Op

public interface Op

Known Indirect Subclasses Abort, Abs<T extends TNumber>, AccumulateN<T extends TType>, AccumulatorApplyGradient, AccumulatorNumAccumulated, AccumulatorSetGlobalStep, AccumulatorTakeGradient<T extends TType>, Acos<T extends TType>, Acosh<T extends TType>, Add<T extends TType>, AddManySparseToTensorsMap, AddN<T extends TType>, AddSparseToTensorsMap, AdjustContrast<T extends TNumber>, AdjustHue<T extends TNumber>, and 1157 others. Abort Raise a exception to abort the process when called.  Abs<T extends TNumber> Computes the absolute value of a tensor.  AccumulateN<T extends TType> Returns the element-wise sum of a list of tensors.  AccumulatorApplyGradient Applies a gradient to a given accumulator.  AccumulatorNumAccumulated Returns the number of gradients aggregated in the given accumulators.  AccumulatorSetGlobalStep Updates the accumulator with a new value for global_step.  AccumulatorTakeGradient<T extends TType> Extracts the average gradient in the given ConditionalAccumulator.  Acos<T extends TType> Computes acos of x element-wise.  Acosh<T extends TType> Computes inverse hyperbolic cosine of x element-wise.  Add<T extends TType> Returns x + y element-wise.  AddManySparseToTensorsMap Add an `N`-minibatch `SparseTensor` to a `SparseTensorsMap`, return `N` handles.  AddN<T extends TType> Add all input tensors element wise.  AddSparseToTensorsMap Add a `SparseTensor` to a `SparseTensorsMap` return its handle.  AdjustContrast<T extends TNumber> Adjust the contrast of one or more images.  AdjustHue<T extends TNumber> Adjust the hue of one or more images.  AdjustSaturation<T extends TNumber> Adjust the saturation of one or more images.  All Computes the "logical and" of elements across dimensions of a tensor.  AllCandidateSampler Generates labels for candidate sampling with a learned unigram distribution.  AllReduce<T extends TNumber> Mutually reduces multiple tensors of identical type and shape.  AllToAll<T extends TType> An Op to exchange data across TPU replicas.  Angle<U extends TNumber> Returns the argument of a complex number.  AnonymousIterator A container for an iterator resource.  AnonymousMemoryCache   AnonymousMultiDeviceIterator A container for a multi device iterator resource.  AnonymousRandomSeedGenerator   AnonymousSeedGenerator   Any Computes the "logical or" of elements across dimensions of a tensor.  ApplyAdaMax<T extends TType> Update '*var' according to the AdaMax algorithm.  ApplyAdadelta<T extends TType> Update '*var' according to the adadelta scheme.  ApplyAdagrad<T extends TType> Update '*var' according to the adagrad scheme.  ApplyAdagradDa<T extends TType> Update '*var' according to the proximal adagrad scheme.  ApplyAdagradV2<T extends TType> Update '*var' according to the adagrad scheme.  ApplyAdam<T extends TType> Update '*var' according to the Adam algorithm.  ApplyAddSign<T extends TType> Update '*var' according to the AddSign update.  ApplyCenteredRmsProp<T extends TType> Update '*var' according to the centered RMSProp algorithm.  ApplyFtrl<T extends TType> Update '*var' according to the Ftrl-proximal scheme.  ApplyGradientDescent<T extends TType> Update '*var' by subtracting 'alpha' * 'delta' from it.  ApplyMomentum<T extends TType> Update '*var' according to the momentum scheme.  ApplyPowerSign<T extends TType> Update '*var' according to the AddSign update.  ApplyProximalAdagrad<T extends TType> Update '*var' and '*accum' according to FOBOS with Adagrad learning rate.  ApplyProximalGradientDescent<T extends TType> Update '*var' as FOBOS algorithm with fixed learning rate.  ApplyRmsProp<T extends TType> Update '*var' according to the RMSProp algorithm.  ApproximateEqual Returns the truth value of abs(x-y) < tolerance element-wise.  ArgMax<V extends TNumber> Returns the index with the largest value across dimensions of a tensor.  ArgMin<V extends TNumber> Returns the index with the smallest value across dimensions of a tensor.  AsString Converts each entry in the given tensor to strings.  Asin<T extends TType> Computes the trignometric inverse sine of x element-wise.  Asinh<T extends TType> Computes inverse hyperbolic sine of x element-wise.  AssertCardinalityDataset   AssertNextDataset   AssertThat Asserts that the given condition is true.  Assign<T extends TType> Update 'ref' by assigning 'value' to it.  AssignAdd<T extends TType> Update 'ref' by adding 'value' to it.  AssignAddVariableOp Adds a value to the current value of a variable.  AssignSub<T extends TType> Update 'ref' by subtracting 'value' from it.  AssignSubVariableOp Subtracts a value from the current value of a variable.  AssignVariableOp Assigns a new value to a variable.  Atan<T extends TType> Computes the trignometric inverse tangent of x element-wise.  Atan2<T extends TNumber> Computes arctangent of `y/x` element-wise, respecting signs of the arguments.  Atanh<T extends TType> Computes inverse hyperbolic tangent of x element-wise.  AudioSpectrogram Produces a visualization of audio data over time.  AudioSummary Outputs a `Summary` protocol buffer with audio.  AutoShardDataset Creates a dataset that shards the input dataset.  AvgPool<T extends TNumber> Performs average pooling on the input.  AvgPool3d<T extends TNumber> Performs 3D average pooling on the input.  AvgPool3dGrad<T extends TNumber> Computes gradients of average pooling function.  AvgPoolGrad<T extends TNumber> Computes gradients of the average pooling function.  BandPart<T extends TType> Copy a tensor setting everything outside a central band in each innermost matrix to zero.  BandedTriangularSolve<T extends TType>   Barrier Defines a barrier that persists across different graph executions.  BarrierClose Closes the given barrier.  BarrierIncompleteSize Computes the number of incomplete elements in the given barrier.  BarrierInsertMany For each key, assigns the respective value to the specified component.  BarrierReadySize Computes the number of complete elements in the given barrier.  BarrierTakeMany Takes the given number of completed elements from a barrier.  Batch Batches all input tensors nondeterministically.  BatchCholesky<T extends TNumber>   BatchCholeskyGrad<T extends TNumber>   BatchDataset Creates a dataset that batches `batch_size` elements from `input_dataset`.  BatchFft   BatchFft2d   BatchFft3d   BatchIfft   BatchIfft2d   BatchIfft3d   BatchMatMul<T extends TType> Multiplies slices of two tensors in batches.  BatchMatrixBandPart<T extends TType>   BatchMatrixDeterminant<T extends TType>   BatchMatrixDiag<T extends TType>   BatchMatrixDiagPart<T extends TType>   BatchMatrixInverse<T extends TNumber>   BatchMatrixSetDiag<T extends TType>   BatchMatrixSolve<T extends TNumber>   BatchMatrixSolveLs<T extends TNumber>   BatchMatrixTriangularSolve<T extends TNumber>   BatchNormWithGlobalNormalization<T extends TType> Batch normalization.  BatchNormWithGlobalNormalizationGrad<T extends TType> Gradients for batch normalization.  BatchSelfAdjointEig<T extends TNumber>   BatchSvd<T extends TType>   BatchToSpace<T extends TType> BatchToSpace for 4-D tensors of type T.  BatchToSpaceNd<T extends TType> BatchToSpace for N-D tensors of type T.  BesselI0<T extends TNumber>   BesselI0e<T extends TNumber>   BesselI1<T extends TNumber>   BesselI1e<T extends TNumber>   BesselJ0<T extends TNumber>   BesselJ1<T extends TNumber>   BesselK0<T extends TNumber>   BesselK0e<T extends TNumber>   BesselK1<T extends TNumber>   BesselK1e<T extends TNumber>   BesselY0<T extends TNumber>   BesselY1<T extends TNumber>   Betainc<T extends TNumber> Compute the regularized incomplete beta integral \\(I_x(a, b)\\).  BiasAdd<T extends TType> Adds `bias` to `value`.  BiasAddGrad<T extends TType> The backward operation for "BiasAdd" on the "bias" tensor.  Bincount<T extends TNumber> Counts the number of occurrences of each value in an integer array.  Bitcast<U extends TType> Bitcasts a tensor from one type to another without copying data.  BitwiseAnd<T extends TNumber> Elementwise computes the bitwise AND of `x` and `y`.  BitwiseOr<T extends TNumber> Elementwise computes the bitwise OR of `x` and `y`.  BitwiseXor<T extends TNumber> Elementwise computes the bitwise XOR of `x` and `y`.  BlockLSTM<T extends TNumber> Computes the LSTM cell forward propagation for all the time steps.  BlockLSTMGrad<T extends TNumber> Computes the LSTM cell backward propagation for the entire time sequence.  BoostedTreesAggregateStats Aggregates the summary of accumulated stats for the batch.  BoostedTreesBucketize Bucketize each feature based on bucket boundaries.  BoostedTreesCalculateBestFeatureSplit Calculates gains for each feature and returns the best possible split information for the feature.  BoostedTreesCalculateBestFeatureSplitV2 Calculates gains for each feature and returns the best possible split information for each node.  BoostedTreesCalculateBestGainsPerFeature Calculates gains for each feature and returns the best possible split information for the feature.  BoostedTreesCenterBias Calculates the prior from the training data (the bias) and fills in the first node with the logits' prior.  BoostedTreesCreateEnsemble Creates a tree ensemble model and returns a handle to it.  BoostedTreesCreateQuantileStreamResource Create the Resource for Quantile Streams.  BoostedTreesDeserializeEnsemble Deserializes a serialized tree ensemble config and replaces current tree ensemble.  BoostedTreesEnsembleResourceHandleOp Creates a handle to a BoostedTreesEnsembleResource   BoostedTreesExampleDebugOutputs Debugging/model interpretability outputs for each example.  BoostedTreesFlushQuantileSummaries Flush the quantile summaries from each quantile stream resource.  BoostedTreesGetEnsembleStates Retrieves the tree ensemble resource stamp token, number of trees and growing statistics.  BoostedTreesMakeQuantileSummaries Makes the summary of quantiles for the batch.  BoostedTreesMakeStatsSummary Makes the summary of accumulated stats for the batch.  BoostedTreesPredict Runs multiple additive regression ensemble predictors on input instances and computes the logits.  BoostedTreesQuantileStreamResourceAddSummaries Add the quantile summaries to each quantile stream resource.  BoostedTreesQuantileStreamResourceDeserialize Deserialize bucket boundaries and ready flag into current QuantileAccumulator.  BoostedTreesQuantileStreamResourceFlush Flush the summaries for a quantile stream resource.  BoostedTreesQuantileStreamResourceGetBucketBoundaries Generate the bucket boundaries for each feature based on accumulated summaries.  BoostedTreesQuantileStreamResourceHandleOp Creates a handle to a BoostedTreesQuantileStreamResource.  BoostedTreesSerializeEnsemble Serializes the tree ensemble to a proto.  BoostedTreesSparseAggregateStats Aggregates the summary of accumulated stats for the batch.  BoostedTreesSparseCalculateBestFeatureSplit Calculates gains for each feature and returns the best possible split information for the feature.  BoostedTreesTrainingPredict Runs multiple additive regression ensemble predictors on input instances and computes the update to cached logits.  BoostedTreesUpdateEnsemble Updates the tree ensemble by either adding a layer to the last tree being grown or by starting a new tree.  BoostedTreesUpdateEnsembleV2 Updates the tree ensemble by adding a layer to the last tree being grown or by starting a new tree.  BroadcastDynamicShape<T extends TNumber> Return the shape of s0 op s1 with broadcast.  BroadcastGradientArgs<T extends TNumber> Return the reduction indices for computing gradients of s0 op s1 with broadcast.  BroadcastHelper<T extends TType> Helper operator for performing XLA-style broadcasts Broadcasts `lhs` and `rhs` to the same rank, by adding size 1 dimensions to whichever of `lhs` and `rhs` has the lower rank, using XLA's broadcasting rules for binary operators.  BroadcastRecv<T extends TType> Receives a tensor value broadcast from another device.  BroadcastSend<T extends TType> Broadcasts a tensor value to one or more other devices.  BroadcastTo<T extends TType> Broadcast an array for a compatible shape.  Bucketize Bucketizes 'input' based on 'boundaries'.  BytesProducedStatsDataset Records the bytes size of each element of `input_dataset` in a StatsAggregator.  