turi::style_transfer::ResNet16ModelTrainer Class Reference

#include <toolkits/style_transfer/st_resnet16_model_trainer.hpp>

Public Member Functions
	ResNet16ModelTrainer (Config config, neural_net::float_array_map weights)
bool	SupportsLossComponents () const override
std::shared_ptr< neural_net::Publisher< std::unique_ptr< Checkpoint > > >	AsCheckpointPublisher () override
virtual std::shared_ptr< neural_net::Publisher< TrainingProgress > >	AsTrainingBatchPublisher (std::unique_ptr< data_iterator > training_data, const std::string &vgg_mlmodel_path, int offset, std::unique_ptr< float > initial_training_loss, neural_net::compute_context *context)
virtual std::shared_ptr< neural_net::Publisher< DataBatch > >	AsInferenceBatchPublisher (std::unique_ptr< data_iterator > test_data, std::vector< int > style_indices, neural_net::compute_context *context)

Detailed Description

Subclass of ModelTrainer encapsulating the resnet-16 architecture.

Definition at line 35 of file st_resnet16_model_trainer.hpp.

Constructor & Destructor Documentation

ResNet16ModelTrainer()

turi::style_transfer::ResNet16ModelTrainer::ResNet16ModelTrainer	(	Config	config,
		neural_net::float_array_map	weights
	)

Initializes a model from a checkpoint.

Member Function Documentation

AsCheckpointPublisher()

std::shared_ptr<neural_net::Publisher<std::unique_ptr<Checkpoint> > > turi::style_transfer::ResNet16ModelTrainer::AsCheckpointPublisher ( )

overridevirtual

Returns a publisher that can be used to request checkpoints.

Implements turi::style_transfer::ModelTrainer.

AsInferenceBatchPublisher()

virtual std::shared_ptr<neural_net::Publisher<DataBatch> > turi::style_transfer::ModelTrainer::AsInferenceBatchPublisher ( std::unique_ptr< data_iterator >  test_data, std::vector< int >  style_indices, neural_net::compute_context *  context  )

virtualinherited

Given a data iterator, return a publisher of inference model outputs.

AsTrainingBatchPublisher()

virtual std::shared_ptr<neural_net::Publisher<TrainingProgress> > turi::style_transfer::ModelTrainer::AsTrainingBatchPublisher ( std::unique_ptr< data_iterator >  training_data, const std::string &  vgg_mlmodel_path, int  offset, std::unique_ptr< float >  initial_training_loss, neural_net::compute_context *  context  )

virtualinherited

Given a data iterator, return a publisher of training model outputs.

SupportsLossComponents()

bool turi::style_transfer::ResNet16ModelTrainer::SupportsLossComponents ( ) const

inlineoverridevirtual

Returns true iff the output from the training batch publisher sets the style_loss and content_loss values.

Implements turi::style_transfer::ModelTrainer.

Definition at line 42 of file st_resnet16_model_trainer.hpp.

---

The documentation for this class was generated from the following file:

• toolkits/style_transfer/st_resnet16_model_trainer.hpp