org.tensorflow.op.nn

Class CtcLoss<T extends Number>

java.lang.Object

org.tensorflow.op.PrimitiveOp

org.tensorflow.op.nn.CtcLoss<T>

Type Parameters:

T - data type for loss() output

All Implemented Interfaces:

Op

@Operator(group="nn")
public final class CtcLoss<T extends Number>
extends PrimitiveOp

Calculates the CTC Loss (log probability) for each batch entry. Also calculates

the gradient. This class performs the softmax operation for you, so inputs should be e.g. linear projections of outputs by an LSTM.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	CtcLoss.Options Optional attributes for CtcLoss

Field Summary

Fields inherited from class org.tensorflow.op.PrimitiveOp

operation

Method Summary

Modifier and Type	Method and Description
static <T extends Number> CtcLoss<T>	create(Scope scope, Operand<T> inputs, Operand<Long> labelsIndices, Operand<Integer> labelsValues, Operand<Integer> sequenceLength, CtcLoss.Options... options) Factory method to create a class wrapping a new CtcLoss operation.
static CtcLoss.Options	ctcMergeRepeated(Boolean ctcMergeRepeated)
Output<T>	gradient() The gradient of `loss`.
static CtcLoss.Options	ignoreLongerOutputsThanInputs(Boolean ignoreLongerOutputsThanInputs)
Output<T>	loss() A vector (batch) containing log-probabilities.
static CtcLoss.Options	preprocessCollapseRepeated(Boolean preprocessCollapseRepeated)

Methods inherited from class org.tensorflow.op.PrimitiveOp

equals, hashCode, op, toString

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Method Detail

create

public static <T extends Number> CtcLoss<T> create(Scope scope,
                                                   Operand<T> inputs,
                                                   Operand<Long> labelsIndices,
                                                   Operand<Integer> labelsValues,
                                                   Operand<Integer> sequenceLength,
                                                   CtcLoss.Options... options)

Factory method to create a class wrapping a new CtcLoss operation.

Parameters:

scope - current scope

inputs - 3-D, shape: `(max_time x batch_size x num_classes)`, the logits.

labelsIndices - The indices of a `SparseTensor`. `labels_indices(i, :) == [b, t]` means `labels_values(i)` stores the id for `(batch b, time t)`.

labelsValues - The values (labels) associated with the given batch and time.

sequenceLength - A vector containing sequence lengths (batch).

options - carries optional attributes values

Returns:

a new instance of CtcLoss

preprocessCollapseRepeated

public static CtcLoss.Options preprocessCollapseRepeated(Boolean preprocessCollapseRepeated)

Parameters:

preprocessCollapseRepeated - Scalar, if true then repeated labels are collapsed prior to the CTC calculation.

ctcMergeRepeated

public static CtcLoss.Options ctcMergeRepeated(Boolean ctcMergeRepeated)

Parameters:

ctcMergeRepeated - Scalar. If set to false, during CTC calculation repeated non-blank labels will not be merged and are interpreted as individual labels. This is a simplified version of CTC.

ignoreLongerOutputsThanInputs

public static CtcLoss.Options ignoreLongerOutputsThanInputs(Boolean ignoreLongerOutputsThanInputs)

Parameters:

ignoreLongerOutputsThanInputs - Scalar. If set to true, during CTC calculation, items that have longer output sequences than input sequences are skipped: they don't contribute to the loss term and have zero-gradient.

loss

public Output<T> loss()

A vector (batch) containing log-probabilities.

gradient

public Output<T> gradient()

The gradient of `loss`. 3-D, shape: `(max_time x batch_size x num_classes)`.