CSRSparseMatrixComponents<T extends TType> Reads out the CSR components at batch `index`.  CSRSparseMatrixToDense<T extends TType> Convert a (possibly batched) CSRSparseMatrix to dense.  CSRSparseMatrixToSparseTensor<T extends TType> Converts a (possibly batched) CSRSparesMatrix to a SparseTensor.  CSVDataset   CSVDatasetV2   CTCLossV2 Calculates the CTC Loss (log probability) for each batch entry.  CacheDataset Creates a dataset that caches elements from `input_dataset`.  CacheDatasetV2   Cast<U extends TType> Cast x of type SrcT to y of DstT.  Ceil<T extends TNumber> Returns element-wise smallest integer not less than x.  CheckNumerics<T extends TNumber> Checks a tensor for NaN, -Inf and +Inf values.  Cholesky<T extends TType> Computes the Cholesky decomposition of one or more square matrices.  CholeskyGrad<T extends TNumber> Computes the reverse mode backpropagated gradient of the Cholesky algorithm.  ChooseFastestDataset   ClipByValue<T extends TType> Clips tensor values to a specified min and max.  CloseSummaryWriter   ClusterOutput<T extends TType> Operator that connects the output of an XLA computation to other consumer graph nodes.  CollectiveGather<T extends TNumber> Mutually accumulates multiple tensors of identical type and shape.  CollectivePermute<T extends TType> An Op to permute tensors across replicated TPU instances.  CombinedNonMaxSuppression Greedily selects a subset of bounding boxes in descending order of score, This operation performs non_max_suppression on the inputs per batch, across all classes.  CompareAndBitpack Compare values of `input` to `threshold` and pack resulting bits into a `uint8`.  CompilationResult Returns the result of a TPU compilation.  CompileSucceededAssert Asserts that compilation succeeded.  Complex<U extends TType> Converts two real numbers to a complex number.  ComplexAbs<U extends TNumber> Computes the complex absolute value of a tensor.  CompressElement Compresses a dataset element.  ComputeAccidentalHits Computes the ids of the positions in sampled_candidates that match true_labels.  ComputeBatchSize Computes the static batch size of a dataset sans partial batches.  Concat<T extends TType> Concatenates tensors along one dimension.  ConcatenateDataset Creates a dataset that concatenates `input_dataset` with `another_dataset`.  ConditionalAccumulator A conditional accumulator for aggregating gradients.  ConfigureDistributedTPU Sets up the centralized structures for a distributed TPU system.  ConfigureTPUEmbedding Sets up TPUEmbedding in a distributed TPU system.  Conj<T extends TType> Returns the complex conjugate of a complex number.  ConjugateTranspose<T extends TType> Shuffle dimensions of x according to a permutation and conjugate the result.  Constant<T extends TType> An operator producing a constant value.  ConsumeMutexLock This op consumes a lock created by `MutexLock`.  ControlTrigger Does nothing.  Conv<T extends TType> Wraps the XLA ConvGeneralDilated operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#conv_convolution .  Conv2d<T extends TNumber> Computes a 2-D convolution given 4-D `input` and `filter` tensors.  Conv2dBackpropFilter<T extends TNumber> Computes the gradients of convolution with respect to the filter.  Conv2dBackpropInput<T extends TNumber> Computes the gradients of convolution with respect to the input.  Conv3d<T extends TNumber> Computes a 3-D convolution given 5-D `input` and `filter` tensors.  Conv3dBackpropFilter<T extends TNumber> Computes the gradients of 3-D convolution with respect to the filter.  Conv3dBackpropInput<U extends TNumber> Computes the gradients of 3-D convolution with respect to the input.  Copy<T extends TType> Copy a tensor from CPU-to-CPU or GPU-to-GPU.  CopyHost<T extends TType> Copy a tensor to host.  Cos<T extends TType> Computes cos of x element-wise.  Cosh<T extends TType> Computes hyperbolic cosine of x element-wise.  CountUpTo<T extends TNumber> Increments 'ref' until it reaches 'limit'.  CreateSummaryDbWriter   CreateSummaryFileWriter   CropAndResize Extracts crops from the input image tensor and resizes them.  CropAndResizeGradBoxes Computes the gradient of the crop_and_resize op wrt the input boxes tensor.  CropAndResizeGradImage<T extends TNumber> Computes the gradient of the crop_and_resize op wrt the input image tensor.  Cross<T extends TNumber> Compute the pairwise cross product.  CrossReplicaSum<T extends TNumber> An Op to sum inputs across replicated TPU instances.  CtcBeamSearchDecoder<T extends TNumber> Performs beam search decoding on the logits given in input.  CtcGreedyDecoder<T extends TNumber> Performs greedy decoding on the logits given in inputs.  CtcLoss<T extends TNumber> Calculates the CTC Loss (log probability) for each batch entry.  CudnnRNN<T extends TNumber> A RNN backed by cuDNN.  CudnnRNNBackprop<T extends TNumber> Backprop step of CudnnRNNV3.  CudnnRNNCanonicalToParams<T extends TNumber> Converts CudnnRNN params from canonical form to usable form.  CudnnRNNParamsToCanonical<T extends TNumber> Retrieves CudnnRNN params in canonical form.  CudnnRnnParamsSize<U extends TNumber> Computes size of weights that can be used by a Cudnn RNN model.  Cumprod<T extends TType> Compute the cumulative product of the tensor `x` along `axis`.  Cumsum<T extends TType> Compute the cumulative sum of the tensor `x` along `axis`.  CumulativeLogsumexp<T extends TNumber> Compute the cumulative product of the tensor `x` along `axis`.  DataFormatDimMap<T extends TNumber> Returns the dimension index in the destination data format given the one in the source data format.  DataFormatVecPermute<T extends TNumber> Permute input tensor from `src_format` to `dst_format`.  DataServiceDataset   DatasetCardinality Returns the cardinality of `input_dataset`.  DatasetFromGraph Creates a dataset from the given `graph_def`.  DatasetToGraph Returns a serialized GraphDef representing `input_dataset`.  DatasetToSingleElement Outputs the single element from the given dataset.  DatasetToTFRecord Writes the given dataset to the given file using the TFRecord format.  DatasetToTfRecord Writes the given dataset to the given file using the TFRecord format.  Dawsn<T extends TNumber>   DebugGradientIdentity<T extends TType> Identity op for gradient debugging.  DebugGradientRefIdentity<T extends TType> Identity op for gradient debugging.  DebugIdentity<T extends TType> Debug Identity V2 Op.  DebugNanCount Debug NaN Value Counter Op.  DebugNumericsSummary<U extends TNumber> Debug Numeric Summary V2 Op.  DecodeAndCropJpeg Decode and Crop a JPEG-encoded image to a uint8 tensor.  DecodeBase64 Decode web-safe base64-encoded strings.  DecodeBmp Decode the first frame of a BMP-encoded image to a uint8 tensor.  DecodeCompressed Decompress strings.  DecodeCsv Convert CSV records to tensors.  DecodeGif Decode the frame(s) of a GIF-encoded image to a uint8 tensor.  DecodeImage<T extends TNumber> Function for decode_bmp, decode_gif, decode_jpeg, and decode_png.  DecodeJpeg Decode a JPEG-encoded image to a uint8 tensor.  DecodeJsonExample Convert JSON-encoded Example records to binary protocol buffer strings.  DecodePaddedRaw<T extends TNumber> Reinterpret the bytes of a string as a vector of numbers.  DecodePng<T extends TNumber> Decode a PNG-encoded image to a uint8 or uint16 tensor.  DecodeProto The op extracts fields from a serialized protocol buffers message into tensors.  DecodeRaw<T extends TType> Reinterpret the bytes of a string as a vector of numbers.  DecodeWav Decode a 16-bit PCM WAV file to a float tensor.  DeepCopy<T extends TType> Makes a copy of `x`.  DeleteIterator A container for an iterator resource.  DeleteMemoryCache   DeleteMultiDeviceIterator A container for an iterator resource.  DeleteRandomSeedGenerator   DeleteSeedGenerator   DeleteSessionTensor Delete the tensor specified by its handle in the session.  DenseBincount<U extends TNumber> Counts the number of occurrences of each value in an integer array.  DenseCountSparseOutput<U extends TNumber> Performs sparse-output bin counting for a tf.tensor input.  DenseToCSRSparseMatrix Converts a dense tensor to a (possibly batched) CSRSparseMatrix.  DenseToDenseSetOperation<T extends TType> Applies set operation along last dimension of 2 `Tensor` inputs.  DenseToSparseBatchDataset Creates a dataset that batches input elements into a SparseTensor.  DenseToSparseSetOperation<T extends TType> Applies set operation along last dimension of `Tensor` and `SparseTensor`.  DepthToSpace<T extends TType> DepthToSpace for tensors of type T.  DepthwiseConv2dNative<T extends TNumber> Computes a 2-D depthwise convolution given 4-D `input` and `filter` tensors.  DepthwiseConv2dNativeBackpropFilter<T extends TNumber> Computes the gradients of depthwise convolution with respect to the filter.  DepthwiseConv2dNativeBackpropInput<T extends TNumber> Computes the gradients of depthwise convolution with respect to the input.  Dequantize Takes the packed uint32 input and unpacks the input to uint8 to do Dequantization on device.  DeserializeIterator Converts the given variant tensor to an iterator and stores it in the given resource.  DeserializeManySparse<T extends TType> Deserialize and concatenate `SparseTensors` from a serialized minibatch.  DeserializeSparse<U extends TType> Deserialize `SparseTensor` objects.  DestroyResourceOp Deletes the resource specified by the handle.  DestroyTemporaryVariable<T extends TType> Destroys the temporary variable and returns its final value.  Det<T extends TType> Computes the determinant of one or more square matrices.  DeviceIndex Return the index of device the op runs.  Digamma<T extends TNumber> Computes Psi, the derivative of Lgamma (the log of the absolute value of `Gamma(x)`), element-wise.  Dilation2d<T extends TNumber> Computes the grayscale dilation of 4-D `input` and 3-D `filter` tensors.  Dilation2dBackpropFilter<T extends TNumber> Computes the gradient of morphological 2-D dilation with respect to the filter.  Dilation2dBackpropInput<T extends TNumber> Computes the gradient of morphological 2-D dilation with respect to the input.  DirectedInterleaveDataset A substitute for `InterleaveDataset` on a fixed list of `N` datasets.  Div<T extends TType> Returns x / y element-wise.  DivNoNan<T extends TType> Returns 0 if the denominator is zero.  Dot<T extends TType> Wraps the XLA DotGeneral operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#dotgeneral .  DrawBoundingBoxes<T extends TNumber> Draw bounding boxes on a batch of images.  DummyIterationCounter   DummyMemoryCache   DummySeedGenerator   DynamicPartition<T extends TType> Partitions `data` into `num_partitions` tensors using indices from `partitions`.  DynamicSlice<T extends TType> Wraps the XLA DynamicSlice operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#dynamicslice .  DynamicStitch<T extends TType> Interleave the values from the `data` tensors into a single tensor.  DynamicUpdateSlice<T extends TType> Wraps the XLA DynamicUpdateSlice operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#dynamicupdateslice .  EditDistance Computes the (possibly normalized) Levenshtein Edit Distance.  Eig<U extends TType> Computes the eigen decomposition of one or more square matrices.  Einsum<T extends TType> An op which supports basic einsum op with 2 inputs and 1 output.  Elu<T extends TNumber> Computes exponential linear: `exp(features) - 1` if < 0, `features` otherwise.  EluGrad<T extends TNumber> Computes gradients for the exponential linear (Elu) operation.  EmbeddingActivations An op enabling differentiation of TPU Embeddings.  Empty<T extends TType> Creates a tensor with the given shape.  EmptyTensorList Creates and returns an empty tensor list.  EmptyTensorMap Creates and returns an empty tensor map.  EncodeBase64 Encode strings into web-safe base64 format.  EncodeJpeg JPEG-encode an image.  EncodeJpegVariableQuality JPEG encode input image with provided compression quality.  EncodePng PNG-encode an image.  EncodeProto The op serializes protobuf messages provided in the input tensors.  EncodeWav Encode audio data using the WAV file format.  EnqueueTPUEmbeddingIntegerBatch An op that enqueues a list of input batch tensors to TPUEmbedding.  EnqueueTPUEmbeddingRaggedTensorBatch Eases the porting of code that uses tf.nn.embedding_lookup().  