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

Refactor configuration [Part II]. (#4577)

Browse Source 
* Refactor configuration [Part II].

* General changes:
** Remove `Init` methods to avoid ambiguity.
** Remove `Configure(std::map<>)` to avoid redundant copying and prepare for
   parameter validation. (`std::vector` is returned from `InitAllowUnknown`).
** Add name to tree updaters for easier debugging.

* Learner changes:
** Make `LearnerImpl` the only source of configuration.

    All configurations are stored and carried out by `LearnerImpl::Configure()`.

** Remove booster in C API.

    Originally kept for "compatibility reason", but did not state why.  So here
    we just remove it.

** Add a `metric_names_` field in `LearnerImpl`.
** Remove `LazyInit`.  Configuration will always be lazy.
** Run `Configure` before every iteration.

* Predictor changes:
** Allocate both cpu and gpu predictor.
** Remove cpu_predictor from gpu_predictor.

    `GBTree` is now used to dispatch the predictor.

** Remove some GPU Predictor tests.

* IO

No IO changes.  The binary model format stability is tested by comparing
hashing value of save models between two commits
This commit is contained in:
Jiaming Yuan
2019-07-20 08:34:56 -04:00
committed by GitHub
parent ad1192e8a3
commit f0064c07ab
69 changed files with 669 additions and 761 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
src/predictor/gpu_predictor.cu
View File

@@ -377,8 +377,7 @@ class GPUPredictor : public xgboost::Predictor {
}
public:
GPUPredictor() // NOLINT
: cpu_predictor_(Predictor::Create("cpu_predictor", learner_param_)) {}
GPUPredictor() = default;
void PredictBatch(DMatrix* dmat, HostDeviceVector<bst_float>* out_preds,
const gbm::GBTreeModel& model, int tree_begin,
@@ -468,12 +467,15 @@ class GPUPredictor : public xgboost::Predictor {
std::vector<bst_float>* out_preds,
const gbm::GBTreeModel& model, unsigned ntree_limit,
unsigned root_index) override {
cpu_predictor_->PredictInstance(inst, out_preds, model, root_index);
LOG(FATAL) << "Internal error: " << __func__
<< " is not implemented in GPU Predictor.";
}
void PredictLeaf(DMatrix* p_fmat, std::vector<bst_float>* out_preds,
const gbm::GBTreeModel& model,
unsigned ntree_limit) override {
cpu_predictor_->PredictLeaf(p_fmat, out_preds, model, ntree_limit);
LOG(FATAL) << "Internal error: " << __func__
<< " is not implemented in GPU Predictor.";
}
void PredictContribution(DMatrix* p_fmat,
@@ -481,9 +483,8 @@ class GPUPredictor : public xgboost::Predictor {
const gbm::GBTreeModel& model, unsigned ntree_limit,
bool approximate, int condition,
unsigned condition_feature) override {
cpu_predictor_->PredictContribution(p_fmat, out_contribs, model, ntree_limit,
approximate, condition,
condition_feature);
LOG(FATAL) << "Internal error: " << __func__
<< " is not implemented in GPU Predictor.";
}
void PredictInteractionContributions(DMatrix* p_fmat,
@@ -491,14 +492,13 @@ class GPUPredictor : public xgboost::Predictor {
const gbm::GBTreeModel& model,
unsigned ntree_limit,
bool approximate) override {
cpu_predictor_->PredictInteractionContributions(p_fmat, out_contribs, model,
ntree_limit, approximate);
LOG(FATAL) << "Internal error: " << __func__
<< " is not implemented in GPU Predictor.";
}
void Init(const std::vector<std::pair<std::string, std::string>>& cfg,
const std::vector<std::shared_ptr<DMatrix>>& cache) override {
Predictor::Init(cfg, cache);
cpu_predictor_->Init(cfg, cache);
void Configure(const std::vector<std::pair<std::string, std::string>>& cfg,
const std::vector<std::shared_ptr<DMatrix>>& cache) override {
Predictor::Configure(cfg, cache);
GPUSet devices = GPUSet::All(learner_param_->gpu_id, learner_param_->n_gpus);
ConfigureShards(devices);
@@ -517,7 +517,6 @@ class GPUPredictor : public xgboost::Predictor {
});
}
std::unique_ptr<Predictor> cpu_predictor_;
std::vector<DeviceShard> shards_;
GPUSet devices_;
common::Monitor monitor_;