org.nd4j.autodiff.samediff.internal

Class DataTypesSession

java.lang.Object

org.nd4j.autodiff.samediff.internal.AbstractSession<DataType,DataTypesSession.DataTypeCalc>

org.nd4j.autodiff.samediff.internal.DataTypesSession

public class DataTypesSession
extends AbstractSession<DataType,DataTypesSession.DataTypeCalc>

Infer datatypes for all variables. Optionally update the datatypes of variables as we go

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
protected static class	DataTypesSession.DataTypeCalc

Nested classes/interfaces inherited from class org.nd4j.autodiff.samediff.internal.AbstractSession

AbstractSession.FrameIter, AbstractSession.VarId

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	dynamicUpdate

Fields inherited from class org.nd4j.autodiff.samediff.internal.AbstractSession

availableForExec, availableForExecSet, execConstInputs, execInputs, execInputsAllIter, frameParents, nodeOutputs, OUTER_FRAME, sameDiff, subgraph, tensorArrays

Constructor Summary

Constructors

Constructor and Description
DataTypesSession(SameDiff sameDiff, boolean dynamicUpdate)

Method Summary

Modifier and Type	Method and Description
DataTypesSession.DataTypeCalc	getAndParameterizeOp(String opName, AbstractSession.FrameIter frameIter, Set<AbstractSession.VarId> inputs, Set<AbstractSession.VarId> allIterInputs, Set<String> constAndPhInputs, Map<String,DataType> placeholderValues) Get the parameterized op to execute - for example, the op/DifferentialFunction with all inputs set
DataType	getConstantOrVariable(String variableName) Get the constant or variable output - for example, constant array or constant shape.
DataType[]	getOutputs(DataTypesSession.DataTypeCalc op, AbstractSession.FrameIter outputFrameIter, Set<AbstractSession.VarId> inputs, Set<AbstractSession.VarId> allIterInputs, Set<String> constAndPhInputs, List<Listener> listeners, At at, MultiDataSet batch) Execute the op - calculate INDArrays, or shape info, etc

Methods inherited from class org.nd4j.autodiff.samediff.internal.AbstractSession

addToExecInputs, allInputsAvailable, contains, get, get, initSubgraph, lookup, newVarId, newVarId, output, output, preprocessPlaceholders, updateDescendentsForExec

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

dynamicUpdate

protected boolean dynamicUpdate

Constructor Detail

DataTypesSession

public DataTypesSession(SameDiff sameDiff,
                        boolean dynamicUpdate)

Parameters:

sameDiff - SameDiff instance

dynamicUpdate - If true: Dynamically update the datatypes as we go

Method Detail

getConstantOrVariable

public DataType getConstantOrVariable(String variableName)

Description copied from class: AbstractSession

Get the constant or variable output - for example, constant array or constant shape. Note that both constants and variables (i.e., VariableType.CONSTANT and VariableType.VARIABLE) are the same for all frames and iterations.

Specified by:

getConstantOrVariable in class AbstractSession<DataType,DataTypesSession.DataTypeCalc>

Parameters:

variableName - The name of the variable to get the constant for

Returns:

The constant

getAndParameterizeOp

public DataTypesSession.DataTypeCalc getAndParameterizeOp(String opName,
                                                          AbstractSession.FrameIter frameIter,
                                                          Set<AbstractSession.VarId> inputs,
                                                          Set<AbstractSession.VarId> allIterInputs,
                                                          Set<String> constAndPhInputs,
                                                          Map<String,DataType> placeholderValues)

Description copied from class: AbstractSession

Get the parameterized op to execute - for example, the op/DifferentialFunction with all inputs set

Specified by:

getAndParameterizeOp in class AbstractSession<DataType,DataTypesSession.DataTypeCalc>

Parameters:

opName - Name of the op

frameIter - The frame and iteration of the op outputs

inputs - The inputs to the op (excluding constants/placeholders) - for the specific frame + iteration

allIterInputs - The inputs - those that are not iteration-specific (mainly Enter op vars, which might be used in all iterations but are only executed once on iter 0)

constAndPhInputs - The constant and placeholder inputs - used for all frames/iterations

Returns:

The parameterized op

getOutputs

public DataType[] getOutputs(DataTypesSession.DataTypeCalc op,
                             AbstractSession.FrameIter outputFrameIter,
                             Set<AbstractSession.VarId> inputs,
                             Set<AbstractSession.VarId> allIterInputs,
                             Set<String> constAndPhInputs,
                             List<Listener> listeners,
                             At at,
                             MultiDataSet batch)

Description copied from class: AbstractSession

Execute the op - calculate INDArrays, or shape info, etc

Specified by:

getOutputs in class AbstractSession<DataType,DataTypesSession.DataTypeCalc>

Parameters:

op - Operation to exit. This should be parameterized (i.e., all inputs set)

outputFrameIter - The frame and iteration of the outputs

inputs - The specific input arrays for the op

Returns:

The outputs of the op