EnqueueTPUEmbeddingSparseBatch An op that enqueues TPUEmbedding input indices from a SparseTensor.  EnqueueTPUEmbeddingSparseTensorBatch Eases the porting of code that uses tf.nn.embedding_lookup_sparse().  EnsureShape<T extends TType> Ensures that the tensor's shape matches the expected shape.  Enter<T extends TType> Creates or finds a child frame, and makes `data` available to the child frame.  Equal Returns the truth value of (x == y) element-wise.  Erf<T extends TNumber> Computes the Gauss error function of `x` element-wise.  Erfc<T extends TNumber> Computes the complementary error function of `x` element-wise.  EuclideanNorm<T extends TType> Computes the euclidean norm of elements across dimensions of a tensor.  Execute Op that loads and executes a TPU program on a TPU device.  ExecuteAndUpdateVariables Op that executes a program with optional in-place variable updates.  Exit<T extends TType> Exits the current frame to its parent frame.  Exp<T extends TType> Computes exponential of x element-wise.  ExpandDims<T extends TType> Inserts a dimension of 1 into a tensor's shape.  Expint<T extends TNumber>   Expm1<T extends TType> Computes `exp(x) - 1` element-wise.  ExtractGlimpse Extracts a glimpse from the input tensor.  ExtractImagePatches<T extends TType> Extract `patches` from `images` and put them in the "depth" output dimension.  ExtractJpegShape<T extends TNumber> Extract the shape information of a JPEG-encoded image.  ExtractVolumePatches<T extends TNumber> Extract `patches` from `input` and put them in the `"depth"` output dimension.  Fact Output a fact about factorials.  FakeQuantWithMinMaxArgs Fake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same type.  FakeQuantWithMinMaxArgsGradient Compute gradients for a FakeQuantWithMinMaxArgs operation.  FakeQuantWithMinMaxVars Fake-quantize the 'inputs' tensor of type float via global float scalars Fake-quantize the `inputs` tensor of type float via global float scalars `min` and `max` to `outputs` tensor of same shape as `inputs`.  FakeQuantWithMinMaxVarsGradient Compute gradients for a FakeQuantWithMinMaxVars operation.  FakeQuantWithMinMaxVarsPerChannel Fake-quantize the 'inputs' tensor of type float via per-channel floats Fake-quantize the `inputs` tensor of type float per-channel and one of the shapes: `[d]`, `[b, d]` `[b, h, w, d]` via per-channel floats `min` and `max` of shape `[d]` to `outputs` tensor of same shape as `inputs`.  FakeQuantWithMinMaxVarsPerChannelGradient Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.  Fft<T extends TType> Fast Fourier transform.  Fft2d<T extends TType> 2D fast Fourier transform.  Fft3d<T extends TType> 3D fast Fourier transform.  FifoQueue A queue that produces elements in first-in first-out order.  Fill<U extends TType> Creates a tensor filled with a scalar value.  FilterByLastComponentDataset Creates a dataset containing elements of first component of `input_dataset` having true in the last component.  Fingerprint Generates fingerprint values.  FixedLengthRecordDataset   FixedLengthRecordReader A Reader that outputs fixed-length records from a file.  FixedUnigramCandidateSampler Generates labels for candidate sampling with a learned unigram distribution.  Floor<T extends TNumber> Returns element-wise largest integer not greater than x.  FloorDiv<T extends TType> Returns x // y element-wise.  FloorMod<T extends TNumber> Returns element-wise remainder of division.  FlushSummaryWriter   FractionalAvgPool<T extends TNumber> Performs fractional average pooling on the input.  FractionalAvgPoolGrad<T extends TNumber> Computes gradient of the FractionalAvgPool function.  FractionalMaxPool<T extends TNumber> Performs fractional max pooling on the input.  FractionalMaxPoolGrad<T extends TNumber> Computes gradient of the FractionalMaxPool function.  FresnelCos<T extends TNumber>   FresnelSin<T extends TNumber>   FusedBatchNorm<T extends TNumber, U extends TNumber> Batch normalization.  FusedBatchNormGrad<T extends TNumber, U extends TNumber> Gradient for batch normalization.  FusedPadConv2d<T extends TNumber> Performs a padding as a preprocess during a convolution.  FusedResizeAndPadConv2d<T extends TNumber> Performs a resize and padding as a preprocess during a convolution.  GRUBlockCell<T extends TNumber> Computes the GRU cell forward propagation for 1 time step.  GRUBlockCellGrad<T extends TNumber> Computes the GRU cell back-propagation for 1 time step.  Gather<T extends TType> Wraps the XLA Gather operator documented at https://www.tensorflow.org/xla/operation_semantics#gather  GatherNd<T extends TType> Gather slices from `params` into a Tensor with shape specified by `indices`.  GatherV2<T extends TNumber> Mutually accumulates multiple tensors of identical type and shape.  GenerateBoundingBoxProposals This op produces Region of Interests from given bounding boxes(bbox_deltas) encoded wrt anchors according to eq.2 in arXiv:1506.01497 The op selects top `pre_nms_topn` scoring boxes, decodes them with respect to anchors, applies non-maximal suppression on overlapping boxes with higher than `nms_threshold` intersection-over-union (iou) value, discarding boxes where shorter side is less than `min_size`.  GenerateVocabRemapping Given a path to new and old vocabulary files, returns a remapping Tensor of length `num_new_vocab`, where `remapping[i]` contains the row number in the old vocabulary that corresponds to row `i` in the new vocabulary (starting at line `new_vocab_offset` and up to `num_new_vocab` entities), or `-1` if entry `i` in the new vocabulary is not in the old vocabulary.  GetSessionHandle Store the input tensor in the state of the current session.  GetSessionTensor<T extends TType> Get the value of the tensor specified by its handle.  Greater Returns the truth value of (x > y) element-wise.  GreaterEqual Returns the truth value of (x >= y) element-wise.  GuaranteeConst<T extends TType> Gives a guarantee to the TF runtime that the input tensor is a constant.  HashTable Creates a non-initialized hash table.  HistogramFixedWidth<U extends TNumber> Return histogram of values.  HistogramSummary Outputs a `Summary` protocol buffer with a histogram.  HsvToRgb<T extends TNumber> Convert one or more images from HSV to RGB.  Identity<T extends TType> Return a tensor with the same shape and contents as the input tensor or value.  IdentityN Returns a list of tensors with the same shapes and contents as the input tensors.  IdentityReader A Reader that outputs the queued work as both the key and value.  Ifft<T extends TType> Inverse fast Fourier transform.  Ifft2d<T extends TType> Inverse 2D fast Fourier transform.  Ifft3d<T extends TType> Inverse 3D fast Fourier transform.  Igamma<T extends TNumber> Compute the lower regularized incomplete Gamma function `P(a, x)`.  IgammaGradA<T extends TNumber> Computes the gradient of `igamma(a, x)` wrt `a`.  Igammac<T extends TNumber> Compute the upper regularized incomplete Gamma function `Q(a, x)`.  IgnoreErrorsDataset Creates a dataset that contains the elements of `input_dataset` ignoring errors.  Imag<U extends TNumber> Returns the imaginary part of a complex number.  ImageProjectiveTransformV2<T extends TNumber> Applies the given transform to each of the images.  ImageProjectiveTransformV3<T extends TNumber> Applies the given transform to each of the images.  ImageSummary Outputs a `Summary` protocol buffer with images.  ImmutableConst<T extends TType> Returns immutable tensor from memory region.  ImportEvent   InTopK Says whether the targets are in the top `K` predictions.  InfeedDequeue<T extends TType> A placeholder op for a value that will be fed into the computation.  InfeedDequeueTuple Fetches multiple values from infeed as an XLA tuple.  InfeedEnqueue An op which feeds a single Tensor value into the computation.  InfeedEnqueuePrelinearizedBuffer An op which enqueues prelinearized buffer into TPU infeed.  InfeedEnqueueTuple Feeds multiple Tensor values into the computation as an XLA tuple.  Init   InitializeTable Table initializer that takes two tensors for keys and values respectively.  InitializeTableFromDataset   InitializeTableFromTextFile Initializes a table from a text file.  InplaceAdd<T extends TType> Adds v into specified rows of x.  InplaceSub<T extends TType> Subtracts `v` into specified rows of `x`.  InplaceUpdate<T extends TType> Updates specified rows 'i' with values 'v'.  Inv<T extends TType> Computes the inverse of one or more square invertible matrices or their adjoints (conjugate transposes).  InvGrad<T extends TType> Computes the gradient for the inverse of `x` wrt its input.  Invert<T extends TNumber> Invert (flip) each bit of supported types; for example, type `uint8` value 01010101 becomes 10101010.  InvertPermutation<T extends TNumber> Computes the inverse permutation of a tensor.  Irfft<U extends TNumber> Inverse real-valued fast Fourier transform.  Irfft2d<U extends TNumber> Inverse 2D real-valued fast Fourier transform.  Irfft3d<U extends TNumber> Inverse 3D real-valued fast Fourier transform.  IsBoostedTreesEnsembleInitialized Checks whether a tree ensemble has been initialized.  IsBoostedTreesQuantileStreamResourceInitialized Checks whether a quantile stream has been initialized.  IsFinite Returns which elements of x are finite.  IsInf Returns which elements of x are Inf.  IsNan Returns which elements of x are NaN.  IsVariableInitialized Checks whether a tensor has been initialized.  IsotonicRegression<U extends TNumber> Solves a batch of isotonic regression problems.  Iterator   IteratorFromStringHandle   IteratorGetDevice Returns the name of the device on which `resource` has been placed.  IteratorGetNext Gets the next output from the given iterator .  IteratorGetNextAsOptional Gets the next output from the given iterator as an Optional variant.  IteratorGetNextSync Gets the next output from the given iterator.  IteratorToStringHandle Converts the given `resource_handle` representing an iterator to a string.  Join Joins the strings in the given list of string tensors into one tensor; with the given separator (default is an empty separator).  KMC2ChainInitialization Returns the index of a data point that should be added to the seed set.  KeyValueSort<T extends TNumber, U extends TType> Wraps the XLA Sort operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#sort .  KmeansPlusPlusInitialization Selects num_to_sample rows of input using the KMeans++ criterion.  KthOrderStatistic Computes the Kth order statistic of a data set.  L2Loss<T extends TNumber> L2 Loss.  LMDBDataset Creates a dataset that emits the key-value pairs in one or more LMDB files.  LSTMBlockCell<T extends TNumber> Computes the LSTM cell forward propagation for 1 time step.  LSTMBlockCellGrad<T extends TNumber> Computes the LSTM cell backward propagation for 1 timestep.  LatencyStatsDataset Records the latency of producing `input_dataset` elements in a StatsAggregator.  LeakyRelu<T extends TNumber> Computes rectified linear: `max(features, features * alpha)`.  LeakyReluGrad<T extends TNumber> Computes rectified linear gradients for a LeakyRelu operation.  LearnedUnigramCandidateSampler Generates labels for candidate sampling with a learned unigram distribution.  LeftShift<T extends TNumber> Elementwise computes the bitwise left-shift of `x` and `y`.  Less Returns the truth value of (x < y) element-wise.  LessEqual Returns the truth value of (x <= y) element-wise.  Lgamma<T extends TNumber> Computes the log of the absolute value of `Gamma(x)` element-wise.  LinSpace<T extends TNumber> Generates values in an interval.  LmdbDataset   LmdbReader A Reader that outputs the records from a LMDB file.  LoadAndRemapMatrix Loads a 2-D (matrix) `Tensor` with name `old_tensor_name` from the checkpoint at `ckpt_path` and potentially reorders its rows and columns using the specified remappings.  LoadTPUEmbeddingADAMParameters Load ADAM embedding parameters.  LoadTPUEmbeddingADAMParametersGradAccumDebug Load ADAM embedding parameters with debug support.  LoadTPUEmbeddingAdadeltaParameters Load Adadelta embedding parameters.  LoadTPUEmbeddingAdadeltaParametersGradAccumDebug Load Adadelta parameters with debug support.  LoadTPUEmbeddingAdagradParameters Load Adagrad embedding parameters.  