CTCLossV2

public final class CTCLossV2

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.

class CTCLossV2.Options Optional attributes for CTCLossV2

String OP_NAME The name of this op, as known by TensorFlow core engine

static CTCLossV2	create(Scope scope, Operand<TFloat32> inputs, Operand<TInt64> labelsIndices, Operand<TInt32> labelsValues, Operand<TInt32> sequenceLength, Options... options) Factory method to create a class wrapping a new CTCLossV2 operation.
static CTCLossV2.Options	ctcMergeRepeated(Boolean ctcMergeRepeated)
Output<TFloat32>	gradient() The gradient of `loss`.
static CTCLossV2.Options	ignoreLongerOutputsThanInputs(Boolean ignoreLongerOutputsThanInputs)
Output<TFloat32>	loss() A vector (batch) containing log-probabilities.
static CTCLossV2.Options	preprocessCollapseRepeated(Boolean preprocessCollapseRepeated)

From class org.tensorflow.op.RawOp

final boolean	equals(Object obj)
final int	hashCode()
Operation	op() Return this unit of computation as a single `Operation`.
final String	toString()

From class java.lang.Object

From interface org.tensorflow.op.Op

abstract ExecutionEnvironment	env() Return the execution environment this op was created in.
abstract Operation	op() Return this unit of computation as a single `Operation`.

Constants

The name of this op, as known by TensorFlow core engine

Constant Value: "CTCLossV2"

Factory method to create a class wrapping a new CTCLossV2 operation.

scope	current scope
inputs	3-D, shape: `(max_time x batch_size x num_classes)`, the logits. Default blank label is 0 rather num_classes - 1.
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

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.

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

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.

A vector (batch) containing log-probabilities.

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