onnxruntime/winml/lib/Api.Experimental/LearningModelBuilder.cpp

#include "lib/Api.Experimental/pch/pch.h"
#include "LearningModelBuilder.h"
#include "LearningModel.h"
#include "TensorFeatureDescriptor.h"
#include "LearningModelSession.h"
#include "LearningModelInputs.h"
#include "LearningModelOutputs.h"
#include "LearningModelOperatorSet.h"
#include "OnnxruntimeProvider.h"

namespace WINML_EXPERIMENTALP {

LearningModelBuilder::LearningModelBuilder(int64_t opset)
  : inert_session_(nullptr),
    inputs_(nullptr),
    outputs_(nullptr),
    operators_(nullptr) {
  telemetry_helper.LogApiUsage("LearningModelBuilder::LearningModelBuilder");

  WINML_THROW_IF_FAILED(CreateOnnxruntimeEngineFactory(engine_factory_.put()));
  WINML_THROW_IF_FAILED(engine_factory_->CreateEmptyModel(opset, model_.put()));
  inputs_ = winrt::make<winml_experimentalp::LearningModelInputs>(*this);
  outputs_ = winrt::make<winml_experimentalp::LearningModelOutputs>(*this);
  operators_ = winrt::make<winml_experimentalp::LearningModelOperatorSet>(*this);

  winrt::com_ptr<_winml::IEngineBuilder> builder;
  WINML_THROW_IF_FAILED(engine_factory_->CreateEngineBuilder(builder.put()));
  winrt::com_ptr<_winml::IEngine> engine;
  WINML_THROW_IF_FAILED(builder->CreateEngine(engine.put()));
  inert_session_ = winmlp::LearningModelSession::CreateInertSession(engine.get());
}

LearningModelBuilder::LearningModelBuilder(LearningModelBuilder& builder)
  : inert_session_(nullptr),
    inputs_(builder.inputs_),
    outputs_(builder.outputs_),
    operators_(builder.operators_) {
}

winml_experimental::LearningModelInputs LearningModelBuilder::Inputs() {
  return inputs_;
}

winml_experimental::LearningModelOutputs LearningModelBuilder::Outputs() {
  return outputs_;
}

winml_experimental::LearningModelOperatorSet LearningModelBuilder::Operators() {
  return operators_;
}

winml::LearningModel LearningModelBuilder::CreateModel() {
  telemetry_helper.LogApiUsage("LearningModelBuilder::CreateModel");

  com_ptr<_winml::IModel> model_clone;
  model_->CloneModel(model_clone.put());
  return winrt::make<winmlp::LearningModel>(engine_factory_.get(), model_clone.get(), nullptr);
}

void LearningModelBuilder::Save(const winrt::hstring& file_name) {
  telemetry_helper.LogApiUsage("LearningModelBuilder::Save");

  model_->SaveModel(file_name.c_str(), file_name.size());
}

winml_experimental::LearningModelBuilder LearningModelBuilder::Create(int32_t opset) {
  return winrt::make<LearningModelBuilder>(static_cast<int64_t>(opset));
}

winml::TensorFeatureDescriptor LearningModelBuilder::CreateTensorFeatureDescriptor(
  hstring const& name, winml::TensorKind const& kind, array_view<int64_t const> shape
) {
  return winrt::make<winmlp::TensorFeatureDescriptor>(name, L"", kind, shape);
}

winml::TensorFeatureDescriptor LearningModelBuilder::CreateTensorFeatureDescriptor(
  hstring const& name, hstring const& description, winml::TensorKind const& kind, array_view<int64_t const> shape
) {
  return winrt::make<winmlp::TensorFeatureDescriptor>(name, description, kind, shape);
}

_winml::IModel* LearningModelBuilder::UseModel() {
  return model_.get();
}

}  // namespace WINML_EXPERIMENTALP