LoadTPUEmbeddingAdagradParametersGradAccumDebug Load Adagrad embedding parameters with debug support.  LoadTPUEmbeddingCenteredRMSPropParameters Load centered RMSProp embedding parameters.  LoadTPUEmbeddingFTRLParameters Load FTRL embedding parameters.  LoadTPUEmbeddingFTRLParametersGradAccumDebug Load FTRL embedding parameters with debug support.  LoadTPUEmbeddingMDLAdagradLightParameters Load MDL Adagrad Light embedding parameters.  LoadTPUEmbeddingMomentumParameters Load Momentum embedding parameters.  LoadTPUEmbeddingMomentumParametersGradAccumDebug Load Momentum embedding parameters with debug support.  LoadTPUEmbeddingProximalAdagradParameters Load proximal Adagrad embedding parameters.  LoadTPUEmbeddingProximalAdagradParametersGradAccumDebug Load proximal Adagrad embedding parameters with debug support.  LoadTPUEmbeddingProximalYogiParameters   LoadTPUEmbeddingProximalYogiParametersGradAccumDebug   LoadTPUEmbeddingRMSPropParameters Load RMSProp embedding parameters.  LoadTPUEmbeddingRMSPropParametersGradAccumDebug Load RMSProp embedding parameters with debug support.  LoadTPUEmbeddingStochasticGradientDescentParameters Load SGD embedding parameters.  LoadTPUEmbeddingStochasticGradientDescentParametersGradAccumDebug Load SGD embedding parameters.  LocalResponseNormalization<T extends TNumber> Local Response Normalization.  LocalResponseNormalizationGrad<T extends TNumber> Gradients for Local Response Normalization.  Log<T extends TType> Computes natural logarithm of x element-wise.  Log1p<T extends TType> Computes natural logarithm of (1 + x) element-wise.  LogMatrixDeterminant<T extends TType> Computes the sign and the log of the absolute value of the determinant of one or more square matrices.  LogSoftmax<T extends TNumber> Computes log softmax activations.  LogUniformCandidateSampler Generates labels for candidate sampling with a log-uniform distribution.  LogicalAnd Returns the truth value of x AND y element-wise.  LogicalNot Returns the truth value of `NOT x` element-wise.  LogicalOr Returns the truth value of x OR y element-wise.  LookupTableExport<T extends TType, U extends TType> Outputs all keys and values in the table.  LookupTableFind<U extends TType> Looks up keys in a table, outputs the corresponding values.  LookupTableImport Replaces the contents of the table with the specified keys and values.  LookupTableInsert Updates the table to associates keys with values.  LookupTableRemove Removes keys and its associated values from a table.  LookupTableSize Computes the number of elements in the given table.  LoopCond Forwards the input to the output.  Lower Converts all uppercase characters into their respective lowercase replacements.  LowerBound<U extends TNumber> Applies lower_bound(sorted_search_values, values) along each row.  Lu<T extends TType, U extends TNumber> Computes the LU decomposition of one or more square matrices.  MakeIterator Makes a new iterator from the given `dataset` and stores it in `iterator`.  MakeUnique Make all elements in the non-Batch dimension unique, but \"close\" to their initial value.  MapClear Op removes all elements in the underlying container.  MapIncompleteSize Op returns the number of incomplete elements in the underlying container.  MapPeek Op peeks at the values at the specified key.  MapSize Op returns the number of elements in the underlying container.  MapStage Stage (key, values) in the underlying container which behaves like a hashtable.  MapUnstage Op removes and returns the values associated with the key from the underlying container.  MapUnstageNoKey Op removes and returns a random (key, value) from the underlying container.  MatMul<T extends TType> Multiply the matrix "a" by the matrix "b".  MatchingFiles Returns the set of files matching one or more glob patterns.  MatchingFilesDataset   MatrixDiag<T extends TType> Returns a batched diagonal tensor with given batched diagonal values.  MatrixDiagPart<T extends TType> Returns the batched diagonal part of a batched tensor.  MatrixDiagPartV3<T extends TType> Returns the batched diagonal part of a batched tensor.  MatrixDiagV3<T extends TType> Returns a batched diagonal tensor with given batched diagonal values.  MatrixLogarithm<T extends TType> Computes the matrix logarithm of one or more square matrices: \\(log(exp(A)) = A\\) This op is only defined for complex matrices.  MatrixSetDiag<T extends TType> Returns a batched matrix tensor with new batched diagonal values.  MatrixSolveLs<T extends TType> Solves one or more linear least-squares problems.  Max<T extends TType> Computes the maximum of elements across dimensions of a tensor.  MaxIntraOpParallelismDataset Creates a dataset that overrides the maximum intra-op parallelism.  MaxPool<T extends TType> Performs max pooling on the input.  MaxPool3d<T extends TNumber> Performs 3D max pooling on the input.  MaxPool3dGrad<U extends TNumber> Computes gradients of 3D max pooling function.  MaxPool3dGradGrad<T extends TNumber> Computes second-order gradients of the maxpooling function.  MaxPoolGrad<T extends TNumber> Computes gradients of the maxpooling function.  MaxPoolGradGrad<T extends TNumber> Computes second-order gradients of the maxpooling function.  MaxPoolGradGradWithArgmax<T extends TNumber> Computes second-order gradients of the maxpooling function.  MaxPoolGradWithArgmax<T extends TNumber> Computes gradients of the maxpooling function.  MaxPoolWithArgmax<T extends TNumber, U extends TNumber> Performs max pooling on the input and outputs both max values and indices.  Maximum<T extends TNumber> Returns the max of x and y (i.e.  Mean<T extends TType> Computes the mean of elements across dimensions of a tensor.  Merge<T extends TType> Forwards the value of an available tensor from `inputs` to `output`.  MergeSummary Merges summaries.  MergeV2Checkpoints V2 format specific: merges the metadata files of sharded checkpoints.  Mfcc Transforms a spectrogram into a form that's useful for speech recognition.  Min<T extends TType> Computes the minimum of elements across dimensions of a tensor.  Minimum<T extends TNumber> Returns the min of x and y (i.e.  MirrorPad<T extends TType> Pads a tensor with mirrored values.  MirrorPadGrad<T extends TType> Gradient op for `MirrorPad` op.  MlirPassthroughOp Wraps an arbitrary MLIR computation expressed as a module with a main() function.  Mod<T extends TNumber> Returns element-wise remainder of division.  ModelDataset Identity transformation that models performance.  Mul<T extends TType> Returns x * y element-wise.  MulNoNan<T extends TType> Returns x * y element-wise.  MultiDeviceIterator Creates a MultiDeviceIterator resource.  MultiDeviceIteratorFromStringHandle Generates a MultiDeviceIterator resource from its provided string handle.  MultiDeviceIteratorGetNextFromShard Gets next element for the provided shard number.  MultiDeviceIteratorInit Initializes the multi device iterator with the given dataset.  MultiDeviceIteratorToStringHandle Produces a string handle for the given MultiDeviceIterator.  Multinomial<U extends TNumber> Draws samples from a multinomial distribution.  MutableDenseHashTable Creates an empty hash table that uses tensors as the backing store.  MutableHashTable Creates an empty hash table.  MutableHashTableOfTensors Creates an empty hash table.  Mutex Creates a Mutex resource that can be locked by `MutexLock`.  MutexLock Locks a mutex resource.  NcclAllReduce<T extends TNumber> Outputs a tensor containing the reduction across all input tensors.  NcclBroadcast<T extends TNumber> Sends `input` to all devices that are connected to the output.  NcclReduce<T extends TNumber> Reduces `input` from `num_devices` using `reduction` to a single device.  Ndtri<T extends TNumber>   NearestNeighbors Selects the k nearest centers for each point.  Neg<T extends TType> Computes numerical negative value element-wise.  NegTrain Training via negative sampling.  NextAfter<T extends TNumber> Returns the next representable value of `x1` in the direction of `x2`, element-wise.  NextIteration<T extends TType> Makes its input available to the next iteration.  NoOp Does nothing.  NonDeterministicInts<U extends TType> Non-deterministically generates some integers.  NonMaxSuppression<T extends TNumber> Greedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high intersection-over-union (IOU) overlap with previously selected boxes.  NonMaxSuppressionWithOverlaps Greedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high overlaps with previously selected boxes.  NonSerializableDataset   NotEqual Returns the truth value of (x != y) element-wise.  NthElement<T extends TNumber> Finds values of the `n`-th order statistic for the last dimension.  OneHot<U extends TType> Returns a one-hot tensor.  Ones<T extends TType> An operator creating a constant initialized with ones of the shape given by `dims`.  OnesLike<T extends TType> Returns a tensor of ones with the same shape and type as x.  Operand<T extends TType> Interface implemented by operands of a TensorFlow operation.  OptimizeDataset Creates a dataset by applying optimizations to `input_dataset`.  OptimizeDatasetV2 Creates a dataset by applying related optimizations to `input_dataset`.  OptionalFromValue Constructs an Optional variant from a tuple of tensors.  OptionalGetValue Returns the value stored in an Optional variant or raises an error if none exists.  OptionalHasValue Returns true if and only if the given Optional variant has a value.  OptionalNone Creates an Optional variant with no value.  OrderedMapClear Op removes all elements in the underlying container.  OrderedMapIncompleteSize Op returns the number of incomplete elements in the underlying container.  OrderedMapPeek Op peeks at the values at the specified key.  OrderedMapSize Op returns the number of elements in the underlying container.  OrderedMapStage Stage (key, values) in the underlying container which behaves like a ordered associative container.  OrderedMapUnstage Op removes and returns the values associated with the key from the underlying container.  OrderedMapUnstageNoKey Op removes and returns the (key, value) element with the smallest key from the underlying container.  OrdinalSelector A TPU core selector Op.  OutfeedDequeue<T extends TType> Retrieves a single tensor from the computation outfeed.  OutfeedDequeueTuple Retrieve multiple values from the computation outfeed.  OutfeedDequeueTupleV2 Retrieve multiple values from the computation outfeed.  OutfeedDequeueV2<T extends TType> Retrieves a single tensor from the computation outfeed.  OutfeedEnqueue Enqueue a Tensor on the computation outfeed.  OutfeedEnqueueTuple Enqueue multiple Tensor values on the computation outfeed.  Output<T extends TType> A symbolic handle to a tensor produced by an Operation.  Pad<T extends TType> Wraps the XLA Pad operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#pad .  PaddedBatchDataset Creates a dataset that batches and pads `batch_size` elements from the input.  PaddingFifoQueue A queue that produces elements in first-in first-out order.  ParallelConcat<T extends TType> Concatenates a list of `N` tensors along the first dimension.  ParallelDynamicStitch<T extends TType> Interleave the values from the `data` tensors into a single tensor.  ParameterizedTruncatedNormal<U extends TNumber> Outputs random values from a normal distribution.  ParseExample Transforms a vector of tf.Example protos (as strings) into typed tensors.  ParseExampleDataset Transforms `input_dataset` containing `Example` protos as vectors of DT_STRING into a dataset of `Tensor` or `SparseTensor` objects representing the parsed features.  ParseSequenceExample Transforms a vector of tf.io.SequenceExample protos (as strings) into typed tensors.  ParseSingleExample Transforms a tf.Example proto (as a string) into typed tensors.  ParseSingleSequenceExample Transforms a scalar brain.SequenceExample proto (as strings) into typed tensors.  ParseTensor<T extends TType> Transforms a serialized tensorflow.TensorProto proto into a Tensor.  PartitionedInput<T extends TType> An op that groups a list of partitioned inputs together.  PartitionedOutput<T extends TType> An op that demultiplexes a tensor to be sharded by XLA to a list of partitioned outputs outside the XLA computation.  