Expand Down
108 changes: 105 additions & 3 deletions tests/cpp/plugin/test_sycl_hist_updater.cc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,8 @@
#include "../../../plugin/sycl/tree/hist_updater.h"
#include "../../../plugin/sycl/device_manager.h"

#include "../../../src/tree/common_row_partitioner.h"

#include "../helpers.h"

namespace xgboost::sycl::tree {
Expand Down Expand Up @@ -59,6 +61,12 @@ class TestHistUpdater : public HistUpdater<GradientSumT> {
HistUpdater<GradientSumT>::EvaluateSplits(nodes_set, gmat, tree);
return HistUpdater<GradientSumT>::snode_host_;
}

auto TestApplySplit(const std::vector<ExpandEntry> nodes,
const common::GHistIndexMatrix& gmat,
RegTree* p_tree) {
HistUpdater<GradientSumT>::ApplySplit(nodes, gmat, p_tree);
}
};

void GenerateRandomGPairs(::sycl::queue* qu, GradientPair* gpair_ptr, size_t num_rows, bool has_neg_hess) {
Expand Down Expand Up @@ -385,6 +393,88 @@ void TestHistUpdaterEvaluateSplits(const xgboost::tree::TrainParam& param) {
ASSERT_NEAR(best_loss_chg_des[0], best_loss_chg, 1e-6);
}

template <typename GradientSumT>
void TestHistUpdaterApplySplit(const xgboost::tree::TrainParam& param, float sparsity, int max_bins) {
const size_t num_rows = 16;
const size_t num_columns = 1;

Context ctx;
ctx.UpdateAllowUnknown(Args{{"device", "sycl"}});

DeviceManager device_manager;
auto qu = device_manager.GetQueue(ctx.Device());

auto p_fmat = RandomDataGenerator{num_rows, num_columns, sparsity}.GenerateDMatrix();
sycl::DeviceMatrix dmat;
dmat.Init(qu, p_fmat.get());

common::GHistIndexMatrix gmat;
gmat.Init(qu, &ctx, dmat, max_bins);

RegTree tree;
tree.ExpandNode(0, 0, 0, false, 0, 0, 0, 0, 0, 0, 0);

std::vector<tree::ExpandEntry> nodes;
nodes.emplace_back(tree::ExpandEntry(0, tree.GetDepth(0)));

FeatureInteractionConstraintHost int_constraints;
TestHistUpdater<GradientSumT> updater(&ctx, qu, param, int_constraints, p_fmat.get());
USMVector<GradientPair, MemoryType::on_device> gpair(&qu, num_rows);
GenerateRandomGPairs(&qu, gpair.Data(), num_rows, false);

auto* row_set_collection = updater.TestInitData(gmat, gpair, *p_fmat, tree);
updater.TestApplySplit(nodes, gmat, &tree);

// Copy indexes to host
std::vector<size_t> row_indices_host(num_rows);
qu.memcpy(row_indices_host.data(), row_set_collection->Data().Data(), sizeof(size_t)*num_rows).wait();

// Reference Implementation
std::vector<size_t> row_indices_desired_host(num_rows);
{
TestHistUpdater<GradientSumT> updater4verification(&ctx, qu, param, int_constraints, p_fmat.get());
auto* row_set_collection4verification = updater4verification.TestInitData(gmat, gpair, *p_fmat, tree);

size_t n_nodes = nodes.size();
std::vector<int32_t> split_conditions(n_nodes);
xgboost::tree::CommonRowPartitioner::FindSplitConditions(nodes, tree, gmat, &split_conditions);

common::PartitionBuilder partition_builder;
partition_builder.Init(&qu, n_nodes, [&](size_t node_in_set) {
const int32_t nid = nodes[node_in_set].nid;
return (*row_set_collection4verification)[nid].Size();
});

::sycl::event event;
partition_builder.Partition(gmat, nodes, (*row_set_collection4verification),
split_conditions, &tree, &event);
qu.wait_and_throw();

for (size_t node_in_set = 0; node_in_set < n_nodes; node_in_set++) {
const int32_t nid = nodes[node_in_set].nid;
size_t* data_result = const_cast<size_t*>((*row_set_collection4verification)[nid].begin);
partition_builder.MergeToArray(node_in_set, data_result, &event);
}
qu.wait_and_throw();

for (size_t i = 0; i < n_nodes; ++i) {
const int32_t nid = nodes[i].nid;
const size_t n_left = partition_builder.GetNLeftElems(i);
const size_t n_right = partition_builder.GetNRightElems(i);

row_set_collection4verification->AddSplit(nid, tree[nid].LeftChild(),
tree[nid].RightChild(), n_left, n_right);
}

qu.memcpy(row_indices_desired_host.data(), row_set_collection4verification->Data().Data(), sizeof(size_t)*num_rows).wait();
}

for (size_t row = 0; row < num_rows; ++row) {
ASSERT_EQ(row_indices_desired_host[row], row_indices_host[row]);
}

}

TEST(SyclHistUpdater, Sampling) {
xgboost::tree::TrainParam param;
param.UpdateAllowUnknown(Args{{"subsample", "0.7"}});
Expand Down Expand Up @@ -432,12 +522,24 @@ TEST(SyclHistUpdater, EvaluateSplits) {
TestHistUpdaterEvaluateSplits<double>(param);
}

TEST(SyclHistUpdater, ApplySplit) {
TEST(SyclHistUpdater, ApplySplitSparce) {
xgboost::tree::TrainParam param;
param.UpdateAllowUnknown(Args{{"max_depth", "3"}});

TestHistUpdaterApplySplit<float>(param, 0.3, 256);
TestHistUpdaterApplySplit<double>(param, 0.3, 256);
}

TEST(SyclHistUpdater, ApplySplitDence) {
xgboost::tree::TrainParam param;
param.UpdateAllowUnknown(Args{{"max_depth", "3"}});

// TestHistUpdaterApplySplit<float>(param);
// TestHistUpdaterApplySplit<double>(param);
TestHistUpdaterApplySplit<float>(param, 0.0, 256);
TestHistUpdaterApplySplit<float>(param, 0.0, 256+1);
TestHistUpdaterApplySplit<float>(param, 0.0, (1u << 16) + 1);
TestHistUpdaterApplySplit<double>(param, 0.0, 256);
TestHistUpdaterApplySplit<double>(param, 0.0, 256+1);
TestHistUpdaterApplySplit<double>(param, 0.0, (1u << 16) + 1);
}

} // namespace xgboost::sycl::tree

0 comments on commit 8743b00

Please sign in to comment.