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

Avoid thread block with sparse data. (#7255)

Browse Source
This commit is contained in:
Jiaming Yuan
2021-09-25 13:11:34 +08:00
committed by GitHub
parent ca17f8a5fc
commit d8a549e6ac
6 changed files with 97 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
src/predictor/cpu_predictor.cc
View File

@@ -253,17 +253,32 @@ class CPUPredictor : public Predictor {
gbm::GBTreeModel const &model, int32_t tree_begin,
int32_t tree_end) const {
const int threads = omp_get_max_threads();
constexpr double kDensityThresh = .5;
size_t total = std::max(p_fmat->Info().num_row_ * p_fmat->Info().num_col_,
static_cast<uint64_t>(1));
double density = static_cast<double>(p_fmat->Info().num_nonzero_) /
static_cast<double>(total);
bool blocked = density > kDensityThresh;
std::vector<RegTree::FVec> feat_vecs;
InitThreadTemp(threads * kBlockOfRowsSize,
InitThreadTemp(threads * (blocked ? kBlockOfRowsSize : 1),
model.learner_model_param->num_feature, &feat_vecs);
for (auto const& batch : p_fmat->GetBatches<SparsePage>()) {
for (auto const &batch : p_fmat->GetBatches<SparsePage>()) {
CHECK_EQ(out_preds->size(),
p_fmat->Info().num_row_ * model.learner_model_param->num_output_group);
p_fmat->Info().num_row_ *
model.learner_model_param->num_output_group);
size_t constexpr kUnroll = 8;
PredictBatchByBlockOfRowsKernel<SparsePageView<kUnroll>,
kBlockOfRowsSize>(SparsePageView<kUnroll>{&batch},
out_preds, model, tree_begin,
tree_end, &feat_vecs);
if (blocked) {
PredictBatchByBlockOfRowsKernel<SparsePageView<kUnroll>,
kBlockOfRowsSize>(
SparsePageView<kUnroll>{&batch}, out_preds, model, tree_begin,
tree_end, &feat_vecs);
} else {
PredictBatchByBlockOfRowsKernel<SparsePageView<kUnroll>, 1>(
SparsePageView<kUnroll>{&batch}, out_preds, model, tree_begin,
tree_end, &feat_vecs);
}
}
}
@@ -316,7 +331,7 @@ class CPUPredictor : public Predictor {
tree_end);
}
template <typename Adapter>
template <typename Adapter, size_t kBlockSize>
void DispatchedInplacePredict(dmlc::any const &x, std::shared_ptr<DMatrix> p_m,
const gbm::GBTreeModel &model, float missing,
PredictionCacheEntry *out_preds,
@@ -336,9 +351,9 @@ class CPUPredictor : public Predictor {
std::vector<Entry> workspace(m->NumColumns() * 8 * threads);
auto &predictions = out_preds->predictions.HostVector();
std::vector<RegTree::FVec> thread_temp;
InitThreadTemp(threads * kBlockOfRowsSize,
model.learner_model_param->num_feature, &thread_temp);
PredictBatchByBlockOfRowsKernel<AdapterView<Adapter>, kBlockOfRowsSize>(
InitThreadTemp(threads * kBlockSize, model.learner_model_param->num_feature,
&thread_temp);
PredictBatchByBlockOfRowsKernel<AdapterView<Adapter>, kBlockSize>(
AdapterView<Adapter>(m.get(), missing, common::Span<Entry>{workspace}),
&predictions, model, tree_begin, tree_end, &thread_temp);
}
@@ -348,16 +363,16 @@ class CPUPredictor : public Predictor {
PredictionCacheEntry *out_preds, uint32_t tree_begin,
unsigned tree_end) const override {
if (x.type() == typeid(std::shared_ptr<data::DenseAdapter>)) {
this->DispatchedInplacePredict<data::DenseAdapter>(
this->DispatchedInplacePredict<data::DenseAdapter, kBlockOfRowsSize>(
x, p_m, model, missing, out_preds, tree_begin, tree_end);
} else if (x.type() == typeid(std::shared_ptr<data::CSRAdapter>)) {
this->DispatchedInplacePredict<data::CSRAdapter>(
this->DispatchedInplacePredict<data::CSRAdapter, 1>(
x, p_m, model, missing, out_preds, tree_begin, tree_end);
} else if (x.type() == typeid(std::shared_ptr<data::ArrayAdapter>)) {
this->DispatchedInplacePredict<data::ArrayAdapter> (
this->DispatchedInplacePredict<data::ArrayAdapter, kBlockOfRowsSize> (
x, p_m, model, missing, out_preds, tree_begin, tree_end);
} else if (x.type() == typeid(std::shared_ptr<data::CSRArrayAdapter>)) {
this->DispatchedInplacePredict<data::CSRArrayAdapter> (
this->DispatchedInplacePredict<data::CSRArrayAdapter, 1> (
x, p_m, model, missing, out_preds, tree_begin, tree_end);
} else {
return false;