Placeholder<T extends TType> A placeholder op for a value that will be fed into the computation.  PlaceholderWithDefault<T extends TType> A placeholder op that passes through `input` when its output is not fed.  Polygamma<T extends TNumber> Compute the polygamma function \\(\psi^{(n)}(x)\\).  PopulationCount Computes element-wise population count (a.k.a.  Pow<T extends TType> Computes the power of one value to another.  PrefetchDataset Creates a dataset that asynchronously prefetches elements from `input_dataset`.  Prelinearize An op which linearizes one Tensor value to an opaque variant tensor.  PrelinearizeTuple An op which linearizes multiple Tensor values to an opaque variant tensor.  PreventGradient<T extends TType> An identity op that triggers an error if a gradient is requested.  Print Prints a string scalar.  PriorityQueue A queue that produces elements sorted by the first component value.  PrivateThreadPoolDataset Creates a dataset that uses a custom thread pool to compute `input_dataset`.  Prod<T extends TType> Computes the product of elements across dimensions of a tensor.  Qr<T extends TType> Computes the QR decompositions of one or more matrices.  Quantize<T extends TType> Quantize the 'input' tensor of type float to 'output' tensor of type 'T'.  QuantizeAndDequantize<T extends TNumber> Quantizes then dequantizes a tensor.  QuantizeAndDequantizeV3<T extends TNumber> Quantizes then dequantizes a tensor.  QuantizeAndDequantizeV4<T extends TNumber> Returns the gradient of `quantization.QuantizeAndDequantizeV4`.  QuantizeAndDequantizeV4Grad<T extends TNumber> Returns the gradient of `QuantizeAndDequantizeV4`.  QuantizeDownAndShrinkRange<U extends TType> Convert the quantized 'input' tensor into a lower-precision 'output', using the actual distribution of the values to maximize the usage of the lower bit depth and adjusting the output min and max ranges accordingly.  QuantizedAdd<V extends TType> Returns x + y element-wise, working on quantized buffers.  QuantizedAvgPool<T extends TType> Produces the average pool of the input tensor for quantized types.  QuantizedBatchNormWithGlobalNormalization<U extends TType> Quantized Batch normalization.  QuantizedBiasAdd<V extends TType> Adds Tensor 'bias' to Tensor 'input' for Quantized types.  QuantizedConcat<T extends TType> Concatenates quantized tensors along one dimension.  QuantizedConv2DAndRelu<V extends TType>   QuantizedConv2DAndReluAndRequantize<V extends TType>   QuantizedConv2DAndRequantize<V extends TType>   QuantizedConv2DPerChannel<V extends TType> Computes QuantizedConv2D per channel.  QuantizedConv2DWithBias<V extends TType>   QuantizedConv2DWithBiasAndRelu<V extends TType>   QuantizedConv2DWithBiasAndReluAndRequantize<W extends TType>   QuantizedConv2DWithBiasAndRequantize<W extends TType>   QuantizedConv2DWithBiasSignedSumAndReluAndRequantize<X extends TType>   QuantizedConv2DWithBiasSumAndRelu<V extends TType>   QuantizedConv2DWithBiasSumAndReluAndRequantize<X extends TType>   QuantizedConv2d<V extends TType> Computes a 2D convolution given quantized 4D input and filter tensors.  QuantizedDepthwiseConv2D<V extends TType> Computes quantized depthwise Conv2D.  QuantizedDepthwiseConv2DWithBias<V extends TType> Computes quantized depthwise Conv2D with Bias.  QuantizedDepthwiseConv2DWithBiasAndRelu<V extends TType> Computes quantized depthwise Conv2D with Bias and Relu.  QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize<W extends TType> Computes quantized depthwise Conv2D with Bias, Relu and Requantize.  QuantizedInstanceNorm<T extends TType> Quantized Instance normalization.  QuantizedMatMul<V extends TType> Perform a quantized matrix multiplication of `a` by the matrix `b`.  QuantizedMatMulWithBias<W extends TType> Performs a quantized matrix multiplication of `a` by the matrix `b` with bias add.  QuantizedMatMulWithBiasAndDequantize<W extends TNumber>   QuantizedMatMulWithBiasAndRelu<V extends TType> Perform a quantized matrix multiplication of `a` by the matrix `b` with bias add and relu fusion.  QuantizedMatMulWithBiasAndReluAndRequantize<W extends TType> Perform a quantized matrix multiplication of `a` by the matrix `b` with bias add and relu and requantize fusion.  QuantizedMatMulWithBiasAndRequantize<W extends TType>   QuantizedMaxPool<T extends TType> Produces the max pool of the input tensor for quantized types.  QuantizedMul<V extends TType> Returns x * y element-wise, working on quantized buffers.  QuantizedRelu<U extends TType> Computes Quantized Rectified Linear: `max(features, 0)`  QuantizedRelu6<U extends TType> Computes Quantized Rectified Linear 6: `min(max(features, 0), 6)`  QuantizedReluX<U extends TType> Computes Quantized Rectified Linear X: `min(max(features, 0), max_value)`  QuantizedReshape<T extends TType> Reshapes a quantized tensor as per the Reshape op.  QuantizedResizeBilinear<T extends TType> Resize quantized `images` to `size` using quantized bilinear interpolation.  QueueClose Closes the given queue.  QueueDequeue Dequeues a tuple of one or more tensors from the given queue.  QueueDequeueMany Dequeues `n` tuples of one or more tensors from the given queue.  QueueDequeueUpTo Dequeues `n` tuples of one or more tensors from the given queue.  QueueEnqueue Enqueues a tuple of one or more tensors in the given queue.  QueueEnqueueMany Enqueues zero or more tuples of one or more tensors in the given queue.  QueueIsClosed Returns true if queue is closed.  QueueSize Computes the number of elements in the given queue.  RaggedBincount<U extends TNumber> Counts the number of occurrences of each value in an integer array.  RaggedCountSparseOutput<U extends TNumber> Performs sparse-output bin counting for a ragged tensor input.  RaggedCross<T extends TType, U extends TNumber> Generates a feature cross from a list of tensors, and returns it as a RaggedTensor.  RaggedGather<T extends TNumber, U extends TType> Gather ragged slices from `params` axis `0` according to `indices`.  RaggedRange<U extends TNumber, T extends TNumber> Returns a `RaggedTensor` containing the specified sequences of numbers.  RaggedTensorFromVariant<U extends TNumber, T extends TType> Decodes a `variant` Tensor into a `RaggedTensor`.  RaggedTensorToSparse<U extends TType> Converts a `RaggedTensor` into a `SparseTensor` with the same values.  RaggedTensorToTensor<U extends TType> Create a dense tensor from a ragged tensor, possibly altering its shape.  RaggedTensorToVariant Encodes a `RaggedTensor` into a `variant` Tensor.  RaggedTensorToVariantGradient<U extends TType> Helper used to compute the gradient for `RaggedTensorToVariant`.  RandomCrop<T extends TNumber> Randomly crop `image`.  RandomDataset Creates a Dataset that returns pseudorandom numbers.  RandomGamma<U extends TNumber> Outputs random values from the Gamma distribution(s) described by alpha.  RandomGammaGrad<T extends TNumber> Computes the derivative of a Gamma random sample w.r.t.  RandomPoisson<V extends TNumber> Outputs random values from the Poisson distribution(s) described by rate.  RandomShuffle<T extends TType> Randomly shuffles a tensor along its first dimension.  RandomShuffleQueue A queue that randomizes the order of elements.  RandomStandardNormal<U extends TNumber> Outputs random values from a normal distribution.  RandomUniform<U extends TNumber> Outputs random values from a uniform distribution.  RandomUniformInt<U extends TNumber> Outputs random integers from a uniform distribution.  Range<T extends TNumber> Creates a sequence of numbers.  RangeDataset Creates a dataset with a range of values.  Rank Returns the rank of a tensor.  RawOp A base class for Op implementations that are backed by a single Operation.  ReadFile Reads and outputs the entire contents of the input filename.  ReadVariableOp<T extends TType> Reads the value of a variable.  ReaderNumRecordsProduced Returns the number of records this Reader has produced.  ReaderNumWorkUnitsCompleted Returns the number of work units this Reader has finished processing.  ReaderRead Returns the next record (key, value pair) produced by a Reader.  ReaderReadUpTo Returns up to `num_records` (key, value) pairs produced by a Reader.  ReaderReset Restore a Reader to its initial clean state.  ReaderRestoreState Restore a reader to a previously saved state.  ReaderSerializeState Produce a string tensor that encodes the state of a Reader.  Real<U extends TNumber> Returns the real part of a complex number.  RealDiv<T extends TType> Returns x / y element-wise for real types.  RebatchDataset Creates a dataset that changes the batch size.  RebatchDatasetV2 Creates a dataset that changes the batch size.  Reciprocal<T extends TType> Computes the reciprocal of x element-wise.  ReciprocalGrad<T extends TType> Computes the gradient for the inverse of `x` wrt its input.  RecordInput Emits randomized records.  Recv<T extends TType> Receives the named tensor from another XLA computation.  RecvTPUEmbeddingActivations An op that receives embedding activations on the TPU.  Reduce<T extends TNumber> Mutually reduces multiple tensors of identical type and shape.  ReduceAll Computes the "logical and" of elements across dimensions of a tensor.  ReduceAny Computes the "logical or" of elements across dimensions of a tensor.  ReduceJoin Joins a string Tensor across the given dimensions.  ReduceMax<T extends TType> Computes the maximum of elements across dimensions of a tensor.  ReduceMin<T extends TType> Computes the minimum of elements across dimensions of a tensor.  ReduceProd<T extends TType> Computes the product of elements across dimensions of a tensor.  ReduceSum<T extends TType> Computes the sum of elements across dimensions of a tensor.  ReduceV2<T extends TNumber> Mutually reduces multiple tensors of identical type and shape.  RefEnter<T extends TType> Creates or finds a child frame, and makes `data` available to the child frame.  RefExit<T extends TType> Exits the current frame to its parent frame.  RefIdentity<T extends TType> Return the same ref tensor as the input ref tensor.  RefMerge<T extends TType> Forwards the value of an available tensor from `inputs` to `output`.  RefNextIteration<T extends TType> Makes its input available to the next iteration.  RefSelect<T extends TType> Forwards the `index`th element of `inputs` to `output`.  RefSwitch<T extends TType> Forwards the ref tensor `data` to the output port determined by `pred`.  RegexFullMatch Check if the input matches the regex pattern.  RegexReplace Replaces matches of the `pattern` regular expression in `input` with the replacement string provided in `rewrite`.  RegisterDataset Registers a dataset with the tf.data service.  Relu<T extends TType> Computes rectified linear: `max(features, 0)`.  Relu6<T extends TNumber> Computes rectified linear 6: `min(max(features, 0), 6)`.  Relu6Grad<T extends TNumber> Computes rectified linear 6 gradients for a Relu6 operation.  ReluGrad<T extends TNumber> Computes rectified linear gradients for a Relu operation.  RemoteFusedGraphExecute Execute a sub graph on a remote processor.  RepeatDataset Creates a dataset that emits the outputs of `input_dataset` `count` times.  ReplicaId Replica ID.  ReplicateMetadata Metadata indicating how the TPU computation should be replicated.  ReplicatedInput<T extends TType> Connects N inputs to an N-way replicated TPU computation.  ReplicatedOutput<T extends TType> Connects N outputs from an N-way replicated TPU computation.  RequantizationRange Computes a range that covers the actual values present in a quantized tensor.  RequantizationRangePerChannel Computes requantization range per channel.  Requantize<U extends TType> Converts the quantized `input` tensor into a lower-precision `output`.  RequantizePerChannel<U extends TType> Requantizes input with min and max values known per channel.  Reshape<T extends TType> Reshapes a tensor.  ResizeArea Resize `images` to `size` using area interpolation.  ResizeBicubic Resize `images` to `size` using bicubic interpolation.  ResizeBicubicGrad<T extends TNumber> Computes the gradient of bicubic interpolation.  ResizeBilinear Resize `images` to `size` using bilinear interpolation.  ResizeBilinearGrad<T extends TNumber> Computes the gradient of bilinear interpolation.  