st_model_trainer.hpp File Reference

#include <ml/neural_net/combine.hpp>
#include <ml/neural_net/compute_context.hpp>
#include <ml/neural_net/model_spec.hpp>
#include <toolkits/style_transfer/style_transfer_data_iterator.hpp>

Go to the source code of this file.

Classes
struct	turi::style_transfer::DataBatch
struct	turi::style_transfer::EncodedBatch
struct	turi::style_transfer::EncodedInferenceBatch
struct	turi::style_transfer::TrainingProgress
struct	turi::style_transfer::Config
class	turi::style_transfer::DataIterator
class	turi::style_transfer::InferenceDataIterator
class	turi::style_transfer::ProgressUpdater
class	turi::style_transfer::Checkpoint
class	turi::style_transfer::ModelTrainer

Namespaces
	turi
	SKD.

Functions
EncodedBatch	turi::style_transfer::EncodeTrainingBatch (DataBatch batch, int width, int height)
EncodedInferenceBatch	turi::style_transfer::EncodeInferenceBatch (DataBatch batch)
DataBatch	turi::style_transfer::DecodeInferenceBatch (EncodedInferenceBatch batch)

Detailed Description

Defines the value types representing each stage of a style-transfer training pipeline, and the virtual interface for arbitrary style-transfer models.

Definition in file st_model_trainer.hpp.

Function Documentation

DecodeInferenceBatch()

DataBatch turi::style_transfer::DecodeInferenceBatch

(

EncodedInferenceBatch

batch

)

Converts the raw output from an inference backend into images.

EncodeInferenceBatch()

EncodedInferenceBatch turi::style_transfer::EncodeInferenceBatch

(

DataBatch

batch

)

Converts native images into tensors that can be fed into the model backend.

EncodeTrainingBatch()

EncodedBatch turi::style_transfer::EncodeTrainingBatch	(	DataBatch	batch,
		int	width,
		int	height
	)

Converts native images into tensors that can be fed into the model backend.