Interface TrainerContext
-
- All Known Implementing Classes:
DefaultTrainerContext,SymmetricTrainerContext
public interface TrainerContext
-
-
Method Summary
All Methods Instance Methods Abstract Methods Modifier and Type Method Description Trainercreate(String uuid, int threadId, org.deeplearning4j.nn.api.Model model, int rootDevice, boolean useMDS, ParallelWrapper wrapper, org.deeplearning4j.nn.conf.WorkspaceMode workspaceMode, int averagingFrequency)Create aTrainerbased on the given parametersvoidfinalizeRound(org.deeplearning4j.nn.api.Model originalModel, org.deeplearning4j.nn.api.Model... models)This method is called at averagingFrequencyvoidfinalizeTraining(org.deeplearning4j.nn.api.Model originalModel, org.deeplearning4j.nn.api.Model... models)This method is calledvoidinit(org.deeplearning4j.nn.api.Model model, Object... args)Initialize the context
-
-
-
Method Detail
-
init
void init(org.deeplearning4j.nn.api.Model model, Object... args)Initialize the context- Parameters:
model-args- the arguments to initialize with (maybe null)
-
create
Trainer create(String uuid, int threadId, org.deeplearning4j.nn.api.Model model, int rootDevice, boolean useMDS, ParallelWrapper wrapper, org.deeplearning4j.nn.conf.WorkspaceMode workspaceMode, int averagingFrequency)
Create aTrainerbased on the given parameters- Parameters:
threadId- the thread id to use for this workermodel- the model to start the trainer withrootDevice- the root device iduseMDS- whether to use MultiDataSet or DataSet or notwrapper- the wrapper instance to use with this trainer (this refernece is needed for coordination with theParallelWrapper'sTrainingListener- Returns:
- the created training instance
-
finalizeRound
void finalizeRound(org.deeplearning4j.nn.api.Model originalModel, org.deeplearning4j.nn.api.Model... models)This method is called at averagingFrequency- Parameters:
originalModel-models-
-
finalizeTraining
void finalizeTraining(org.deeplearning4j.nn.api.Model originalModel, org.deeplearning4j.nn.api.Model... models)This method is called- Parameters:
originalModel-models-
-
-