ResizeNearestNeighbor<T extends TNumber> Resize `images` to `size` using nearest neighbor interpolation.  ResizeNearestNeighborGrad<T extends TNumber> Computes the gradient of nearest neighbor interpolation.  ResourceAccumulatorApplyGradient Applies a gradient to a given accumulator.  ResourceAccumulatorNumAccumulated Returns the number of gradients aggregated in the given accumulators.  ResourceAccumulatorSetGlobalStep Updates the accumulator with a new value for global_step.  ResourceAccumulatorTakeGradient<T extends TType> Extracts the average gradient in the given ConditionalAccumulator.  ResourceApplyAdaMax Update '*var' according to the AdaMax algorithm.  ResourceApplyAdadelta Update '*var' according to the adadelta scheme.  ResourceApplyAdagrad Update '*var' according to the adagrad scheme.  ResourceApplyAdagradDa Update '*var' according to the proximal adagrad scheme.  ResourceApplyAdam Update '*var' according to the Adam algorithm.  ResourceApplyAdamWithAmsgrad Update '*var' according to the Adam algorithm.  ResourceApplyAddSign Update '*var' according to the AddSign update.  ResourceApplyCenteredRmsProp Update '*var' according to the centered RMSProp algorithm.  ResourceApplyFtrl Update '*var' according to the Ftrl-proximal scheme.  ResourceApplyGradientDescent Update '*var' by subtracting 'alpha' * 'delta' from it.  ResourceApplyKerasMomentum Update '*var' according to the momentum scheme.  ResourceApplyMomentum Update '*var' according to the momentum scheme.  ResourceApplyPowerSign Update '*var' according to the AddSign update.  ResourceApplyProximalAdagrad Update '*var' and '*accum' according to FOBOS with Adagrad learning rate.  ResourceApplyProximalGradientDescent Update '*var' as FOBOS algorithm with fixed learning rate.  ResourceApplyRmsProp Update '*var' according to the RMSProp algorithm.  ResourceConditionalAccumulator A conditional accumulator for aggregating gradients.  ResourceCountUpTo<T extends TNumber> Increments variable pointed to by 'resource' until it reaches 'limit'.  ResourceGather<U extends TType> Gather slices from the variable pointed to by `resource` according to `indices`.  ResourceGatherNd<U extends TType>   ResourceScatterAdd Adds sparse updates to the variable referenced by `resource`.  ResourceScatterDiv Divides sparse updates into the variable referenced by `resource`.  ResourceScatterMax Reduces sparse updates into the variable referenced by `resource` using the `max` operation.  ResourceScatterMin Reduces sparse updates into the variable referenced by `resource` using the `min` operation.  ResourceScatterMul Multiplies sparse updates into the variable referenced by `resource`.  ResourceScatterNdAdd Applies sparse addition to individual values or slices in a Variable.  ResourceScatterNdMax   ResourceScatterNdMin   ResourceScatterNdSub Applies sparse subtraction to individual values or slices in a Variable.  ResourceScatterNdUpdate Applies sparse `updates` to individual values or slices within a given variable according to `indices`.  ResourceScatterSub Subtracts sparse updates from the variable referenced by `resource`.  ResourceScatterUpdate Assigns sparse updates to the variable referenced by `resource`.  ResourceSparseApplyAdadelta var: Should be from a Variable().  ResourceSparseApplyAdagrad Update relevant entries in '*var' and '*accum' according to the adagrad scheme.  ResourceSparseApplyAdagradDa Update entries in '*var' and '*accum' according to the proximal adagrad scheme.  ResourceSparseApplyAdagradV2 Update relevant entries in '*var' and '*accum' according to the adagrad scheme.  ResourceSparseApplyCenteredRmsProp Update '*var' according to the centered RMSProp algorithm.  ResourceSparseApplyFtrl Update relevant entries in '*var' according to the Ftrl-proximal scheme.  ResourceSparseApplyKerasMomentum Update relevant entries in '*var' and '*accum' according to the momentum scheme.  ResourceSparseApplyMomentum Update relevant entries in '*var' and '*accum' according to the momentum scheme.  ResourceSparseApplyProximalAdagrad Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.  ResourceSparseApplyProximalGradientDescent Sparse update '*var' as FOBOS algorithm with fixed learning rate.  ResourceSparseApplyRmsProp Update '*var' according to the RMSProp algorithm.  ResourceStridedSliceAssign Assign `value` to the sliced l-value reference of `ref`.  Restore Restores tensors from a V2 checkpoint.  RestoreSlice<T extends TType> Restores a tensor from checkpoint files.  RetrieveTPUEmbeddingADAMParameters Retrieve ADAM embedding parameters.  RetrieveTPUEmbeddingADAMParametersGradAccumDebug Retrieve ADAM embedding parameters with debug support.  RetrieveTPUEmbeddingAdadeltaParameters Retrieve Adadelta embedding parameters.  RetrieveTPUEmbeddingAdadeltaParametersGradAccumDebug Retrieve Adadelta embedding parameters with debug support.  RetrieveTPUEmbeddingAdagradParameters Retrieve Adagrad embedding parameters.  RetrieveTPUEmbeddingAdagradParametersGradAccumDebug Retrieve Adagrad embedding parameters with debug support.  RetrieveTPUEmbeddingCenteredRMSPropParameters Retrieve centered RMSProp embedding parameters.  RetrieveTPUEmbeddingFTRLParameters Retrieve FTRL embedding parameters.  RetrieveTPUEmbeddingFTRLParametersGradAccumDebug Retrieve FTRL embedding parameters with debug support.  RetrieveTPUEmbeddingMDLAdagradLightParameters Retrieve MDL Adagrad Light embedding parameters.  RetrieveTPUEmbeddingMomentumParameters Retrieve Momentum embedding parameters.  RetrieveTPUEmbeddingMomentumParametersGradAccumDebug Retrieve Momentum embedding parameters with debug support.  RetrieveTPUEmbeddingProximalAdagradParameters Retrieve proximal Adagrad embedding parameters.  RetrieveTPUEmbeddingProximalAdagradParametersGradAccumDebug Retrieve proximal Adagrad embedding parameters with debug support.  RetrieveTPUEmbeddingProximalYogiParameters   RetrieveTPUEmbeddingProximalYogiParametersGradAccumDebug   RetrieveTPUEmbeddingRMSPropParameters Retrieve RMSProp embedding parameters.  RetrieveTPUEmbeddingRMSPropParametersGradAccumDebug Retrieve RMSProp embedding parameters with debug support.  RetrieveTPUEmbeddingStochasticGradientDescentParameters Retrieve SGD embedding parameters.  RetrieveTPUEmbeddingStochasticGradientDescentParametersGradAccumDebug Retrieve SGD embedding parameters with debug support.  Reverse<T extends TType> Reverses specific dimensions of a tensor.  ReverseSequence<T extends TType> Reverses variable length slices.  Rfft<U extends TType> Real-valued fast Fourier transform.  Rfft2d<U extends TType> 2D real-valued fast Fourier transform.  Rfft3d<U extends TType> 3D real-valued fast Fourier transform.  RgbToHsv<T extends TNumber> Converts one or more images from RGB to HSV.  RightShift<T extends TNumber> Elementwise computes the bitwise right-shift of `x` and `y`.  Rint<T extends TNumber> Returns element-wise integer closest to x.  RngReadAndSkip Advance the counter of a counter-based RNG.  RngSkip Advance the counter of a counter-based RNG.  Roll<T extends TType> Rolls the elements of a tensor along an axis.  Round<T extends TType> Rounds the values of a tensor to the nearest integer, element-wise.  Rpc Perform batches of RPC requests.  Rsqrt<T extends TType> Computes reciprocal of square root of x element-wise.  RsqrtGrad<T extends TType> Computes the gradient for the rsqrt of `x` wrt its input.  SampleDistortedBoundingBox<T extends TNumber> Generate a single randomly distorted bounding box for an image.  SamplingDataset Creates a dataset that takes a Bernoulli sample of the contents of another dataset.  Save Saves tensors in V2 checkpoint format.  SaveSlices Saves input tensors slices to disk.  ScalarSummary Outputs a `Summary` protocol buffer with scalar values.  ScaleAndTranslate   ScaleAndTranslateGrad<T extends TNumber>   ScatterAdd<T extends TType> Adds sparse updates to a variable reference.  ScatterDiv<T extends TType> Divides a variable reference by sparse updates.  ScatterMax<T extends TNumber> Reduces sparse updates into a variable reference using the `max` operation.  ScatterMin<T extends TNumber> Reduces sparse updates into a variable reference using the `min` operation.  ScatterMul<T extends TType> Multiplies sparse updates into a variable reference.  ScatterNd<U extends TType> Scatter `updates` into a new tensor according to `indices`.  ScatterNdAdd<T extends TType> Applies sparse addition to individual values or slices in a Variable.  ScatterNdMax<T extends TType> Computes element-wise maximum.  ScatterNdMin<T extends TType> Computes element-wise minimum.  ScatterNdNonAliasingAdd<T extends TType> Applies sparse addition to `input` using individual values or slices from `updates` according to indices `indices`.  ScatterNdSub<T extends TType> Applies sparse subtraction to individual values or slices in a Variable.  ScatterNdUpdate<T extends TType> Applies sparse `updates` to individual values or slices within a given variable according to `indices`.  ScatterSub<T extends TType> Subtracts sparse updates to a variable reference.  ScatterUpdate<T extends TType> Applies sparse updates to a variable reference.  SdcaFprint Computes fingerprints of the input strings.  SdcaOptimizer Distributed version of Stochastic Dual Coordinate Ascent (SDCA) optimizer for linear models with L1 + L2 regularization.  SdcaShrinkL1 Applies L1 regularization shrink step on the parameters.  SegmentMax<T extends TNumber> Computes the maximum along segments of a tensor.  SegmentMean<T extends TType> Computes the mean along segments of a tensor.  SegmentMin<T extends TNumber> Computes the minimum along segments of a tensor.  SegmentProd<T extends TType> Computes the product along segments of a tensor.  SegmentSum<T extends TType> Computes the sum along segments of a tensor.  Select<T extends TType>   SelfAdjointEig<T extends TType> Computes the eigen decomposition of a batch of self-adjoint matrices (Note: Only real inputs are supported).  Selu<T extends TNumber> Computes scaled exponential linear: `scale * alpha * (exp(features) - 1)` if < 0, `scale * features` otherwise.  SeluGrad<T extends TNumber> Computes gradients for the scaled exponential linear (Selu) operation.  Send Sends the named tensor to another XLA computation.  SendTPUEmbeddingGradients Performs gradient updates of embedding tables.  SerializeIterator Converts the given `resource_handle` representing an iterator to a variant tensor.  SerializeManySparse<U extends TType> Serialize an `N`-minibatch `SparseTensor` into an `[N, 3]` `Tensor` object.  SerializeSparse<U extends TType> Serialize a `SparseTensor` into a `[3]` `Tensor` object.  SerializeTensor Transforms a Tensor into a serialized TensorProto proto.  SetDiff1d<T extends TType, U extends TNumber> Computes the difference between two lists of numbers or strings.  SetSize Number of unique elements along last dimension of input `set`.  SetStatsAggregatorDataset   Shape<U extends TNumber> Returns the shape of a tensor.  ShapeN<U extends TNumber> Returns shape of tensors.  ShardDataset Creates a `Dataset` that includes only 1/`num_shards` of this dataset.  ShardedFilename Generate a sharded filename.  ShardedFilespec Generate a glob pattern matching all sharded file names.  Sharding<T extends TType> An op which shards the input based on the given sharding attribute.  ShuffleAndRepeatDataset   ShuffleDataset   ShutdownDistributedTPU Shuts down a running distributed TPU system.  Sigmoid<T extends TType> Computes sigmoid of `x` element-wise.  SigmoidGrad<T extends TType> Computes the gradient of the sigmoid of `x` wrt its input.  Sign<T extends TType> Returns an element-wise indication of the sign of a number.  Sin<T extends TType> Computes sine of x element-wise.  Sinh<T extends TType> Computes hyperbolic sine of x element-wise.  Size<U extends TNumber> Returns the size of a tensor.  SkipDataset Creates a dataset that skips `count` elements from the `input_dataset`.  Skipgram Parses a text file and creates a batch of examples.  SleepDataset   Slice<T extends TType> Return a slice from 'input'.  SlidingWindowDataset Creates a dataset that passes a sliding window over `input_dataset`.  Snapshot<T extends TType> Returns a copy of the input tensor.  SobolSample<T extends TNumber> Generates points from the Sobol sequence.  Softmax<T extends TNumber> Computes softmax activations.  SoftmaxCrossEntropyWithLogits<T extends TNumber> Computes softmax cross entropy cost and gradients to backpropagate.  Softplus<T extends TNumber> Computes softplus: `log(exp(features) + 1)`.  SoftplusGrad<T extends TNumber> Computes softplus gradients for a softplus operation.  Softsign<T extends TNumber> Computes softsign: `features / (abs(features) + 1)`.  SoftsignGrad<T extends TNumber> Computes softsign gradients for a softsign operation.  Solve<T extends TType> Solves systems of linear equations.  Sort<T extends TType> Wraps the XLA Sort operator, documented at https://www.tensorflow.org/performance/xla/operation_semantics#sort .  SpaceToBatch<T extends TType> SpaceToBatch for 4-D tensors of type T.  SpaceToBatchNd<T extends TType> SpaceToBatch for N-D tensors of type T.  SpaceToDepth<T extends TType> SpaceToDepth for tensors of type T.  SparseAccumulatorApplyGradient Applies a sparse gradient to a given accumulator.  SparseAccumulatorTakeGradient<T extends TType> Extracts the average sparse gradient in a SparseConditionalAccumulator.  SparseAdd<T extends TType> Adds two `SparseTensor` objects to produce another `SparseTensor`.  SparseAddGrad<T extends TType> The gradient operator for the SparseAdd op.  SparseApplyAdadelta<T extends TType> var: Should be from a Variable().  SparseApplyAdagrad<T extends TType> Update relevant entries in '*var' and '*accum' according to the adagrad scheme.  SparseApplyAdagradDa<T extends TType> Update entries in '*var' and '*accum' according to the proximal adagrad scheme.  SparseApplyCenteredRmsProp<T extends TType> Update '*var' according to the centered RMSProp algorithm.  SparseApplyFtrl<T extends TType> Update relevant entries in '*var' according to the Ftrl-proximal scheme.  SparseApplyMomentum<T extends TType> Update relevant entries in '*var' and '*accum' according to the momentum scheme.  SparseApplyProximalAdagrad<T extends TType> Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.  SparseApplyProximalGradientDescent<T extends TType> Sparse update '*var' as FOBOS algorithm with fixed learning rate.  SparseApplyRmsProp<T extends TType> Update '*var' according to the RMSProp algorithm.  SparseBincount<U extends TNumber> Counts the number of occurrences of each value in an integer array.  SparseConcat<T extends TType> Concatenates a list of `SparseTensor` along the specified dimension.  SparseConditionalAccumulator A conditional accumulator for aggregating sparse gradients.  SparseCountSparseOutput<U extends TNumber> Performs sparse-output bin counting for a sparse tensor input.  SparseCross Generates sparse cross from a list of sparse and dense tensors.  SparseCrossHashed Generates sparse cross from a list of sparse and dense tensors.  SparseDenseCwiseAdd<T extends TType> Adds up a SparseTensor and a dense Tensor, using these special rules: (1) Broadcasts the dense side to have the same shape as the sparse side, if eligible; (2) Then, only the dense values pointed to by the indices of the SparseTensor participate in the cwise addition.  SparseDenseCwiseDiv<T extends TType> Component-wise divides a SparseTensor by a dense Tensor.  SparseDenseCwiseMul<T extends TType> Component-wise multiplies a SparseTensor by a dense Tensor.  SparseFillEmptyRows<T extends TType> Fills empty rows in the input 2-D `SparseTensor` with a default value.  SparseFillEmptyRowsGrad<T extends TType> The gradient of SparseFillEmptyRows.  SparseMatMul Multiply matrix "a" by matrix "b".  SparseMatrixAdd Sparse addition of two CSR matrices, C = alpha * A + beta * B.  SparseMatrixMatMul<T extends TType> Matrix-multiplies a sparse matrix with a dense matrix.  SparseMatrixMul Element-wise multiplication of a sparse matrix with a dense tensor.  SparseMatrixNNZ Returns the number of nonzeroes of `sparse_matrix`.  SparseMatrixOrderingAMD Computes the Approximate Minimum Degree (AMD) ordering of `input`.  SparseMatrixSoftmax Calculates the softmax of a CSRSparseMatrix.  SparseMatrixSoftmaxGrad Calculates the gradient of the SparseMatrixSoftmax op.  SparseMatrixSparseCholesky Computes the sparse Cholesky decomposition of `input`.  SparseMatrixSparseMatMul Sparse-matrix-multiplies two CSR matrices `a` and `b`.  SparseMatrixTranspose Transposes the inner (matrix) dimensions of a CSRSparseMatrix.  SparseMatrixZeros Creates an all-zeros CSRSparseMatrix with shape `dense_shape`.  SparseReduceMax<T extends TNumber> Computes the max of elements across dimensions of a SparseTensor.  SparseReduceMaxSparse<T extends TNumber> Computes the max of elements across dimensions of a SparseTensor.  SparseReduceSum<T extends TType> Computes the sum of elements across dimensions of a SparseTensor.  SparseReduceSumSparse<T extends TType> Computes the sum of elements across dimensions of a SparseTensor.  SparseReorder<T extends TType> Reorders a SparseTensor into the canonical, row-major ordering.  SparseReshape Reshapes a SparseTensor to represent values in a new dense shape.  SparseSegmentMean<T extends TNumber> Computes the mean along sparse segments of a tensor.  SparseSegmentMeanGrad<T extends TNumber> Computes gradients for SparseSegmentMean.  SparseSegmentMeanWithNumSegments<T extends TNumber> Computes the mean along sparse segments of a tensor.  SparseSegmentSqrtN<T extends TNumber> Computes the sum along sparse segments of a tensor divided by the sqrt of N.  SparseSegmentSqrtNGrad<T extends TNumber> Computes gradients for SparseSegmentSqrtN.  SparseSegmentSqrtNWithNumSegments<T extends TNumber> Computes the sum along sparse segments of a tensor divided by the sqrt of N.  SparseSegmentSum<T extends TNumber> Computes the sum along sparse segments of a tensor.  SparseSegmentSumWithNumSegments<T extends TNumber> Computes the sum along sparse segments of a tensor.  SparseSlice<T extends TType> Slice a `SparseTensor` based on the `start` and `size`.  SparseSliceGrad<T extends TType> The gradient operator for the SparseSlice op.  SparseSoftmax<T extends TNumber> Applies softmax to a batched N-D `SparseTensor`.  SparseSoftmaxCrossEntropyWithLogits<T extends TNumber> Computes softmax cross entropy cost and gradients to backpropagate.  SparseSparseMaximum<T extends TNumber> Returns the element-wise max of two SparseTensors.  SparseSparseMinimum<T extends TType> Returns the element-wise min of two SparseTensors.  SparseSplit<T extends TType> Split a `SparseTensor` into `num_split` tensors along one dimension.  SparseTensorDenseAdd<U extends TType> Adds up a `SparseTensor` and a dense `Tensor`, producing a dense `Tensor`.  SparseTensorDenseMatMul<U extends TType> Multiply SparseTensor (of rank 2) "A" by dense matrix "B".  SparseTensorSliceDataset Creates a dataset that splits a SparseTensor into elements row-wise.  SparseTensorToCSRSparseMatrix Converts a SparseTensor to a (possibly batched) CSRSparseMatrix.  SparseToDense<U extends TType> Converts a sparse representation into a dense tensor.  SparseToSparseSetOperation<T extends TType> Applies set operation along last dimension of 2 `SparseTensor` inputs.  Spence<T extends TNumber>   Split<T extends TType> Splits a tensor into `num_split` tensors along one dimension.  SplitV<T extends TType> Splits a tensor into `num_split` tensors along one dimension.  SqlDataset Creates a dataset that executes a SQL query and emits rows of the result set.  Sqrt<T extends TType> Computes square root of x element-wise.  SqrtGrad<T extends TType> Computes the gradient for the sqrt of `x` wrt its input.  Sqrtm<T extends TType> Computes the matrix square root of one or more square matrices: matmul(sqrtm(A), sqrtm(A)) = A The input matrix should be invertible.  Square<T extends TType> Computes square of x element-wise.  SquaredDifference<T extends TType> Returns conj(x - y)(x - y) element-wise.  Squeeze<T extends TType> Removes dimensions of size 1 from the shape of a tensor.  Stack<T extends TType> Packs a list of `N` rank-`R` tensors into one rank-`(R+1)` tensor.  Stage Stage values similar to a lightweight Enqueue.  StageClear Op removes all elements in the underlying container.  StagePeek Op peeks at the values at the specified index.  StageSize Op returns the number of elements in the underlying container.  StatefulRandomBinomial<V extends TNumber>   StatefulStandardNormal<U extends TType> Outputs random values from a normal distribution.  StatefulTruncatedNormal<U extends TType> Outputs random values from a truncated normal distribution.  StatefulUniform<U extends TType> Outputs random values from a uniform distribution.  StatefulUniformFullInt<U extends TType> Outputs random integers from a uniform distribution.  StatefulUniformInt<U extends TType> Outputs random integers from a uniform distribution.  StatelessMultinomial<V extends TNumber> Draws samples from a multinomial distribution.  StatelessParameterizedTruncatedNormal<V extends TNumber>   StatelessRandomBinomial<W extends TNumber> Outputs deterministic pseudorandom random numbers from a binomial distribution.  StatelessRandomGamma<V extends TNumber> Outputs deterministic pseudorandom random numbers from a gamma distribution.  StatelessRandomGetKeyCounterAlg Picks the best algorithm based on device, and scrambles seed into key and counter.  StatelessRandomNormal<V extends TNumber> Outputs deterministic pseudorandom values from a normal distribution.  StatelessRandomNormalV2<U extends TNumber> Outputs deterministic pseudorandom values from a normal distribution.  StatelessRandomPoisson<W extends TNumber> Outputs deterministic pseudorandom random numbers from a Poisson distribution.  StatelessRandomUniform<V extends TNumber> Outputs deterministic pseudorandom random values from a uniform distribution.  StatelessRandomUniformFullInt<V extends TNumber> Outputs deterministic pseudorandom random integers from a uniform distribution.  StatelessRandomUniformFullIntV2<U extends TNumber> Outputs deterministic pseudorandom random integers from a uniform distribution.  StatelessRandomUniformInt<V extends TNumber> Outputs deterministic pseudorandom random integers from a uniform distribution.  StatelessRandomUniformIntV2<U extends TNumber> Outputs deterministic pseudorandom random integers from a uniform distribution.  StatelessRandomUniformV2<U extends TNumber> Outputs deterministic pseudorandom random values from a uniform distribution.  StatelessSampleDistortedBoundingBox<T extends TNumber> Generate a randomly distorted bounding box for an image deterministically.  StatelessTruncatedNormal<V extends TNumber> Outputs deterministic pseudorandom values from a truncated normal distribution.  StatelessTruncatedNormalV2<U extends TNumber> Outputs deterministic pseudorandom values from a truncated normal distribution.  StaticRegexFullMatch Check if the input matches the regex pattern.  StaticRegexReplace Replaces the match of pattern in input with rewrite.  StatsAggregatorHandle   StatsAggregatorSetSummaryWriter Set a summary_writer_interface to record statistics using given stats_aggregator.  StatsAggregatorSummary Produces a summary of any statistics recorded by the given statistics manager.  StopGradient<T extends TType> Stops gradient computation.  StridedSlice<T extends TType> Return a strided slice from `input`.  StridedSliceAssign<T extends TType> Assign `value` to the sliced l-value reference of `ref`.  StridedSliceGrad<U extends TType> Returns the gradient of `StridedSlice`.  StringFormat Formats a string template using a list of tensors.  StringLength String lengths of `input`.  StringNGrams<T extends TNumber> Creates ngrams from ragged string data.  StringSplit Split elements of `source` based on `sep` into a `SparseTensor`.  Strip Strip leading and trailing whitespaces from the Tensor.  Sub<T extends TType> Returns x - y element-wise.  Substr Return substrings from `Tensor` of strings.  Sum<T extends TType> Computes the sum of elements across dimensions of a tensor.  SummaryWriter   Svd<T extends TType> Computes the eigen decomposition of a batch of self-adjoint matrices (Note: Only real inputs are supported).  SwitchCond<T extends TType> Forwards `data` to the output port determined by `pred`.  TPUCompilationResult Returns the result of a TPU compilation.  TPUEmbeddingActivations An op enabling differentiation of TPU Embeddings.  TPUReplicateMetadata Metadata indicating how the TPU computation should be replicated.  TPUReplicatedInput<T extends TType> Connects N inputs to an N-way replicated TPU computation.  TPUReplicatedOutput<T extends TType> Connects N outputs from an N-way replicated TPU computation.  TakeDataset Creates a dataset that contains `count` elements from the `input_dataset`.  TakeManySparseFromTensorsMap<T extends TType> Read `SparseTensors` from a `SparseTensorsMap` and concatenate them.  Tan<T extends TType> Computes tan of x element-wise.  Tanh<T extends TType> Computes hyperbolic tangent of `x` element-wise.  TanhGrad<T extends TType> Computes the gradient for the tanh of `x` wrt its input.  TemporaryVariable<T extends TType> Returns a tensor that may be mutated, but only persists within a single step.  TensorArray An array of Tensors of given size.  TensorArrayClose Delete the TensorArray from its resource container.  TensorArrayConcat<T extends TType> Concat the elements from the TensorArray into value `value`.  TensorArrayGather<T extends TType> Gather specific elements from the TensorArray into output `value`.  TensorArrayGrad Creates a TensorArray for storing the gradients of values in the given handle.  TensorArrayGradWithShape Creates a TensorArray for storing multiple gradients of values in the given handle.  TensorArrayPack<T extends TType>   TensorArrayRead<T extends TType> Read an element from the TensorArray into output `value`.  TensorArrayScatter Scatter the data from the input value into specific TensorArray elements.  TensorArraySize Get the current size of the TensorArray.  TensorArraySplit Split the data from the input value into TensorArray elements.  TensorArrayUnpack   TensorArrayWrite Push an element onto the tensor_array.  TensorDataset Creates a dataset that emits `components` as a tuple of tensors once.  TensorDiag<T extends TType> Returns a diagonal tensor with a given diagonal values.  TensorDiagPart<T extends TType> Returns the diagonal part of the tensor.  TensorForestCreateTreeVariable Creates a tree resource and returns a handle to it.  TensorForestTreeDeserialize Deserializes a proto into the tree handle   TensorForestTreeIsInitializedOp Checks whether a tree has been initialized.  TensorForestTreePredict Output the logits for the given input data   TensorForestTreeResourceHandleOp Creates a handle to a TensorForestTreeResource   TensorForestTreeSerialize Serializes the tree handle to a proto   TensorForestTreeSize Get the number of nodes in a tree   TensorListConcat<U extends TType> Concats all tensors in the list along the 0th dimension.  TensorListConcatLists   TensorListElementShape<T extends TNumber> The shape of the elements of the given list, as a tensor.  TensorListFromTensor Creates a TensorList which, when stacked, has the value of `tensor`.  TensorListGather<T extends TType> Creates a Tensor by indexing into the TensorList.  TensorListGetItem<T extends TType>   TensorListLength Returns the number of tensors in the input tensor list.  TensorListPopBack<T extends TType> Returns the last element of the input list as well as a list with all but that element.  TensorListPushBack Returns a list which has the passed-in `Tensor` as last element and the other elements of the given list in `input_handle`.  TensorListPushBackBatch   TensorListReserve List of the given size with empty elements.  TensorListResize Resizes the list.  TensorListScatter Creates a TensorList by indexing into a Tensor.  TensorListScatterIntoExistingList Scatters tensor at indices in an input list.  TensorListSetItem   TensorListSplit Splits a tensor into a list.  TensorListStack<T extends TType> Stacks all tensors in the list.  TensorMapErase Returns a tensor map with item from given key erased.  TensorMapHasKey Returns whether the given key exists in the map.  TensorMapInsert Returns a map that is the 'input_handle' with the given key-value pair inserted.  TensorMapLookup<U extends TType> Returns the value from a given key in a tensor map.  TensorMapSize Returns the number of tensors in the input tensor map.  TensorMapStackKeys<T extends TType> Returns a Tensor stack of all keys in a tensor map.  TensorScatterNdAdd<T extends TType> Adds sparse `updates` to an existing tensor according to `indices`.  TensorScatterNdMax<T extends TType>   TensorScatterNdMin<T extends TType>   TensorScatterNdSub<T extends TType> Subtracts sparse `updates` from an existing tensor according to `indices`.  TensorScatterNdUpdate<T extends TType> Scatter `updates` into an existing tensor according to `indices`.  TensorSliceDataset Creates a dataset that emits each dim-0 slice of `components` once.  TensorStridedSliceUpdate<T extends TType> Assign `value` to the sliced l-value reference of `input`.  TensorSummary Outputs a `Summary` protocol buffer with a tensor and per-plugin data.  TextLineDataset Creates a dataset that emits the lines of one or more text files.  TextLineReader A Reader that outputs the lines of a file delimited by '\n'.  TfRecordDataset Creates a dataset that emits the records from one or more TFRecord files.  TfRecordReader A Reader that outputs the records from a TensorFlow Records file.  ThreadPoolDataset Creates a dataset that uses a custom thread pool to compute `input_dataset`.  ThreadPoolHandle Creates a dataset that uses a custom thread pool to compute `input_dataset`.  Tile<T extends TType> Constructs a tensor by tiling a given tensor.  TileGrad<T extends TType> Returns the gradient of `Tile`.  Timestamp Provides the time since epoch in seconds.  ToBool Converts a tensor to a scalar predicate.  ToHashBucket Converts each string in the input Tensor to its hash mod by a number of buckets.  ToHashBucketFast Converts each string in the input Tensor to its hash mod by a number of buckets.  ToHashBucketStrong Converts each string in the input Tensor to its hash mod by a number of buckets.  ToNumber<T extends TNumber> Converts each string in the input Tensor to the specified numeric type.  TopK<T extends TNumber> Finds values and indices of the `k` largest elements for the last dimension.  TopKUnique Returns the TopK unique values in the array in sorted order.  TopKWithUnique Returns the TopK values in the array in sorted order.  Transpose<T extends TType> Shuffle dimensions of x according to a permutation.  TriangularSolve<T extends TType> Solves systems of linear equations with upper or lower triangular matrices by backsubstitution.  TridiagonalMatMul<T extends TType> Calculate product with tridiagonal matrix.  TridiagonalSolve<T extends TType> Solves tridiagonal systems of equations.  TruncateDiv<T extends TType> Returns x / y element-wise for integer types.  TruncateMod<T extends TNumber> Returns element-wise remainder of division.  TruncatedNormal<U extends TNumber> Outputs random values from a truncated normal distribution.  TryRpc Perform batches of RPC requests.  Unbatch<T extends TType> Reverses the operation of Batch for a single output Tensor.  UnbatchDataset A dataset that splits the elements of its input into multiple elements.  UnbatchGrad<T extends TType> Gradient of Unbatch.  UncompressElement Uncompresses a compressed dataset element.  UnicodeDecode<T extends TNumber> Decodes each string in `input` into a sequence of Unicode code points.  UnicodeDecodeWithOffsets<T extends TNumber> Decodes each string in `input` into a sequence of Unicode code points.  UnicodeEncode Encode a tensor of ints into unicode strings.  UnicodeScript Determine the script codes of a given tensor of Unicode integer code points.  UnicodeTranscode Transcode the input text from a source encoding to a destination encoding.  UniformCandidateSampler Generates labels for candidate sampling with a uniform distribution.  Unique<T extends TType, V extends TNumber> Finds unique elements along an axis of a tensor.  UniqueDataset Creates a dataset that contains the unique elements of `input_dataset`.  UniqueWithCounts<T extends TType, V extends TNumber> Finds unique elements along an axis of a tensor.  UnravelIndex<T extends TNumber> Converts an array of flat indices into a tuple of coordinate arrays.  UnsortedSegmentJoin Joins the elements of `inputs` based on `segment_ids`.  UnsortedSegmentMax<T extends TNumber> Computes the maximum along segments of a tensor.  UnsortedSegmentMin<T extends TNumber> Computes the minimum along segments of a tensor.  UnsortedSegmentProd<T extends TType> Computes the product along segments of a tensor.  UnsortedSegmentSum<T extends TType> Computes the sum along segments of a tensor.  Unstack<T extends TType> Unpacks a given dimension of a rank-`R` tensor into `num` rank-`(R-1)` tensors.  Unstage Op is similar to a lightweight Dequeue.  UnwrapDatasetVariant   Upper Converts all lowercase characters into their respective uppercase replacements.  UpperBound<U extends TNumber> Applies upper_bound(sorted_search_values, values) along each row.  VarHandleOp Creates a handle to a Variable resource.  VarIsInitializedOp Checks whether a resource handle-based variable has been initialized.  Variable<T extends TType> Holds state in the form of a tensor that persists across steps.  VariableShape<T extends TNumber> Returns the shape of the variable pointed to by `resource`.  Where Returns locations of nonzero / true values in a tensor.  WholeFileReader A Reader that outputs the entire contents of a file as a value.  WindowDataset Combines (nests of) input elements into a dataset of (nests of) windows.  WorkerHeartbeat Worker heartbeat op.  WrapDatasetVariant   WriteAudioSummary Writes an audio summary.  WriteFile Writes contents to the file at input filename.  WriteGraphSummary Writes a graph summary.  WriteHistogramSummary Writes a histogram summary.  WriteImageSummary Writes an image summary.  WriteRawProtoSummary Writes a serialized proto summary.  WriteScalarSummary Writes a scalar summary.  WriteSummary Writes a tensor summary.  Xdivy<T extends TType> Returns 0 if x == 0, and x / y otherwise, elementwise.  XlaRecvFromHost<T extends TType> An op to receive a tensor from the host.  XlaSendToHost An op to send a tensor to the host.  XlaSetBound Set a bound for the given input value as a hint to Xla compiler, returns the same value.  XlaSpmdFullToShardShape<T extends TType> An op used by XLA SPMD partitioner to switch from automatic partitioning to manual partitioning.  XlaSpmdShardToFullShape<T extends TType> An op used by XLA SPMD partitioner to switch from manual partitioning to automatic partitioning.  Xlog1py<T extends TType> Returns 0 if x == 0, and x * log1p(y) otherwise, elementwise.  Xlogy<T extends TType> Returns 0 if x == 0, and x * log(y) otherwise, elementwise.  Zeros<T extends TType> An operator creating a constant initialized with zeros of the shape given by `dims`.  ZerosLike<T extends TType> Returns a tensor of zeros with the same shape and type as x.  Zeta<T extends TNumber> Compute the Hurwitz zeta function \\(\zeta(x, q)\\).  ZipDataset Creates a dataset that zips together `input_datasets`.  erfinv<T extends TNumber>

A logical unit of computation.

Op implementations provide a strongly typed API for building and executing operations without the use of literals and indexes, as required in internal classes like Operation.

Ops can be classified under two categories:

• Raw ops target a single TensorFlow operation and are, in most cases, generated automatically from the OpDef proto definitions exposed by TensorFlow runtime library.
• Composite ops execute a series of other ops to accomplish a task logically presented as a single unit of computation.

Public Methods

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

Public Methods

public abstract ExecutionEnvironment env ()

Return the execution environment this op was created in.

public abstract Operation op ()

Return this unit of computation as a single Operation.

For a raw op, the returned value correspond to the TensorFlow operation wrapped by this op.

For a composite op, the returned value usually correspond to the last operation invoked in a chain or to a NoOp grouping one or more operations as a single unit of computation.

Returns

• an Operation